accelerometer
Archiver
Posts: 46,084
Pop quiz:
If I have only the inputs of acceleration and time, how do
I determine the time that elapses between 2 speeds (like 0 and 60 MPH)?
anyone remember the physics equation(s)? [noparse]:)[/noparse]
-Ryan
If I have only the inputs of acceleration and time, how do
I determine the time that elapses between 2 speeds (like 0 and 60 MPH)?
anyone remember the physics equation(s)? [noparse]:)[/noparse]
-Ryan
Comments
a = (vf - vi)/ t
so for 0 to 60 in 2.75 sec...
a = (60 - 0 mph) / 2.75 sec
As this leads to awkward units (miles/ hr/ sec), convert 60 mph to ft/ sec
(60 mi/hr)(5280 ft/ mi)(1 hr/ 3600 sec) = 88 ft/sec
solving for a...
a = (88 ft/ sec)/2.75 sec = 32 ft/ sec^2
And is (coincidentally...) how long it takes your car to reach 60 mph if
you drop it off a tall building (and ignore air resistance).
Nick
On Wed, 9 May 2001 rpogge@3... wrote:
> Pop quiz:
>
> If I have only the inputs of acceleration and time, how do
> I determine the time that elapses between 2 speeds (like 0 and 60 MPH)?
>
> anyone remember the physics equation(s)? [noparse]:)[/noparse]
>
> -Ryan
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
Acceleration = (Final Speed - Initial Speed)/Time Elapsed.
hope it helps
Vetri
If acceleration is constant in m/sec**2 then multiplying by time in sec
gives you m/sec which is velocity.
So ignoring terminal velocity and other weird issues, if you drop a ball on
Earth, it drops at 9.81m/sec**2
So if the ball was at rest at t=0 then at t=2 it will be going 19.62m/sec
and at t=3 it goes 29.43m/sec (again, ignore air resistance, etc.).
If the ball was already moving at, say, 5m/sec (v0) at t=0 then you have
9.81 * t + v0.
So you could solve algebraically.
9.81m/sec**2 * X + v0 = 33m/sec
Solve for X. Of course if you want miles per hour, you have to have your
acceleration in miles/hour**2 or convert as appropriate one way or the
other.
Disclaimer: been a long time since I've been in a physics class.
Also, keep in mind that this is for constant acceleration, which does not
apply in all cases. If your acceleration is a curve then you have to start
doing calculus to model that (acceleration is the slope of the velocity, so
the derivative of the velocity is acceleration and the integral of
acceleration is velocity).
Regards,
Al Williams
AWC
* 50% off sale in progress: http://www.al-williams.com/awce
>
Original Message
> From: rpogge@3... [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=oD3IpGg_6VGSGpCm8B10HYw7EBSn-4qCiTbT0-gE77DzgnB5Vcu5DsL-Q4KxRzGvT74pLH3V5Ho]rpogge@3...[/url
> Sent: Wednesday, May 09, 2001 10:12 AM
> To: basicstamps@yahoogroups.com
> Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
>
>
> Pop quiz:
>
> If I have only the inputs of acceleration and time, how do
> I determine the time that elapses between 2 speeds (like 0 and 60 MPH)?
>
> anyone remember the physics equation(s)? [noparse]:)[/noparse]
>
> -Ryan
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>If I have only the inputs of acceleration and time, how do
>I determine the time that elapses between 2 speeds (like 0 and 60 MPH)?
>
>anyone remember the physics equation(s)? [noparse]:)[/noparse]
>
>-Ryan
Hi Ryan,
Integrate acceleration over time. In the computer, that will mean
taking samples of the acceleration at, say, one second increments,
and adding them up, second by second.
velocity(t) = velocity(t-1) + [noparse][[/noparse]accel(t) * oneSecond]
Of course (caveat emptor) this can lead to serious errors over a long
time period, depending on accuracy of the measurements, accuracy of
the clock, and accuracy of the discrete time calculations. If the
acceleration is steep, you'd do better to average the acceleration
for each second.
velocity(t) = velocity(t-1)
+ [noparse][[/noparse](accel(t)+accel(t-1))/2 * oneSecond]
Better yet, take the samples every 0.1 second, say, then divide the
velocity by 10 at the end.
'BS2 integrator
' enter with initial velocity
' and initial accel0
loop:
gosub waitforsecond ' do this by the clock
seconds=seconds+1
gosub getacceleration ' current value accel1
velocity = (accel0+accel1)/2 + velocity
accel0=accel1
if velocity<60 then loop ' grab the target time
debug ? seconds
end
If acceleration is constant, you can use
v = 1/2 a t^2.
-- regards,
Tracy Allen
electronically monitored ecosystems
mailto:tracy@e...
http://www.emesystems.com
I have seen the units advertised in magazines,
they appear to have no interface to the car at all other than
the cigiret lighter for power, so I assume it is using an
accelerometer chip, and timer to take readings. perhapse
this is a bit to complicated for a stamp?
>If your acceleration is a curve then you have to
> start
> doing calculus to model that (acceleration is the slope of the velocity,
> so
> the derivative of the velocity is acceleration and the integral of
> acceleration is velocity).
>
> Regards,
>
> Al Williams
> AWC
> * 50% off sale in progress: http://www.al-williams.com/awce
>
> >
Original Message
> > From: rpogge@3... [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=ym35auqPnAYxIFNNpAlUwuzCppJDp_hSrg2Uh7RMetGP3H7roV3w3ug0ANLHHsjLfWMWjBsYYA]rpogge@3...[/url
> > Sent: Wednesday, May 09, 2001 10:12 AM
> > To: basicstamps@yahoogroups.com
> > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> >
> >
> > Pop quiz:
> >
> > If I have only the inputs of acceleration and time, how do
> > I determine the time that elapses between 2 speeds (like 0 and 60
> MPH)?
> >
> > anyone remember the physics equation(s)? [noparse]:)[/noparse]
> >
> > -Ryan
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>
It will be very steep and over short periods of time-
aproxamatly 4 seconds for 0-60 MPH, and I would also like 0-100 MPH.
I need accuracy to tenths of a second, so I will go in incrememts of 0.1
seconds instead of 1 second, I would realy liek it to be acurate to hundredths,
but I wonder if the amount of error at this level would make that pointless.
-Ryan
> >Pop quiz:
> >
> >If I have only the inputs of acceleration and time, how do
> >I determine the time that elapses between 2 speeds (like 0 and 60
> MPH)?
> >
> >anyone remember the physics equation(s)? [noparse]:)[/noparse]
> >
> >-Ryan
>
> Hi Ryan,
>
> Integrate acceleration over time. In the computer, that will mean
> taking samples of the acceleration at, say, one second increments,
> and adding them up, second by second.
> velocity(t) = velocity(t-1) + [noparse][[/noparse]accel(t) * oneSecond]
> Of course (caveat emptor) this can lead to serious errors over a long
> time period, depending on accuracy of the measurements, accuracy of
> the clock, and accuracy of the discrete time calculations. If the
> acceleration is steep, you'd do better to average the acceleration
> for each second.
> velocity(t) = velocity(t-1)
> + [noparse][[/noparse](accel(t)+accel(t-1))/2 * oneSecond]
> Better yet, take the samples every 0.1 second, say, then divide the
> velocity by 10 at the end.
>
>
> 'BS2 integrator
> ' enter with initial velocity
> ' and initial accel0
> loop:
> gosub waitforsecond ' do this by the clock
> seconds=seconds+1
> gosub getacceleration ' current value accel1
> velocity = (accel0+accel1)/2 + velocity
> accel0=accel1
> if velocity<60 then loop ' grab the target time
> debug ? seconds
> end
>
>
> If acceleration is constant, you can use
> v = 1/2 a t^2.
>
> -- regards,
> Tracy Allen
> electronically monitored ecosystems
> mailto:tracy@e...
> http://www.emesystems.com
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>
>Well it will be on a curve - its for a car.
>I have seen the units advertised in magazines,
>they appear to have no interface to the car at all other than
>the cigiret lighter for power, so I assume it is using an
>accelerometer chip, and timer to take readings. perhapse
>this is a bit to complicated for a stamp?
>...
>It will be very steep and over short periods of time-
>aproxamatly 4 seconds for 0-60 MPH, and I would also like 0-100 MPH.
>I need accuracy to tenths of a second, so I will go in incrememts of 0.1
>seconds instead of 1 second, I would realy liek it to be acurate to
>hundredths,
>but I wonder if the amount of error at this level would make that pointless.
What is the interface to the unit? The devil is in the details.. If
the unit delivers acelleration readings at 10 millisecond intervals
(timed by the module) then the stamp program could be very simple and
could possibly track at 0.01 second.
-- Tracy
the 202 which is PWM output. Im not so great at PCB design so the ADC
readings are never very stable for me.
-Ryan
>What is the interface to the unit? The devil is in the details.. If
>the unit delivers acelleration readings at 10 millisecond intervals
>(timed by the module) then the stamp program could be very simple and
>could possibly track at 0.01 second.
>
> -- Tracy
>
>
>
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
Note the Analog Devices application notes about distinguishing tilt from
linear acceleration, orientation for best accuracy, etc.
I'm starting a major project using the 202e, which is a chip scale version,
combined with some very small rotational velocity sensors, so contact me
later if you run into problems. My goal is to build a very small package
for monitoring human movements, for a commercial venture. It will require
some miniaturization using chip-on-flex and flip-chip designs, working with
a company that designs ultra-small in-the-ear hearing aids.
Dennis
Original Message
From: ryan pogge [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=xs213xh8fON2rVlwW5jO5EKfkOh6u3Te13chnrSFhaVL2MhepLtaqeZ7rKYDWPAo00Rr7l8]rpogge@3...[/url
Sent: Thursday, May 10, 2001 6:07 AM
To: basicstamps@yahoogroups.com
Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
one of the ADXL accelerometer chips, either the one that needs an ADC, or
the 202 which is PWM output. Im not so great at PCB design so the ADC
readings are never very stable for me.
-Ryan
>What is the interface to the unit? The devil is in the details.. If
>the unit delivers acelleration readings at 10 millisecond intervals
>(timed by the module) then the stamp program could be very simple and
>could possibly track at 0.01 second.
>
> -- Tracy
>
>
>
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
0-60 0-100, latteral G's(cornering), 60-0 braking,
quarter mile times and so fourth. my first goal is just to
get a 0-60 reading. should be about 4.1 sec (=
My concern is that the front end of the car rises
slightly during 'take off' (=
Any ideas on overcomming this?
-Ryan
> Note the Analog Devices application notes about distinguishing tilt
> from
> linear acceleration, orientation for best accuracy, etc.
>
> I'm starting a major project using the 202e, which is a chip scale
> version,
> combined with some very small rotational velocity sensors, so contact
> me
> later if you run into problems. My goal is to build a very small
> package
> for monitoring human movements, for a commercial venture. It will
> require
> some miniaturization using chip-on-flex and flip-chip designs, working
> with
> a company that designs ultra-small in-the-ear hearing aids.
>
> Dennis
>
>
Original Message
> From: ryan pogge [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=mamrBq5WzS6_hRQ3T6fBsbz8GECRP4u9EfVIWgDZepS3LKX9VkT7dTn1_rmhD45XliPmZkY5Sg]rpogge@3...[/url
> Sent: Thursday, May 10, 2001 6:07 AM
> To: basicstamps@yahoogroups.com
> Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
>
>
> one of the ADXL accelerometer chips, either the one that needs an ADC,
> or
> the 202 which is PWM output. Im not so great at PCB design so the ADC
> readings are never very stable for me.
> -Ryan
>
> >What is the interface to the unit? The devil is in the details.. If
> >the unit delivers acelleration readings at 10 millisecond intervals
> >(timed by the module) then the stamp program could be very simple and
> >could possibly track at 0.01 second.
> >
> > -- Tracy
> >
> >
> >
> >
> >Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> >
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>
0-60 in 4.1 sec? Nice to know someone on this list can afford a Porsche
twin turbo.
If you mean overcoming front end lift of the car, I'm way out of my league.
But if you want to separate 'tilt' of the car from linear acceleration, that
can be done by using 2 sensors. Ideally, one should be a strictly
rotational sensor, to monitor angular degrees of lift during takeoff. Then,
multiply the linear sensor's output by the cosine of the lift angle to get
the linear acceleration. By Einstein's equivalence principal, it's not
possible to distinguish gravitational 'pull', due to tilting, from linear
acceleration, using only one linear accelerometer. However, Analog Devices
application notes shows how to add some filtering to the ADXL series, to
make them AC-coupled for 'vibration sensing', which minimizes static tilt.
Dennis
Original Message
From: rpogge@3... [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=MDRJweMZRvZky2e92Ghv1NbrHu9azinoU7ZCWEAAps0KN2zFVdxomJK0qd_iFSuawz2q8sc]rpogge@3...[/url
Sent: Thursday, May 10, 2001 12:37 PM
To: basicstamps@yahoogroups.com
Subject: RE: [noparse][[/noparse]basicstamps] accelerometer
I'm interested in measuring different aspects of car performance,
0-60 0-100, latteral G's(cornering), 60-0 braking,
quarter mile times and so fourth. my first goal is just to
get a 0-60 reading. should be about 4.1 sec (=
My concern is that the front end of the car rises
slightly during 'take off' (=
Any ideas on overcomming this?
-Ryan
Volkswagon employed special suspension devices in their Corrado line
(and hopefully we'll see 'em again in the Steppenwolf) to decrease the
effects of this. It didn't hurt that it also featured a variable spoiler
on the back, but I don't think it affected the lift during initial
acceleration (I think I remember it shifting twice, once at 45 and again
at 140).
My first question would be, where is the engine seated? If it's a front
mounted engine and you're having lift problems, it might be a totally
different scenario than if your engine is sitting above the rear wheels.
Also, are your rear wheels skidding at all during acceleration? And
lastly, is this a front wheel or rear wheel drive vehicle? If it's a
front wheel drive vehicle and you're seeing lift, eh, you may have a
whole other series of problems :-)
(A hotrod junkie friend of mine tells me that the easiest way to
overcome this is to use two different engines, one for the front wheels
and one for the back. Then he started talking about gear ratios and
efficient torque balancing or something and I fell asleep.)
If you *really* need a professional answer to this, I live 4 miles away
from a race track, and I'm sure I can find someone out there who would
know. Next month we have the open Ferrari races, so I'm always looking
for a reason to tell someone, "It'd be easier if I just show you what I
mean. Can I have your keys?"
-Chilton
> My concern is that the front end of the car rises
> slightly during 'take off' (=
> Any ideas on overcomming this?
>
the sine of 0.9 degrees is 0.01
It looks like less than 1% error
richard
Original Message
From: "ryan" <rpogge@3...>
To: <basicstamps@yahoogroups.com>
Sent: Friday, May 11, 2001 11:39 AM
Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> Chilton-
> Its a Dodge Viper, there is VERY little lift.
> but I wonder if even an inch or two would effect the reading enough to
throw
> the results. yes the rear wheels slip durring acceleration, now that I
think
> about it, this
> also causes a 'jerk' of forward acceleration when they catch again.
> ....another problem to overcome.
>
> I mean maybee its not even an issue, ...unless I decide to use it with the
> 66 vett.
> front end of that lifts so much I loose sight of the road (=
>
> I wanted an elegant solution, however I may resort to just taking readings
> from
> the speedometer sensor instead of an accelerometer, this would simplify
> things
> a great deal.
>
> Thanks for all the input guys and gals.
>
> -Ryan
>
> > Hey there SpeedRacer Ryan,
> >
> > Volkswagon employed special suspension devices in their Corrado line
> > (and hopefully we'll see 'em again in the Steppenwolf) to decrease the
> > effects of this. It didn't hurt that it also featured a variable spoiler
> > on the back, but I don't think it affected the lift during initial
> > acceleration (I think I remember it shifting twice, once at 45 and again
> > at 140).
> >
> > My first question would be, where is the engine seated? If it's a front
> > mounted engine and you're having lift problems, it might be a totally
> > different scenario than if your engine is sitting above the rear wheels.
> > Also, are your rear wheels skidding at all during acceleration? And
> > lastly, is this a front wheel or rear wheel drive vehicle? If it's a
> > front wheel drive vehicle and you're seeing lift, eh, you may have a
> > whole other series of problems :-)
> >
> > (A hotrod junkie friend of mine tells me that the easiest way to
> > overcome this is to use two different engines, one for the front wheels
> > and one for the back. Then he started talking about gear ratios and
> > efficient torque balancing or something and I fell asleep.)
> >
> > If you *really* need a professional answer to this, I live 4 miles away
> > from a race track, and I'm sure I can find someone out there who would
> > know. Next month we have the open Ferrari races, so I'm always looking
> > for a reason to tell someone, "It'd be easier if I just show you what I
> > mean. Can I have your keys?"
> >
> > -Chilton
> >
> > > My concern is that the front end of the car rises
> > > slightly during 'take off' (=
> > > Any ideas on overcomming this?
> > >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
range - of course total error is based on range
the 1% I had reference to was 1% of the acceleration value - not time
I dont know how the 1% could ever be related to time - presumably your time
measurement is based on the accuracy of your clock or crystal - If it is a
microprocessor then a thousandth of a second is easly achieved and maybe
even could get millionth of a second.
if you want to know acceleration then accuracy is 1%
(a=a)
if you integrate once to get velocity the accuracy loss due to
accelerometer data is still 1% (v=a*t)
if you integrate twice to get distance the accuracy loss due to
accelerometer data is still 1%
(x=0.5*a*t^2)
for example, If you skid your car to stop and you measured 0.85 g then you
could calculate the velocity change with respect to time (v=0.85*32.2*t) if
it took 4 seconds for the skid then the starting velocity would have been
109 ft/sec (74 mph).
The distance traveled would have been (x=0.5*0.85*32.2*4^2) 219 feet.
The computer would measure the time to something like 3.9999987 sec.
The 1% error on 74 mph would amount to 73.3 mph
The 1% error on 219 feet would amount to 217 feet
Your program would do a similar calculation 100 times a second using a
delta-t rather than total time
performance computer manufacturers like gforce and vc-200, 10 years ago,
just used single accelerometer to measure acceleration and then made a
correction to the data to account for "dive" on braking or "lift" on
acceleration - this brute force correction attempt was somewhat mitigated by
their software asking if the vehicle had stiff suspension (less dive) or
luxury car suspension (more dive). Then their program added or subtracted 1%
or 1/2% to the measured data.
The right way to do it (and maybe they do it that way now) is to use two
accelerometers one measures the acceleration and the others basicallly
monitors the tilt of the vehicle. The program could then correctly compute
the actual acceleration of the vehicle regardless of the lift or dive. In
all cases it would be good to to get the best accelerometer you can afford
with the lowest coss-axis sensitivity possible. The $20 Analog Devices brand
work great - I have used them.
For the accuracy you need I would ignore all of the other effects (assuming
12 bit a/d making 1000 samples per second, using 0-2 g accelerometer)
richard
Original Message
From: "ryan" <rpogge@3...>
To: <basicstamps@yahoogroups.com>
Sent: Friday, May 11, 2001 12:23 PM
Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> Ah but you are assuming that the car can produce 1 G or more of
> acceleration. What if the car only produces only 0.5G's(i have no idea)
then
> you have more like 2% error.
> beacause you will be adding (1deg/90deg * 1G(gravity)) + 0.5G(from car).
>
> And that is assuming that you only have an inch of lift. Then you have
to
> take into consideration the error from accelerometer measurments, time
> measurments, the math(averaging) , the error that I will get when the
wheels
> slip and create acceleration 'spikes', vibration, sun spots, women drivers
>
>
> and a 2% error on a 5 second run is 0.1 seconds. and I want acuracy to the
> 0.1 at least, 0.01 would be ideal.
> -i.e. being able to distinguish a 4.1 second time, from a 4.2 second time
is
> minimal requirment, being able to distinguish between 4.11 and 4.19
second
> runs is ideal.
>
> I guess it is imposible to get that kind of accuracy from something
onboard.
> need to set up lasers on the racetrack if i want accuracy. *shrug*
>
> thanks for all the input
> Ciao
> -Ryan
>
> > one inch lift on a 8 foot wheelbase is about 0.9 degree change
> > the sine of 0.9 degrees is 0.01
> > It looks like less than 1% error
> >
> >
> > richard
> >
> >
> >
> >
Original Message
> > From: "ryan" <rpogge@3...>
> > To: <basicstamps@yahoogroups.com>
> > Sent: Friday, May 11, 2001 11:39 AM
> > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> >
> >
> > > Chilton-
> > > Its a Dodge Viper, there is VERY little lift.
> > > but I wonder if even an inch or two would effect the reading enough to
> > throw
> > > the results. yes the rear wheels slip durring acceleration, now that I
> > think
> > > about it, this
> > > also causes a 'jerk' of forward acceleration when they catch again.
> > > ....another problem to overcome.
> > >
> > > I mean maybee its not even an issue, ...unless I decide to use it with
> the
> > > 66 vett.
> > > front end of that lifts so much I loose sight of the road (=
> > >
> > > I wanted an elegant solution, however I may resort to just taking
> readings
> > > from
> > > the speedometer sensor instead of an accelerometer, this would
simplify
> > > things
> > > a great deal.
> > >
> > > Thanks for all the input guys and gals.
> > >
> > > -Ryan
> > >
> > > > Hey there SpeedRacer Ryan,
> > > >
> > > > Volkswagon employed special suspension devices in their Corrado line
> > > > (and hopefully we'll see 'em again in the Steppenwolf) to decrease
the
> > > > effects of this. It didn't hurt that it also featured a variable
> spoiler
> > > > on the back, but I don't think it affected the lift during initial
> > > > acceleration (I think I remember it shifting twice, once at 45 and
> again
> > > > at 140).
> > > >
> > > > My first question would be, where is the engine seated? If it's a
> front
> > > > mounted engine and you're having lift problems, it might be a
totally
> > > > different scenario than if your engine is sitting above the rear
> wheels.
> > > > Also, are your rear wheels skidding at all during acceleration? And
> > > > lastly, is this a front wheel or rear wheel drive vehicle? If it's a
> > > > front wheel drive vehicle and you're seeing lift, eh, you may have a
> > > > whole other series of problems :-)
> > > >
> > > > (A hotrod junkie friend of mine tells me that the easiest way to
> > > > overcome this is to use two different engines, one for the front
> wheels
> > > > and one for the back. Then he started talking about gear ratios and
> > > > efficient torque balancing or something and I fell asleep.)
> > > >
> > > > If you *really* need a professional answer to this, I live 4 miles
> away
> > > > from a race track, and I'm sure I can find someone out there who
would
> > > > know. Next month we have the open Ferrari races, so I'm always
looking
> > > > for a reason to tell someone, "It'd be easier if I just show you
what
> I
> > > > mean. Can I have your keys?"
> > > >
> > > > -Chilton
> > > >
> > > > > My concern is that the front end of the car rises
> > > > > slightly during 'take off' (=
> > > > > Any ideas on overcomming this?
> > > > >
> > > >
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
Its a Dodge Viper, there is VERY little lift.
but I wonder if even an inch or two would effect the reading enough to throw
the results. yes the rear wheels slip durring acceleration, now that I think
about it, this
also causes a 'jerk' of forward acceleration when they catch again.
....another problem to overcome.
I mean maybee its not even an issue, ...unless I decide to use it with the
66 vett.
front end of that lifts so much I loose sight of the road (=
I wanted an elegant solution, however I may resort to just taking readings
from
the speedometer sensor instead of an accelerometer, this would simplify
things
a great deal.
Thanks for all the input guys and gals.
-Ryan
> Hey there SpeedRacer Ryan,
>
> Volkswagon employed special suspension devices in their Corrado line
> (and hopefully we'll see 'em again in the Steppenwolf) to decrease the
> effects of this. It didn't hurt that it also featured a variable spoiler
> on the back, but I don't think it affected the lift during initial
> acceleration (I think I remember it shifting twice, once at 45 and again
> at 140).
>
> My first question would be, where is the engine seated? If it's a front
> mounted engine and you're having lift problems, it might be a totally
> different scenario than if your engine is sitting above the rear wheels.
> Also, are your rear wheels skidding at all during acceleration? And
> lastly, is this a front wheel or rear wheel drive vehicle? If it's a
> front wheel drive vehicle and you're seeing lift, eh, you may have a
> whole other series of problems :-)
>
> (A hotrod junkie friend of mine tells me that the easiest way to
> overcome this is to use two different engines, one for the front wheels
> and one for the back. Then he started talking about gear ratios and
> efficient torque balancing or something and I fell asleep.)
>
> If you *really* need a professional answer to this, I live 4 miles away
> from a race track, and I'm sure I can find someone out there who would
> know. Next month we have the open Ferrari races, so I'm always looking
> for a reason to tell someone, "It'd be easier if I just show you what I
> mean. Can I have your keys?"
>
> -Chilton
>
> > My concern is that the front end of the car rises
> > slightly during 'take off' (=
> > Any ideas on overcomming this?
> >
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
acceleration. What if the car only produces only 0.5G's(i have no idea) then
you have more like 2% error.
beacause you will be adding (1deg/90deg * 1G(gravity)) + 0.5G(from car).
And that is assuming that you only have an inch of lift. Then you have to
take into consideration the error from accelerometer measurments, time
measurments, the math(averaging) , the error that I will get when the wheels
slip and create acceleration 'spikes', vibration, sun spots, women drivers
and a 2% error on a 5 second run is 0.1 seconds. and I want acuracy to the
0.1 at least, 0.01 would be ideal.
-i.e. being able to distinguish a 4.1 second time, from a 4.2 second time is
minimal requirment, being able to distinguish between 4.11 and 4.19 second
runs is ideal.
I guess it is imposible to get that kind of accuracy from something onboard.
need to set up lasers on the racetrack if i want accuracy. *shrug*
thanks for all the input
Ciao
-Ryan
> one inch lift on a 8 foot wheelbase is about 0.9 degree change
> the sine of 0.9 degrees is 0.01
> It looks like less than 1% error
>
>
> richard
>
>
>
>
Original Message
> From: "ryan" <rpogge@3...>
> To: <basicstamps@yahoogroups.com>
> Sent: Friday, May 11, 2001 11:39 AM
> Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
>
>
> > Chilton-
> > Its a Dodge Viper, there is VERY little lift.
> > but I wonder if even an inch or two would effect the reading enough to
> throw
> > the results. yes the rear wheels slip durring acceleration, now that I
> think
> > about it, this
> > also causes a 'jerk' of forward acceleration when they catch again.
> > ....another problem to overcome.
> >
> > I mean maybee its not even an issue, ...unless I decide to use it with
the
> > 66 vett.
> > front end of that lifts so much I loose sight of the road (=
> >
> > I wanted an elegant solution, however I may resort to just taking
readings
> > from
> > the speedometer sensor instead of an accelerometer, this would simplify
> > things
> > a great deal.
> >
> > Thanks for all the input guys and gals.
> >
> > -Ryan
> >
> > > Hey there SpeedRacer Ryan,
> > >
> > > Volkswagon employed special suspension devices in their Corrado line
> > > (and hopefully we'll see 'em again in the Steppenwolf) to decrease the
> > > effects of this. It didn't hurt that it also featured a variable
spoiler
> > > on the back, but I don't think it affected the lift during initial
> > > acceleration (I think I remember it shifting twice, once at 45 and
again
> > > at 140).
> > >
> > > My first question would be, where is the engine seated? If it's a
front
> > > mounted engine and you're having lift problems, it might be a totally
> > > different scenario than if your engine is sitting above the rear
wheels.
> > > Also, are your rear wheels skidding at all during acceleration? And
> > > lastly, is this a front wheel or rear wheel drive vehicle? If it's a
> > > front wheel drive vehicle and you're seeing lift, eh, you may have a
> > > whole other series of problems :-)
> > >
> > > (A hotrod junkie friend of mine tells me that the easiest way to
> > > overcome this is to use two different engines, one for the front
wheels
> > > and one for the back. Then he started talking about gear ratios and
> > > efficient torque balancing or something and I fell asleep.)
> > >
> > > If you *really* need a professional answer to this, I live 4 miles
away
> > > from a race track, and I'm sure I can find someone out there who would
> > > know. Next month we have the open Ferrari races, so I'm always looking
> > > for a reason to tell someone, "It'd be easier if I just show you what
I
> > > mean. Can I have your keys?"
> > >
> > > -Chilton
> > >
> > > > My concern is that the front end of the car rises
> > > > slightly during 'take off' (=
> > > > Any ideas on overcomming this?
> > > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
What ADC would you recomend,
I usualy have problems with ADC's I can never get a stable reading, I was
told it was /bc I was breadboarding it, and it needed a properly designed
PCB with good ground planes etc....
I have a couple of ADXL chips, the ADXL 150, 250, and 202. You refer to a
+-2g chip the only one I see that is +-2 is the 202 which is duty cycle
output. - I belive I could just use
pulsin to read it. What is the advantage of useing one of the others over a
PWM output chip? Analog sells a nice little eval board for the 202 .. might
check it out.
Thanks
-Ryan
Original Message
> 1% error due to angle is 1% error - does not have anything to do with
> range - of course total error is based on range
>
> the 1% I had reference to was 1% of the acceleration value - not time
>
> I dont know how the 1% could ever be related to time - presumably your
time
> measurement is based on the accuracy of your clock or crystal - If it is a
> microprocessor then a thousandth of a second is easly achieved and maybe
> even could get millionth of a second.
>
> if you want to know acceleration then accuracy is 1%
> (a=a)
> if you integrate once to get velocity the accuracy loss due to
> accelerometer data is still 1% (v=a*t)
> if you integrate twice to get distance the accuracy loss due to
> accelerometer data is still 1%
> (x=0.5*a*t^2)
>
> for example, If you skid your car to stop and you measured 0.85 g then
you
> could calculate the velocity change with respect to time (v=0.85*32.2*t)
if
> it took 4 seconds for the skid then the starting velocity would have been
> 109 ft/sec (74 mph).
>
> The distance traveled would have been (x=0.5*0.85*32.2*4^2) 219 feet.
>
> The computer would measure the time to something like 3.9999987 sec.
>
> The 1% error on 74 mph would amount to 73.3 mph
> The 1% error on 219 feet would amount to 217 feet
>
> Your program would do a similar calculation 100 times a second using a
> delta-t rather than total time
>
> performance computer manufacturers like gforce and vc-200, 10 years ago,
> just used single accelerometer to measure acceleration and then made a
> correction to the data to account for "dive" on braking or "lift" on
> acceleration - this brute force correction attempt was somewhat mitigated
by
> their software asking if the vehicle had stiff suspension (less dive) or
> luxury car suspension (more dive). Then their program added or subtracted
1%
> or 1/2% to the measured data.
>
> The right way to do it (and maybe they do it that way now) is to use two
> accelerometers one measures the acceleration and the others basicallly
> monitors the tilt of the vehicle. The program could then correctly compute
> the actual acceleration of the vehicle regardless of the lift or dive. In
> all cases it would be good to to get the best accelerometer you can afford
> with the lowest coss-axis sensitivity possible. The $20 Analog Devices
brand
> work great - I have used them.
>
> For the accuracy you need I would ignore all of the other effects
(assuming
> 12 bit a/d making 1000 samples per second, using 0-2 g accelerometer)
>
> richard
>
>
>
>
>
>
Original Message
> From: "ryan" <rpogge@3...>
> To: <basicstamps@yahoogroups.com>
> Sent: Friday, May 11, 2001 12:23 PM
> Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
>
>
> > Ah but you are assuming that the car can produce 1 G or more of
> > acceleration. What if the car only produces only 0.5G's(i have no idea)
> then
> > you have more like 2% error.
> > beacause you will be adding (1deg/90deg * 1G(gravity)) + 0.5G(from car).
> >
> > And that is assuming that you only have an inch of lift. Then you have
> to
> > take into consideration the error from accelerometer measurments, time
> > measurments, the math(averaging) , the error that I will get when the
> wheels
> > slip and create acceleration 'spikes', vibration, sun spots, women
drivers
> >
> >
> > and a 2% error on a 5 second run is 0.1 seconds. and I want acuracy to
the
> > 0.1 at least, 0.01 would be ideal.
> > -i.e. being able to distinguish a 4.1 second time, from a 4.2 second
time
> is
> > minimal requirment, being able to distinguish between 4.11 and 4.19
> second
> > runs is ideal.
> >
> > I guess it is imposible to get that kind of accuracy from something
> onboard.
> > need to set up lasers on the racetrack if i want accuracy. *shrug*
> >
> > thanks for all the input
> > Ciao
> > -Ryan
> >
> > > one inch lift on a 8 foot wheelbase is about 0.9 degree change
> > > the sine of 0.9 degrees is 0.01
> > > It looks like less than 1% error
> > >
> > >
> > > richard
> > >
> > >
> > >
> > >
Original Message
> > > From: "ryan" <rpogge@3...>
> > > To: <basicstamps@yahoogroups.com>
> > > Sent: Friday, May 11, 2001 11:39 AM
> > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > >
> > >
> > > > Chilton-
> > > > Its a Dodge Viper, there is VERY little lift.
> > > > but I wonder if even an inch or two would effect the reading enough
to
> > > throw
> > > > the results. yes the rear wheels slip durring acceleration, now that
I
> > > think
> > > > about it, this
> > > > also causes a 'jerk' of forward acceleration when they catch again.
> > > > ....another problem to overcome.
> > > >
> > > > I mean maybee its not even an issue, ...unless I decide to use it
with
> > the
> > > > 66 vett.
> > > > front end of that lifts so much I loose sight of the road (=
> > > >
> > > > I wanted an elegant solution, however I may resort to just taking
> > readings
> > > > from
> > > > the speedometer sensor instead of an accelerometer, this would
> simplify
> > > > things
> > > > a great deal.
> > > >
> > > > Thanks for all the input guys and gals.
> > > >
> > > > -Ryan
> > > >
> > > > > Hey there SpeedRacer Ryan,
> > > > >
> > > > > Volkswagon employed special suspension devices in their Corrado
line
> > > > > (and hopefully we'll see 'em again in the Steppenwolf) to decrease
> the
> > > > > effects of this. It didn't hurt that it also featured a variable
> > spoiler
> > > > > on the back, but I don't think it affected the lift during initial
> > > > > acceleration (I think I remember it shifting twice, once at 45 and
> > again
> > > > > at 140).
> > > > >
> > > > > My first question would be, where is the engine seated? If it's a
> > front
> > > > > mounted engine and you're having lift problems, it might be a
> totally
> > > > > different scenario than if your engine is sitting above the rear
> > wheels.
> > > > > Also, are your rear wheels skidding at all during acceleration?
And
> > > > > lastly, is this a front wheel or rear wheel drive vehicle? If it's
a
> > > > > front wheel drive vehicle and you're seeing lift, eh, you may have
a
> > > > > whole other series of problems :-)
> > > > >
> > > > > (A hotrod junkie friend of mine tells me that the easiest way to
> > > > > overcome this is to use two different engines, one for the front
> > wheels
> > > > > and one for the back. Then he started talking about gear ratios
and
> > > > > efficient torque balancing or something and I fell asleep.)
> > > > >
> > > > > If you *really* need a professional answer to this, I live 4 miles
> > away
> > > > > from a race track, and I'm sure I can find someone out there who
> would
> > > > > know. Next month we have the open Ferrari races, so I'm always
> looking
> > > > > for a reason to tell someone, "It'd be easier if I just show you
> what
> > I
> > > > > mean. Can I have your keys?"
> > > > >
> > > > > -Chilton
> > > > >
> > > > > > My concern is that the front end of the car rises
> > > > > > slightly during 'take off' (=
> > > > > > Any ideas on overcomming this?
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> > > >
> > > >
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Maybe an off the wall idea, but have you considered an onboard GPS system to
get accurate position as a function of time, and then differentiate once for
velocity and twice for acceleration?
Dennis
one is that I can actualy sample acceleration.
for latteral acceleration, sample every 1/100th of a second
and display the highest sample on the screen. this would be usefull for
measuring how many G's you can 'pull' cornering.
besides..., my GPS unit is headed for the lawnmower =p
-Ryan
> Ryan -
>
> Maybe an off the wall idea, but have you considered an onboard GPS system
to
> get accurate position as a function of time, and then differentiate once
for
> velocity and twice for acceleration?
>
> Dennis
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
PIC's with built in ADC,
I have picbasic pro compiler, it even has a ADCIN command - altho I haven't
used it yet. good stuff - thx
-Ryan
Original Message
> if you have a pic or stamp or clone with a builtin a/d then use it (I
have
> had fine results using a 10-bit a/d.
>
> if not (like a stampII) then use a pwm one so that you can read in
directly
> to pin input
>
> shouldnt be too hard to get stable reading of +/- 1 or 2 bits an A/D
>
> In past I used something like a adc1034 for a/d worked fine with only
> 10-bits
>
> 5 volts for 2g at 10 bits should get you 0.005g resolution 12-bit will
> get you 0.001g resolution
>
> refer to pwm specs for thoses chip resolutions
>
>
>
> richard
>
>
>
>
Original Message
> From: "ryan" <rpogge@3...>
> To: <basicstamps@yahoogroups.com>
> Sent: Friday, May 11, 2001 12:44 PM
> Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
>
>
> > Yes i see now, I wasn't thinking.
> > What ADC would you recomend,
> > I usualy have problems with ADC's I can never get a stable reading, I
was
> > told it was /bc I was breadboarding it, and it needed a properly
designed
> > PCB with good ground planes etc....
> >
> > I have a couple of ADXL chips, the ADXL 150, 250, and 202. You refer to
a
> > +-2g chip the only one I see that is +-2 is the 202 which is duty cycle
> > output. - I belive I could just use
> > pulsin to read it. What is the advantage of useing one of the others
over
> a
> > PWM output chip? Analog sells a nice little eval board for the 202 ..
> might
> > check it out.
> >
> > Thanks
> > -Ryan
> >
> >
> >
Original Message
> >
> >
> > > 1% error due to angle is 1% error - does not have anything to do with
> > > range - of course total error is based on range
> > >
> > > the 1% I had reference to was 1% of the acceleration value - not time
> > >
> > > I dont know how the 1% could ever be related to time - presumably
your
> > time
> > > measurement is based on the accuracy of your clock or crystal - If it
is
> a
> > > microprocessor then a thousandth of a second is easly achieved and
maybe
> > > even could get millionth of a second.
> > >
> > > if you want to know acceleration then accuracy is 1%
> > > (a=a)
> > > if you integrate once to get velocity the accuracy loss due to
> > > accelerometer data is still 1% (v=a*t)
> > > if you integrate twice to get distance the accuracy loss due to
> > > accelerometer data is still 1%
> > > (x=0.5*a*t^2)
> > >
> > > for example, If you skid your car to stop and you measured 0.85 g
then
> > you
> > > could calculate the velocity change with respect to time
(v=0.85*32.2*t)
> > if
> > > it took 4 seconds for the skid then the starting velocity would have
> been
> > > 109 ft/sec (74 mph).
> > >
> > > The distance traveled would have been (x=0.5*0.85*32.2*4^2) 219 feet.
> > >
> > > The computer would measure the time to something like 3.9999987 sec.
> > >
> > > The 1% error on 74 mph would amount to 73.3 mph
> > > The 1% error on 219 feet would amount to 217 feet
> > >
> > > Your program would do a similar calculation 100 times a second using
a
> > > delta-t rather than total time
> > >
> > > performance computer manufacturers like gforce and vc-200, 10 years
ago,
> > > just used single accelerometer to measure acceleration and then made a
> > > correction to the data to account for "dive" on braking or "lift" on
> > > acceleration - this brute force correction attempt was somewhat
> mitigated
> > by
> > > their software asking if the vehicle had stiff suspension (less dive)
> or
> > > luxury car suspension (more dive). Then their program added or
> subtracted
> > 1%
> > > or 1/2% to the measured data.
> > >
> > > The right way to do it (and maybe they do it that way now) is to use
> two
> > > accelerometers one measures the acceleration and the others basicallly
> > > monitors the tilt of the vehicle. The program could then correctly
> compute
> > > the actual acceleration of the vehicle regardless of the lift or dive.
> In
> > > all cases it would be good to to get the best accelerometer you can
> afford
> > > with the lowest coss-axis sensitivity possible. The $20 Analog Devices
> > brand
> > > work great - I have used them.
> > >
> > > For the accuracy you need I would ignore all of the other effects
> > (assuming
> > > 12 bit a/d making 1000 samples per second, using 0-2 g accelerometer)
> > >
> > > richard
> > >
> > >
> > >
> > >
> > >
> > >
Original Message
> > > From: "ryan" <rpogge@3...>
> > > To: <basicstamps@yahoogroups.com>
> > > Sent: Friday, May 11, 2001 12:23 PM
> > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > >
> > >
> > > > Ah but you are assuming that the car can produce 1 G or more of
> > > > acceleration. What if the car only produces only 0.5G's(i have no
> idea)
> > > then
> > > > you have more like 2% error.
> > > > beacause you will be adding (1deg/90deg * 1G(gravity)) + 0.5G(from
> car).
> > > >
> > > > And that is assuming that you only have an inch of lift. Then you
> have
> > > to
> > > > take into consideration the error from accelerometer measurments,
time
> > > > measurments, the math(averaging) , the error that I will get when
the
> > > wheels
> > > > slip and create acceleration 'spikes', vibration, sun spots, women
> > drivers
> > > >
> > > >
> > > > and a 2% error on a 5 second run is 0.1 seconds. and I want acuracy
to
> > the
> > > > 0.1 at least, 0.01 would be ideal.
> > > > -i.e. being able to distinguish a 4.1 second time, from a 4.2 second
> > time
> > > is
> > > > minimal requirment, being able to distinguish between 4.11 and 4.19
> > > second
> > > > runs is ideal.
> > > >
> > > > I guess it is imposible to get that kind of accuracy from something
> > > onboard.
> > > > need to set up lasers on the racetrack if i want accuracy. *shrug*
> > > >
> > > > thanks for all the input
> > > > Ciao
> > > > -Ryan
> > > >
> > > > > one inch lift on a 8 foot wheelbase is about 0.9 degree change
> > > > > the sine of 0.9 degrees is 0.01
> > > > > It looks like less than 1% error
> > > > >
> > > > >
> > > > > richard
> > > > >
> > > > >
> > > > >
> > > > >
Original Message
> > > > > From: "ryan" <rpogge@3...>
> > > > > To: <basicstamps@yahoogroups.com>
> > > > > Sent: Friday, May 11, 2001 11:39 AM
> > > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > > >
> > > > >
> > > > > > Chilton-
> > > > > > Its a Dodge Viper, there is VERY little lift.
> > > > > > but I wonder if even an inch or two would effect the reading
> enough
> > to
> > > > > throw
> > > > > > the results. yes the rear wheels slip durring acceleration, now
> that
> > I
> > > > > think
> > > > > > about it, this
> > > > > > also causes a 'jerk' of forward acceleration when they catch
> again.
> > > > > > ....another problem to overcome.
> > > > > >
> > > > > > I mean maybee its not even an issue, ...unless I decide to use
it
> > with
> > > > the
> > > > > > 66 vett.
> > > > > > front end of that lifts so much I loose sight of the road (=
> > > > > >
> > > > > > I wanted an elegant solution, however I may resort to just
taking
> > > > readings
> > > > > > from
> > > > > > the speedometer sensor instead of an accelerometer, this would
> > > simplify
> > > > > > things
> > > > > > a great deal.
> > > > > >
> > > > > > Thanks for all the input guys and gals.
> > > > > >
> > > > > > -Ryan
> > > > > >
> > > > > > > Hey there SpeedRacer Ryan,
> > > > > > >
> > > > > > > Volkswagon employed special suspension devices in their
Corrado
> > line
> > > > > > > (and hopefully we'll see 'em again in the Steppenwolf) to
> decrease
> > > the
> > > > > > > effects of this. It didn't hurt that it also featured a
variable
> > > > spoiler
> > > > > > > on the back, but I don't think it affected the lift during
> initial
> > > > > > > acceleration (I think I remember it shifting twice, once at 45
> and
> > > > again
> > > > > > > at 140).
> > > > > > >
> > > > > > > My first question would be, where is the engine seated? If
it's
> a
> > > > front
> > > > > > > mounted engine and you're having lift problems, it might be a
> > > totally
> > > > > > > different scenario than if your engine is sitting above the
rear
> > > > wheels.
> > > > > > > Also, are your rear wheels skidding at all during
acceleration?
> > And
> > > > > > > lastly, is this a front wheel or rear wheel drive vehicle? If
> it's
> > a
> > > > > > > front wheel drive vehicle and you're seeing lift, eh, you may
> have
> > a
> > > > > > > whole other series of problems :-)
> > > > > > >
> > > > > > > (A hotrod junkie friend of mine tells me that the easiest way
to
> > > > > > > overcome this is to use two different engines, one for the
front
> > > > wheels
> > > > > > > and one for the back. Then he started talking about gear
ratios
> > and
> > > > > > > efficient torque balancing or something and I fell asleep.)
> > > > > > >
> > > > > > > If you *really* need a professional answer to this, I live 4
> miles
> > > > away
> > > > > > > from a race track, and I'm sure I can find someone out there
who
> > > would
> > > > > > > know. Next month we have the open Ferrari races, so I'm always
> > > looking
> > > > > > > for a reason to tell someone, "It'd be easier if I just show
you
> > > what
> > > > I
> > > > > > > mean. Can I have your keys?"
> > > > > > >
> > > > > > > -Chilton
> > > > > > >
> > > > > > > > My concern is that the front end of the car rises
> > > > > > > > slightly during 'take off' (=
> > > > > > > > Any ideas on overcomming this?
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Your use of Yahoo! Groups is subject to
> > > > > http://docs.yahoo.com/info/terms/
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > Your use of Yahoo! Groups is subject to
> > > > http://docs.yahoo.com/info/terms/
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> > > >
> > > >
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
had fine results using a 10-bit a/d.
if not (like a stampII) then use a pwm one so that you can read in directly
to pin input
shouldnt be too hard to get stable reading of +/- 1 or 2 bits an A/D
In past I used something like a adc1034 for a/d worked fine with only
10-bits
5 volts for 2g at 10 bits should get you 0.005g resolution 12-bit will
get you 0.001g resolution
refer to pwm specs for thoses chip resolutions
richard
Original Message
From: "ryan" <rpogge@3...>
To: <basicstamps@yahoogroups.com>
Sent: Friday, May 11, 2001 12:44 PM
Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> Yes i see now, I wasn't thinking.
> What ADC would you recomend,
> I usualy have problems with ADC's I can never get a stable reading, I was
> told it was /bc I was breadboarding it, and it needed a properly designed
> PCB with good ground planes etc....
>
> I have a couple of ADXL chips, the ADXL 150, 250, and 202. You refer to a
> +-2g chip the only one I see that is +-2 is the 202 which is duty cycle
> output. - I belive I could just use
> pulsin to read it. What is the advantage of useing one of the others over
a
> PWM output chip? Analog sells a nice little eval board for the 202 ..
might
> check it out.
>
> Thanks
> -Ryan
>
>
>
Original Message
>
>
> > 1% error due to angle is 1% error - does not have anything to do with
> > range - of course total error is based on range
> >
> > the 1% I had reference to was 1% of the acceleration value - not time
> >
> > I dont know how the 1% could ever be related to time - presumably your
> time
> > measurement is based on the accuracy of your clock or crystal - If it is
a
> > microprocessor then a thousandth of a second is easly achieved and maybe
> > even could get millionth of a second.
> >
> > if you want to know acceleration then accuracy is 1%
> > (a=a)
> > if you integrate once to get velocity the accuracy loss due to
> > accelerometer data is still 1% (v=a*t)
> > if you integrate twice to get distance the accuracy loss due to
> > accelerometer data is still 1%
> > (x=0.5*a*t^2)
> >
> > for example, If you skid your car to stop and you measured 0.85 g then
> you
> > could calculate the velocity change with respect to time (v=0.85*32.2*t)
> if
> > it took 4 seconds for the skid then the starting velocity would have
been
> > 109 ft/sec (74 mph).
> >
> > The distance traveled would have been (x=0.5*0.85*32.2*4^2) 219 feet.
> >
> > The computer would measure the time to something like 3.9999987 sec.
> >
> > The 1% error on 74 mph would amount to 73.3 mph
> > The 1% error on 219 feet would amount to 217 feet
> >
> > Your program would do a similar calculation 100 times a second using a
> > delta-t rather than total time
> >
> > performance computer manufacturers like gforce and vc-200, 10 years ago,
> > just used single accelerometer to measure acceleration and then made a
> > correction to the data to account for "dive" on braking or "lift" on
> > acceleration - this brute force correction attempt was somewhat
mitigated
> by
> > their software asking if the vehicle had stiff suspension (less dive)
or
> > luxury car suspension (more dive). Then their program added or
subtracted
> 1%
> > or 1/2% to the measured data.
> >
> > The right way to do it (and maybe they do it that way now) is to use
two
> > accelerometers one measures the acceleration and the others basicallly
> > monitors the tilt of the vehicle. The program could then correctly
compute
> > the actual acceleration of the vehicle regardless of the lift or dive.
In
> > all cases it would be good to to get the best accelerometer you can
afford
> > with the lowest coss-axis sensitivity possible. The $20 Analog Devices
> brand
> > work great - I have used them.
> >
> > For the accuracy you need I would ignore all of the other effects
> (assuming
> > 12 bit a/d making 1000 samples per second, using 0-2 g accelerometer)
> >
> > richard
> >
> >
> >
> >
> >
> >
Original Message
> > From: "ryan" <rpogge@3...>
> > To: <basicstamps@yahoogroups.com>
> > Sent: Friday, May 11, 2001 12:23 PM
> > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> >
> >
> > > Ah but you are assuming that the car can produce 1 G or more of
> > > acceleration. What if the car only produces only 0.5G's(i have no
idea)
> > then
> > > you have more like 2% error.
> > > beacause you will be adding (1deg/90deg * 1G(gravity)) + 0.5G(from
car).
> > >
> > > And that is assuming that you only have an inch of lift. Then you
have
> > to
> > > take into consideration the error from accelerometer measurments, time
> > > measurments, the math(averaging) , the error that I will get when the
> > wheels
> > > slip and create acceleration 'spikes', vibration, sun spots, women
> drivers
> > >
> > >
> > > and a 2% error on a 5 second run is 0.1 seconds. and I want acuracy to
> the
> > > 0.1 at least, 0.01 would be ideal.
> > > -i.e. being able to distinguish a 4.1 second time, from a 4.2 second
> time
> > is
> > > minimal requirment, being able to distinguish between 4.11 and 4.19
> > second
> > > runs is ideal.
> > >
> > > I guess it is imposible to get that kind of accuracy from something
> > onboard.
> > > need to set up lasers on the racetrack if i want accuracy. *shrug*
> > >
> > > thanks for all the input
> > > Ciao
> > > -Ryan
> > >
> > > > one inch lift on a 8 foot wheelbase is about 0.9 degree change
> > > > the sine of 0.9 degrees is 0.01
> > > > It looks like less than 1% error
> > > >
> > > >
> > > > richard
> > > >
> > > >
> > > >
> > > >
Original Message
> > > > From: "ryan" <rpogge@3...>
> > > > To: <basicstamps@yahoogroups.com>
> > > > Sent: Friday, May 11, 2001 11:39 AM
> > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > >
> > > >
> > > > > Chilton-
> > > > > Its a Dodge Viper, there is VERY little lift.
> > > > > but I wonder if even an inch or two would effect the reading
enough
> to
> > > > throw
> > > > > the results. yes the rear wheels slip durring acceleration, now
that
> I
> > > > think
> > > > > about it, this
> > > > > also causes a 'jerk' of forward acceleration when they catch
again.
> > > > > ....another problem to overcome.
> > > > >
> > > > > I mean maybee its not even an issue, ...unless I decide to use it
> with
> > > the
> > > > > 66 vett.
> > > > > front end of that lifts so much I loose sight of the road (=
> > > > >
> > > > > I wanted an elegant solution, however I may resort to just taking
> > > readings
> > > > > from
> > > > > the speedometer sensor instead of an accelerometer, this would
> > simplify
> > > > > things
> > > > > a great deal.
> > > > >
> > > > > Thanks for all the input guys and gals.
> > > > >
> > > > > -Ryan
> > > > >
> > > > > > Hey there SpeedRacer Ryan,
> > > > > >
> > > > > > Volkswagon employed special suspension devices in their Corrado
> line
> > > > > > (and hopefully we'll see 'em again in the Steppenwolf) to
decrease
> > the
> > > > > > effects of this. It didn't hurt that it also featured a variable
> > > spoiler
> > > > > > on the back, but I don't think it affected the lift during
initial
> > > > > > acceleration (I think I remember it shifting twice, once at 45
and
> > > again
> > > > > > at 140).
> > > > > >
> > > > > > My first question would be, where is the engine seated? If it's
a
> > > front
> > > > > > mounted engine and you're having lift problems, it might be a
> > totally
> > > > > > different scenario than if your engine is sitting above the rear
> > > wheels.
> > > > > > Also, are your rear wheels skidding at all during acceleration?
> And
> > > > > > lastly, is this a front wheel or rear wheel drive vehicle? If
it's
> a
> > > > > > front wheel drive vehicle and you're seeing lift, eh, you may
have
> a
> > > > > > whole other series of problems :-)
> > > > > >
> > > > > > (A hotrod junkie friend of mine tells me that the easiest way to
> > > > > > overcome this is to use two different engines, one for the front
> > > wheels
> > > > > > and one for the back. Then he started talking about gear ratios
> and
> > > > > > efficient torque balancing or something and I fell asleep.)
> > > > > >
> > > > > > If you *really* need a professional answer to this, I live 4
miles
> > > away
> > > > > > from a race track, and I'm sure I can find someone out there who
> > would
> > > > > > know. Next month we have the open Ferrari races, so I'm always
> > looking
> > > > > > for a reason to tell someone, "It'd be easier if I just show you
> > what
> > > I
> > > > > > mean. Can I have your keys?"
> > > > > >
> > > > > > -Chilton
> > > > > >
> > > > > > > My concern is that the front end of the car rises
> > > > > > > slightly during 'take off' (=
> > > > > > > Any ideas on overcomming this?
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > Your use of Yahoo! Groups is subject to
> > > > http://docs.yahoo.com/info/terms/
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Original Message
>
>
> > if you have a pic or stamp or clone with a builtin a/d then use it (I
> have
> > had fine results using a 10-bit a/d.
> >
> > if not (like a stampII) then use a pwm one so that you can read in
> directly
> > to pin input
> >
> > shouldnt be too hard to get stable reading of +/- 1 or 2 bits an A/D
> >
> > In past I used something like a adc1034 for a/d worked fine with only
> > 10-bits
> >
> > 5 volts for 2g at 10 bits should get you 0.005g resolution 12-bit
will
> > get you 0.001g resolution
> >
> > refer to pwm specs for thoses chip resolutions
> >
> >
> >
> > richard
> >
> >
> >
> >
Original Message
> > From: "ryan" <rpogge@3...>
> > To: <basicstamps@yahoogroups.com>
> > Sent: Friday, May 11, 2001 12:44 PM
> > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> >
> >
> > > Yes i see now, I wasn't thinking.
> > > What ADC would you recomend,
> > > I usualy have problems with ADC's I can never get a stable reading, I
> was
> > > told it was /bc I was breadboarding it, and it needed a properly
> designed
> > > PCB with good ground planes etc....
> > >
> > > I have a couple of ADXL chips, the ADXL 150, 250, and 202. You refer
to
> a
> > > +-2g chip the only one I see that is +-2 is the 202 which is duty
cycle
> > > output. - I belive I could just use
> > > pulsin to read it. What is the advantage of useing one of the others
> over
> > a
> > > PWM output chip? Analog sells a nice little eval board for the 202 ..
> > might
> > > check it out.
> > >
> > > Thanks
> > > -Ryan
> > >
> > >
> > >
Original Message
> > >
> > >
> > > > 1% error due to angle is 1% error - does not have anything to do
with
> > > > range - of course total error is based on range
> > > >
> > > > the 1% I had reference to was 1% of the acceleration value - not
time
> > > >
> > > > I dont know how the 1% could ever be related to time - presumably
> your
> > > time
> > > > measurement is based on the accuracy of your clock or crystal - If
it
> is
> > a
> > > > microprocessor then a thousandth of a second is easly achieved and
> maybe
> > > > even could get millionth of a second.
> > > >
> > > > if you want to know acceleration then accuracy is 1%
> > > > (a=a)
> > > > if you integrate once to get velocity the accuracy loss due to
> > > > accelerometer data is still 1% (v=a*t)
> > > > if you integrate twice to get distance the accuracy loss due to
> > > > accelerometer data is still 1%
> > > > (x=0.5*a*t^2)
> > > >
> > > > for example, If you skid your car to stop and you measured 0.85 g
> then
> > > you
> > > > could calculate the velocity change with respect to time
> (v=0.85*32.2*t)
> > > if
> > > > it took 4 seconds for the skid then the starting velocity would have
> > been
> > > > 109 ft/sec (74 mph).
> > > >
> > > > The distance traveled would have been (x=0.5*0.85*32.2*4^2) 219
feet.
> > > >
> > > > The computer would measure the time to something like 3.9999987 sec.
> > > >
> > > > The 1% error on 74 mph would amount to 73.3 mph
> > > > The 1% error on 219 feet would amount to 217 feet
> > > >
> > > > Your program would do a similar calculation 100 times a second
using
> a
> > > > delta-t rather than total time
> > > >
> > > > performance computer manufacturers like gforce and vc-200, 10 years
> ago,
> > > > just used single accelerometer to measure acceleration and then made
a
> > > > correction to the data to account for "dive" on braking or "lift"
on
> > > > acceleration - this brute force correction attempt was somewhat
> > mitigated
> > > by
> > > > their software asking if the vehicle had stiff suspension (less
dive)
> > or
> > > > luxury car suspension (more dive). Then their program added or
> > subtracted
> > > 1%
> > > > or 1/2% to the measured data.
> > > >
> > > > The right way to do it (and maybe they do it that way now) is to
use
> > two
> > > > accelerometers one measures the acceleration and the others
basicallly
> > > > monitors the tilt of the vehicle. The program could then correctly
> > compute
> > > > the actual acceleration of the vehicle regardless of the lift or
dive.
> > In
> > > > all cases it would be good to to get the best accelerometer you can
> > afford
> > > > with the lowest coss-axis sensitivity possible. The $20 Analog
Devices
> > > brand
> > > > work great - I have used them.
> > > >
> > > > For the accuracy you need I would ignore all of the other effects
> > > (assuming
> > > > 12 bit a/d making 1000 samples per second, using 0-2 g
accelerometer)
> > > >
> > > > richard
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
Original Message
> > > > From: "ryan" <rpogge@3...>
> > > > To: <basicstamps@yahoogroups.com>
> > > > Sent: Friday, May 11, 2001 12:23 PM
> > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > >
> > > >
> > > > > Ah but you are assuming that the car can produce 1 G or more of
> > > > > acceleration. What if the car only produces only 0.5G's(i have no
> > idea)
> > > > then
> > > > > you have more like 2% error.
> > > > > beacause you will be adding (1deg/90deg * 1G(gravity)) + 0.5G(from
> > car).
> > > > >
> > > > > And that is assuming that you only have an inch of lift. Then you
> > have
> > > > to
> > > > > take into consideration the error from accelerometer measurments,
> time
> > > > > measurments, the math(averaging) , the error that I will get when
> the
> > > > wheels
> > > > > slip and create acceleration 'spikes', vibration, sun spots, women
> > > drivers
> > > > >
> > > > >
> > > > > and a 2% error on a 5 second run is 0.1 seconds. and I want
acuracy
> to
> > > the
> > > > > 0.1 at least, 0.01 would be ideal.
> > > > > -i.e. being able to distinguish a 4.1 second time, from a 4.2
second
> > > time
> > > > is
> > > > > minimal requirment, being able to distinguish between 4.11 and
4.19
> > > > second
> > > > > runs is ideal.
> > > > >
> > > > > I guess it is imposible to get that kind of accuracy from
something
> > > > onboard.
> > > > > need to set up lasers on the racetrack if i want accuracy. *shrug*
> > > > >
> > > > > thanks for all the input
> > > > > Ciao
> > > > > -Ryan
> > > > >
> > > > > > one inch lift on a 8 foot wheelbase is about 0.9 degree change
> > > > > > the sine of 0.9 degrees is 0.01
> > > > > > It looks like less than 1% error
> > > > > >
> > > > > >
> > > > > > richard
> > > > > >
> > > > > >
> > > > > >
> > > > > >
Original Message
> > > > > > From: "ryan" <rpogge@3...>
> > > > > > To: <basicstamps@yahoogroups.com>
> > > > > > Sent: Friday, May 11, 2001 11:39 AM
> > > > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > > > >
> > > > > >
> > > > > > > Chilton-
> > > > > > > Its a Dodge Viper, there is VERY little lift.
> > > > > > > but I wonder if even an inch or two would effect the reading
> > enough
> > > to
> > > > > > throw
> > > > > > > the results. yes the rear wheels slip durring acceleration,
now
> > that
> > > I
> > > > > > think
> > > > > > > about it, this
> > > > > > > also causes a 'jerk' of forward acceleration when they catch
> > again.
> > > > > > > ....another problem to overcome.
> > > > > > >
> > > > > > > I mean maybee its not even an issue, ...unless I decide to use
> it
> > > with
> > > > > the
> > > > > > > 66 vett.
> > > > > > > front end of that lifts so much I loose sight of the road (=
> > > > > > >
> > > > > > > I wanted an elegant solution, however I may resort to just
> taking
> > > > > readings
> > > > > > > from
> > > > > > > the speedometer sensor instead of an accelerometer, this would
> > > > simplify
> > > > > > > things
> > > > > > > a great deal.
> > > > > > >
> > > > > > > Thanks for all the input guys and gals.
> > > > > > >
> > > > > > > -Ryan
> > > > > > >
> > > > > > > > Hey there SpeedRacer Ryan,
> > > > > > > >
> > > > > > > > Volkswagon employed special suspension devices in their
> Corrado
> > > line
> > > > > > > > (and hopefully we'll see 'em again in the Steppenwolf) to
> > decrease
> > > > the
> > > > > > > > effects of this. It didn't hurt that it also featured a
> variable
> > > > > spoiler
> > > > > > > > on the back, but I don't think it affected the lift during
> > initial
> > > > > > > > acceleration (I think I remember it shifting twice, once at
45
> > and
> > > > > again
> > > > > > > > at 140).
> > > > > > > >
> > > > > > > > My first question would be, where is the engine seated? If
> it's
> > a
> > > > > front
> > > > > > > > mounted engine and you're having lift problems, it might be
a
> > > > totally
> > > > > > > > different scenario than if your engine is sitting above the
> rear
> > > > > wheels.
> > > > > > > > Also, are your rear wheels skidding at all during
> acceleration?
> > > And
> > > > > > > > lastly, is this a front wheel or rear wheel drive vehicle?
If
> > it's
> > > a
> > > > > > > > front wheel drive vehicle and you're seeing lift, eh, you
may
> > have
> > > a
> > > > > > > > whole other series of problems :-)
> > > > > > > >
> > > > > > > > (A hotrod junkie friend of mine tells me that the easiest
way
> to
> > > > > > > > overcome this is to use two different engines, one for the
> front
> > > > > wheels
> > > > > > > > and one for the back. Then he started talking about gear
> ratios
> > > and
> > > > > > > > efficient torque balancing or something and I fell asleep.)
> > > > > > > >
> > > > > > > > If you *really* need a professional answer to this, I live 4
> > miles
> > > > > away
> > > > > > > > from a race track, and I'm sure I can find someone out there
> who
> > > > would
> > > > > > > > know. Next month we have the open Ferrari races, so I'm
always
> > > > looking
> > > > > > > > for a reason to tell someone, "It'd be easier if I just show
> you
> > > > what
> > > > > I
> > > > > > > > mean. Can I have your keys?"
> > > > > > > >
> > > > > > > > -Chilton
> > > > > > > >
> > > > > > > > > My concern is that the front end of the car rises
> > > > > > > > > slightly during 'take off' (=
> > > > > > > > > Any ideas on overcomming this?
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Your use of Yahoo! Groups is subject to
> > > > > http://docs.yahoo.com/info/terms/
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > Your use of Yahoo! Groups is subject to
> > > > http://docs.yahoo.com/info/terms/
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
> >
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
of a quarter... very small, I think it is 29 dollars from analog's webpage.
it is basicly a carrier board for the chip and 2 capacitors and a resistor.
I think it can also be plugged directly into a breadboard.
-Im ordering one this afternoon.
-Ryan
> I've been following this thread with a lot of interest. I've also have a
> couple of the '202 accelerometers from Analog Devices that I would like
> to use as a tilt/rotation sensor. They gave me a couple free samples, but
> they are surface mount chips. Are there any sockets made that can be used
> to mount these devices to a circuit board? If not is there a different
> version that has pins that can be easily soldered?
>
> Thanks
> Cory
>
>
>
> >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
>
> On 5/11/01, 12:45:17 PM, "ryan" <rpogge@3...> wrote regarding Re:
> [noparse][[/noparse]basicstamps] accelerometer:
>
>
> > would you recomend a PIC16F870 ?
>
>
> > >
Original Message
> > >
> > >
> > > > if you have a pic or stamp or clone with a builtin a/d then use it
(I
> > > have
> > > > had fine results using a 10-bit a/d.
> > > >
> > > > if not (like a stampII) then use a pwm one so that you can read in
> > > directly
> > > > to pin input
> > > >
> > > > shouldnt be too hard to get stable reading of +/- 1 or 2 bits an
A/D
> > > >
> > > > In past I used something like a adc1034 for a/d worked fine with
only
> > > > 10-bits
> > > >
> > > > 5 volts for 2g at 10 bits should get you 0.005g resolution 12-bit
> > will
> > > > get you 0.001g resolution
> > > >
> > > > refer to pwm specs for thoses chip resolutions
> > > >
> > > >
> > > >
> > > > richard
> > > >
> > > >
> > > >
> > > >
Original Message
> > > > From: "ryan" <rpogge@3...>
> > > > To: <basicstamps@yahoogroups.com>
> > > > Sent: Friday, May 11, 2001 12:44 PM
> > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > >
> > > >
> > > > > Yes i see now, I wasn't thinking.
> > > > > What ADC would you recomend,
> > > > > I usualy have problems with ADC's I can never get a stable
reading, I
> > > was
> > > > > told it was /bc I was breadboarding it, and it needed a properly
> > > designed
> > > > > PCB with good ground planes etc....
> > > > >
> > > > > I have a couple of ADXL chips, the ADXL 150, 250, and 202. You
refer
> > to
> > > a
> > > > > +-2g chip the only one I see that is +-2 is the 202 which is duty
> > cycle
> > > > > output. - I belive I could just use
> > > > > pulsin to read it. What is the advantage of useing one of the
others
> > > over
> > > > a
> > > > > PWM output chip? Analog sells a nice little eval board for the
202 ..
> > > > might
> > > > > check it out.
> > > > >
> > > > > Thanks
> > > > > -Ryan
> > > > >
> > > > >
> > > > >
Original Message
> > > > >
> > > > >
> > > > > > 1% error due to angle is 1% error - does not have anything to
do
> > with
> > > > > > range - of course total error is based on range
> > > > > >
> > > > > > the 1% I had reference to was 1% of the acceleration value - not
> > time
> > > > > >
> > > > > > I dont know how the 1% could ever be related to time -
presumably
> > > your
> > > > > time
> > > > > > measurement is based on the accuracy of your clock or crystal -
If
> > it
> > > is
> > > > a
> > > > > > microprocessor then a thousandth of a second is easly achieved
and
> > > maybe
> > > > > > even could get millionth of a second.
> > > > > >
> > > > > > if you want to know acceleration then accuracy is 1%
> > > > > > (a=a)
> > > > > > if you integrate once to get velocity the accuracy loss due to
> > > > > > accelerometer data is still 1% (v=a*t)
> > > > > > if you integrate twice to get distance the accuracy loss due to
> > > > > > accelerometer data is still 1%
> > > > > > (x=0.5*a*t^2)
> > > > > >
> > > > > > for example, If you skid your car to stop and you measured 0.85
g
> > > then
> > > > > you
> > > > > > could calculate the velocity change with respect to time
> > > (v=0.85*32.2*t)
> > > > > if
> > > > > > it took 4 seconds for the skid then the starting velocity would
have
> > > > been
> > > > > > 109 ft/sec (74 mph).
> > > > > >
> > > > > > The distance traveled would have been (x=0.5*0.85*32.2*4^2) 219
> > feet.
> > > > > >
> > > > > > The computer would measure the time to something like 3.9999987
sec.
> > > > > >
> > > > > > The 1% error on 74 mph would amount to 73.3 mph
> > > > > > The 1% error on 219 feet would amount to 217 feet
> > > > > >
> > > > > > Your program would do a similar calculation 100 times a second
> > using
> > > a
> > > > > > delta-t rather than total time
> > > > > >
> > > > > > performance computer manufacturers like gforce and vc-200, 10
years
> > > ago,
> > > > > > just used single accelerometer to measure acceleration and then
made
> > a
> > > > > > correction to the data to account for "dive" on braking or
"lift"
> > on
> > > > > > acceleration - this brute force correction attempt was somewhat
> > > > mitigated
> > > > > by
> > > > > > their software asking if the vehicle had stiff suspension (less
> > dive)
> > > > or
> > > > > > luxury car suspension (more dive). Then their program added or
> > > > subtracted
> > > > > 1%
> > > > > > or 1/2% to the measured data.
> > > > > >
> > > > > > The right way to do it (and maybe they do it that way now) is
to
> > use
> > > > two
> > > > > > accelerometers one measures the acceleration and the others
> > basicallly
> > > > > > monitors the tilt of the vehicle. The program could then
correctly
> > > > compute
> > > > > > the actual acceleration of the vehicle regardless of the lift or
> > dive.
> > > > In
> > > > > > all cases it would be good to to get the best accelerometer you
can
> > > > afford
> > > > > > with the lowest coss-axis sensitivity possible. The $20 Analog
> > Devices
> > > > > brand
> > > > > > work great - I have used them.
> > > > > >
> > > > > > For the accuracy you need I would ignore all of the other
effects
> > > > > (assuming
> > > > > > 12 bit a/d making 1000 samples per second, using 0-2 g
> > accelerometer)
> > > > > >
> > > > > > richard
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
Original Message
> > > > > > From: "ryan" <rpogge@3...>
> > > > > > To: <basicstamps@yahoogroups.com>
> > > > > > Sent: Friday, May 11, 2001 12:23 PM
> > > > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > > > >
> > > > > >
> > > > > > > Ah but you are assuming that the car can produce 1 G or more
of
> > > > > > > acceleration. What if the car only produces only 0.5G's(i have
no
> > > > idea)
> > > > > > then
> > > > > > > you have more like 2% error.
> > > > > > > beacause you will be adding (1deg/90deg * 1G(gravity)) +
0.5G(from
> > > > car).
> > > > > > >
> > > > > > > And that is assuming that you only have an inch of lift. Then
you
> > > > have
> > > > > > to
> > > > > > > take into consideration the error from accelerometer
measurments,
> > > time
> > > > > > > measurments, the math(averaging) , the error that I will get
when
> > > the
> > > > > > wheels
> > > > > > > slip and create acceleration 'spikes', vibration, sun spots,
women
> > > > > drivers
> > > > > > >
> > > > > > >
> > > > > > > and a 2% error on a 5 second run is 0.1 seconds. and I want
> > acuracy
> > > to
> > > > > the
> > > > > > > 0.1 at least, 0.01 would be ideal.
> > > > > > > -i.e. being able to distinguish a 4.1 second time, from a 4.2
> > second
> > > > > time
> > > > > > is
> > > > > > > minimal requirment, being able to distinguish between 4.11 and
> > 4.19
> > > > > > second
> > > > > > > runs is ideal.
> > > > > > >
> > > > > > > I guess it is imposible to get that kind of accuracy from
> > something
> > > > > > onboard.
> > > > > > > need to set up lasers on the racetrack if i want accuracy.
*shrug*
> > > > > > >
> > > > > > > thanks for all the input
> > > > > > > Ciao
> > > > > > > -Ryan
> > > > > > >
> > > > > > > > one inch lift on a 8 foot wheelbase is about 0.9 degree
change
> > > > > > > > the sine of 0.9 degrees is 0.01
> > > > > > > > It looks like less than 1% error
> > > > > > > >
> > > > > > > >
> > > > > > > > richard
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
Original Message
> > > > > > > > From: "ryan" <rpogge@3...>
> > > > > > > > To: <basicstamps@yahoogroups.com>
> > > > > > > > Sent: Friday, May 11, 2001 11:39 AM
> > > > > > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > > > > > >
> > > > > > > >
> > > > > > > > > Chilton-
> > > > > > > > > Its a Dodge Viper, there is VERY little lift.
> > > > > > > > > but I wonder if even an inch or two would effect the
reading
> > > > enough
> > > > > to
> > > > > > > > throw
> > > > > > > > > the results. yes the rear wheels slip durring
acceleration,
> > now
> > > > that
> > > > > I
> > > > > > > > think
> > > > > > > > > about it, this
> > > > > > > > > also causes a 'jerk' of forward acceleration when they
catch
> > > > again.
> > > > > > > > > ....another problem to overcome.
> > > > > > > > >
> > > > > > > > > I mean maybee its not even an issue, ...unless I decide to
use
> > > it
> > > > > with
> > > > > > > the
> > > > > > > > > 66 vett.
> > > > > > > > > front end of that lifts so much I loose sight of the road
(=
> > > > > > > > >
> > > > > > > > > I wanted an elegant solution, however I may resort to just
> > > taking
> > > > > > > readings
> > > > > > > > > from
> > > > > > > > > the speedometer sensor instead of an accelerometer, this
would
> > > > > > simplify
> > > > > > > > > things
> > > > > > > > > a great deal.
> > > > > > > > >
> > > > > > > > > Thanks for all the input guys and gals.
> > > > > > > > >
> > > > > > > > > -Ryan
> > > > > > > > >
> > > > > > > > > > Hey there SpeedRacer Ryan,
> > > > > > > > > >
> > > > > > > > > > Volkswagon employed special suspension devices in their
> > > Corrado
> > > > > line
> > > > > > > > > > (and hopefully we'll see 'em again in the Steppenwolf)
to
> > > > decrease
> > > > > > the
> > > > > > > > > > effects of this. It didn't hurt that it also featured a
> > > variable
> > > > > > > spoiler
> > > > > > > > > > on the back, but I don't think it affected the lift
during
> > > > initial
> > > > > > > > > > acceleration (I think I remember it shifting twice, once
at
> > 45
> > > > and
> > > > > > > again
> > > > > > > > > > at 140).
> > > > > > > > > >
> > > > > > > > > > My first question would be, where is the engine seated?
If
> > > it's
> > > > a
> > > > > > > front
> > > > > > > > > > mounted engine and you're having lift problems, it might
be
> > a
> > > > > > totally
> > > > > > > > > > different scenario than if your engine is sitting above
the
> > > rear
> > > > > > > wheels.
> > > > > > > > > > Also, are your rear wheels skidding at all during
> > > acceleration?
> > > > > And
> > > > > > > > > > lastly, is this a front wheel or rear wheel drive
vehicle?
> > If
> > > > it's
> > > > > a
> > > > > > > > > > front wheel drive vehicle and you're seeing lift, eh,
you
> > may
> > > > have
> > > > > a
> > > > > > > > > > whole other series of problems :-)
> > > > > > > > > >
> > > > > > > > > > (A hotrod junkie friend of mine tells me that the
easiest
> > way
> > > to
> > > > > > > > > > overcome this is to use two different engines, one for
the
> > > front
> > > > > > > wheels
> > > > > > > > > > and one for the back. Then he started talking about gear
> > > ratios
> > > > > and
> > > > > > > > > > efficient torque balancing or something and I fell
asleep.)
> > > > > > > > > >
> > > > > > > > > > If you *really* need a professional answer to this, I
live 4
> > > > miles
> > > > > > > away
> > > > > > > > > > from a race track, and I'm sure I can find someone out
there
> > > who
> > > > > > would
> > > > > > > > > > know. Next month we have the open Ferrari races, so I'm
> > always
> > > > > > looking
> > > > > > > > > > for a reason to tell someone, "It'd be easier if I just
show
> > > you
> > > > > > what
> > > > > > > I
> > > > > > > > > > mean. Can I have your keys?"
> > > > > > > > > >
> > > > > > > > > > -Chilton
> > > > > > > > > >
> > > > > > > > > > > My concern is that the front end of the car rises
> > > > > > > > > > > slightly during 'take off' (=
> > > > > > > > > > > Any ideas on overcomming this?
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Your use of Yahoo! Groups is subject to
> > > > > http://docs.yahoo.com/info/terms/
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > Your use of Yahoo! Groups is subject to
> > > > http://docs.yahoo.com/info/terms/
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
>
>
>
>
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
couple of the '202 accelerometers from Analog Devices that I would like
to use as a tilt/rotation sensor. They gave me a couple free samples, but
they are surface mount chips. Are there any sockets made that can be used
to mount these devices to a circuit board? If not is there a different
version that has pins that can be easily soldered?
Thanks
Cory
>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
On 5/11/01, 12:45:17 PM, "ryan" <rpogge@3...> wrote regarding Re:
[noparse][[/noparse]basicstamps] accelerometer:
> would you recomend a PIC16F870 ?
> >
Original Message
> >
> >
> > > if you have a pic or stamp or clone with a builtin a/d then use it (I
> > have
> > > had fine results using a 10-bit a/d.
> > >
> > > if not (like a stampII) then use a pwm one so that you can read in
> > directly
> > > to pin input
> > >
> > > shouldnt be too hard to get stable reading of +/- 1 or 2 bits an A/D
> > >
> > > In past I used something like a adc1034 for a/d worked fine with only
> > > 10-bits
> > >
> > > 5 volts for 2g at 10 bits should get you 0.005g resolution 12-bit
> will
> > > get you 0.001g resolution
> > >
> > > refer to pwm specs for thoses chip resolutions
> > >
> > >
> > >
> > > richard
> > >
> > >
> > >
> > >
Original Message
> > > From: "ryan" <rpogge@3...>
> > > To: <basicstamps@yahoogroups.com>
> > > Sent: Friday, May 11, 2001 12:44 PM
> > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > >
> > >
> > > > Yes i see now, I wasn't thinking.
> > > > What ADC would you recomend,
> > > > I usualy have problems with ADC's I can never get a stable reading, I
> > was
> > > > told it was /bc I was breadboarding it, and it needed a properly
> > designed
> > > > PCB with good ground planes etc....
> > > >
> > > > I have a couple of ADXL chips, the ADXL 150, 250, and 202. You refer
> to
> > a
> > > > +-2g chip the only one I see that is +-2 is the 202 which is duty
> cycle
> > > > output. - I belive I could just use
> > > > pulsin to read it. What is the advantage of useing one of the others
> > over
> > > a
> > > > PWM output chip? Analog sells a nice little eval board for the 202 ..
> > > might
> > > > check it out.
> > > >
> > > > Thanks
> > > > -Ryan
> > > >
> > > >
> > > >
Original Message
> > > >
> > > >
> > > > > 1% error due to angle is 1% error - does not have anything to do
> with
> > > > > range - of course total error is based on range
> > > > >
> > > > > the 1% I had reference to was 1% of the acceleration value - not
> time
> > > > >
> > > > > I dont know how the 1% could ever be related to time - presumably
> > your
> > > > time
> > > > > measurement is based on the accuracy of your clock or crystal - If
> it
> > is
> > > a
> > > > > microprocessor then a thousandth of a second is easly achieved and
> > maybe
> > > > > even could get millionth of a second.
> > > > >
> > > > > if you want to know acceleration then accuracy is 1%
> > > > > (a=a)
> > > > > if you integrate once to get velocity the accuracy loss due to
> > > > > accelerometer data is still 1% (v=a*t)
> > > > > if you integrate twice to get distance the accuracy loss due to
> > > > > accelerometer data is still 1%
> > > > > (x=0.5*a*t^2)
> > > > >
> > > > > for example, If you skid your car to stop and you measured 0.85 g
> > then
> > > > you
> > > > > could calculate the velocity change with respect to time
> > (v=0.85*32.2*t)
> > > > if
> > > > > it took 4 seconds for the skid then the starting velocity would have
> > > been
> > > > > 109 ft/sec (74 mph).
> > > > >
> > > > > The distance traveled would have been (x=0.5*0.85*32.2*4^2) 219
> feet.
> > > > >
> > > > > The computer would measure the time to something like 3.9999987 sec.
> > > > >
> > > > > The 1% error on 74 mph would amount to 73.3 mph
> > > > > The 1% error on 219 feet would amount to 217 feet
> > > > >
> > > > > Your program would do a similar calculation 100 times a second
> using
> > a
> > > > > delta-t rather than total time
> > > > >
> > > > > performance computer manufacturers like gforce and vc-200, 10 years
> > ago,
> > > > > just used single accelerometer to measure acceleration and then made
> a
> > > > > correction to the data to account for "dive" on braking or "lift"
> on
> > > > > acceleration - this brute force correction attempt was somewhat
> > > mitigated
> > > > by
> > > > > their software asking if the vehicle had stiff suspension (less
> dive)
> > > or
> > > > > luxury car suspension (more dive). Then their program added or
> > > subtracted
> > > > 1%
> > > > > or 1/2% to the measured data.
> > > > >
> > > > > The right way to do it (and maybe they do it that way now) is to
> use
> > > two
> > > > > accelerometers one measures the acceleration and the others
> basicallly
> > > > > monitors the tilt of the vehicle. The program could then correctly
> > > compute
> > > > > the actual acceleration of the vehicle regardless of the lift or
> dive.
> > > In
> > > > > all cases it would be good to to get the best accelerometer you can
> > > afford
> > > > > with the lowest coss-axis sensitivity possible. The $20 Analog
> Devices
> > > > brand
> > > > > work great - I have used them.
> > > > >
> > > > > For the accuracy you need I would ignore all of the other effects
> > > > (assuming
> > > > > 12 bit a/d making 1000 samples per second, using 0-2 g
> accelerometer)
> > > > >
> > > > > richard
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
Original Message
> > > > > From: "ryan" <rpogge@3...>
> > > > > To: <basicstamps@yahoogroups.com>
> > > > > Sent: Friday, May 11, 2001 12:23 PM
> > > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > > >
> > > > >
> > > > > > Ah but you are assuming that the car can produce 1 G or more of
> > > > > > acceleration. What if the car only produces only 0.5G's(i have no
> > > idea)
> > > > > then
> > > > > > you have more like 2% error.
> > > > > > beacause you will be adding (1deg/90deg * 1G(gravity)) + 0.5G(from
> > > car).
> > > > > >
> > > > > > And that is assuming that you only have an inch of lift. Then you
> > > have
> > > > > to
> > > > > > take into consideration the error from accelerometer measurments,
> > time
> > > > > > measurments, the math(averaging) , the error that I will get when
> > the
> > > > > wheels
> > > > > > slip and create acceleration 'spikes', vibration, sun spots, women
> > > > drivers
> > > > > >
> > > > > >
> > > > > > and a 2% error on a 5 second run is 0.1 seconds. and I want
> acuracy
> > to
> > > > the
> > > > > > 0.1 at least, 0.01 would be ideal.
> > > > > > -i.e. being able to distinguish a 4.1 second time, from a 4.2
> second
> > > > time
> > > > > is
> > > > > > minimal requirment, being able to distinguish between 4.11 and
> 4.19
> > > > > second
> > > > > > runs is ideal.
> > > > > >
> > > > > > I guess it is imposible to get that kind of accuracy from
> something
> > > > > onboard.
> > > > > > need to set up lasers on the racetrack if i want accuracy. *shrug*
> > > > > >
> > > > > > thanks for all the input
> > > > > > Ciao
> > > > > > -Ryan
> > > > > >
> > > > > > > one inch lift on a 8 foot wheelbase is about 0.9 degree change
> > > > > > > the sine of 0.9 degrees is 0.01
> > > > > > > It looks like less than 1% error
> > > > > > >
> > > > > > >
> > > > > > > richard
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
Original Message
> > > > > > > From: "ryan" <rpogge@3...>
> > > > > > > To: <basicstamps@yahoogroups.com>
> > > > > > > Sent: Friday, May 11, 2001 11:39 AM
> > > > > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > > > > >
> > > > > > >
> > > > > > > > Chilton-
> > > > > > > > Its a Dodge Viper, there is VERY little lift.
> > > > > > > > but I wonder if even an inch or two would effect the reading
> > > enough
> > > > to
> > > > > > > throw
> > > > > > > > the results. yes the rear wheels slip durring acceleration,
> now
> > > that
> > > > I
> > > > > > > think
> > > > > > > > about it, this
> > > > > > > > also causes a 'jerk' of forward acceleration when they catch
> > > again.
> > > > > > > > ....another problem to overcome.
> > > > > > > >
> > > > > > > > I mean maybee its not even an issue, ...unless I decide to use
> > it
> > > > with
> > > > > > the
> > > > > > > > 66 vett.
> > > > > > > > front end of that lifts so much I loose sight of the road (=
> > > > > > > >
> > > > > > > > I wanted an elegant solution, however I may resort to just
> > taking
> > > > > > readings
> > > > > > > > from
> > > > > > > > the speedometer sensor instead of an accelerometer, this would
> > > > > simplify
> > > > > > > > things
> > > > > > > > a great deal.
> > > > > > > >
> > > > > > > > Thanks for all the input guys and gals.
> > > > > > > >
> > > > > > > > -Ryan
> > > > > > > >
> > > > > > > > > Hey there SpeedRacer Ryan,
> > > > > > > > >
> > > > > > > > > Volkswagon employed special suspension devices in their
> > Corrado
> > > > line
> > > > > > > > > (and hopefully we'll see 'em again in the Steppenwolf) to
> > > decrease
> > > > > the
> > > > > > > > > effects of this. It didn't hurt that it also featured a
> > variable
> > > > > > spoiler
> > > > > > > > > on the back, but I don't think it affected the lift during
> > > initial
> > > > > > > > > acceleration (I think I remember it shifting twice, once at
> 45
> > > and
> > > > > > again
> > > > > > > > > at 140).
> > > > > > > > >
> > > > > > > > > My first question would be, where is the engine seated? If
> > it's
> > > a
> > > > > > front
> > > > > > > > > mounted engine and you're having lift problems, it might be
> a
> > > > > totally
> > > > > > > > > different scenario than if your engine is sitting above the
> > rear
> > > > > > wheels.
> > > > > > > > > Also, are your rear wheels skidding at all during
> > acceleration?
> > > > And
> > > > > > > > > lastly, is this a front wheel or rear wheel drive vehicle?
> If
> > > it's
> > > > a
> > > > > > > > > front wheel drive vehicle and you're seeing lift, eh, you
> may
> > > have
> > > > a
> > > > > > > > > whole other series of problems :-)
> > > > > > > > >
> > > > > > > > > (A hotrod junkie friend of mine tells me that the easiest
> way
> > to
> > > > > > > > > overcome this is to use two different engines, one for the
> > front
> > > > > > wheels
> > > > > > > > > and one for the back. Then he started talking about gear
> > ratios
> > > > and
> > > > > > > > > efficient torque balancing or something and I fell asleep.)
> > > > > > > > >
> > > > > > > > > If you *really* need a professional answer to this, I live 4
> > > miles
> > > > > > away
> > > > > > > > > from a race track, and I'm sure I can find someone out there
> > who
> > > > > would
> > > > > > > > > know. Next month we have the open Ferrari races, so I'm
> always
> > > > > looking
> > > > > > > > > for a reason to tell someone, "It'd be easier if I just show
> > you
> > > > > what
> > > > > > I
> > > > > > > > > mean. Can I have your keys?"
> > > > > > > > >
> > > > > > > > > -Chilton
> > > > > > > > >
> > > > > > > > > > My concern is that the front end of the car rises
> > > > > > > > > > slightly during 'take off' (=
> > > > > > > > > > Any ideas on overcomming this?
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Your use of Yahoo! Groups is subject to
> > > > > http://docs.yahoo.com/info/terms/
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > Your use of Yahoo! Groups is subject to
> > > > http://docs.yahoo.com/info/terms/
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > > > >
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> > > >
> > > >
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> >
> >
> >
> >
> > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> >
> >
> >
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
I noticed earlier in this post you mentioned that space was at a
premium and that it was difficult to get a stable ADC reading when
breadboarding. I have a prototype design for the BS2 (or e,sx,p)
which has an 8-channel 8-bit ADC, RTC, 256k EEPROM and 8 additional
I/O ports. You "park" the STAMP into this carrier--it's a 0.9" wide
package with 40 pins on 0.1" spacing for breadboarding or integration
into a PCB. The ADC is only an 8-bit, but we're working on a new
revision which has an 8-channel, 12-bit AD. You can get details at:
http://www.High-TechGarage.com/products/supermax.php
What kind of Viper is it? RT-10 or GTS? Hope you're not paying too
much for gas (here in WI we're at $2.00 a gallon).
Good luck!
--- In basicstamps@y..., "ryan" <rpogge@3...> wrote:
> Yes i see now, I wasn't thinking.
> What ADC would you recomend,
> I usualy have problems with ADC's I can never get a stable reading,
> I was told it was /bc I was breadboarding it, and it needed a
> properly designed PCB with good ground planes etc....
Its a GTS.
premium gas is at 1.81 a gallon(in VA) as of this morning if I remember
correctly.
I will check out your board, I
will actually be using a PIC uC, 16C87x I *think*
which has a built in ADC, however your board might be useful for
prototyping (I need to try some of these PIC bootloaders that are floating
around one of these days though!)
Thanks!
-Ryan
> Hi Ryan,
>
> I noticed earlier in this post you mentioned that space was at a
> premium and that it was difficult to get a stable ADC reading when
> breadboarding. I have a prototype design for the BS2 (or e,sx,p)
> which has an 8-channel 8-bit ADC, RTC, 256k EEPROM and 8 additional
> I/O ports. You "park" the STAMP into this carrier--it's a 0.9" wide
> package with 40 pins on 0.1" spacing for breadboarding or integration
> into a PCB. The ADC is only an 8-bit, but we're working on a new
> revision which has an 8-channel, 12-bit AD. You can get details at:
> http://www.High-TechGarage.com/products/supermax.php
>
> What kind of Viper is it? RT-10 or GTS? Hope you're not paying too
> much for gas (here in WI we're at $2.00 a gallon).
>
> Good luck!
>
> --- In basicstamps@y..., "ryan" <rpogge@3...> wrote:
> > Yes i see now, I wasn't thinking.
> > What ADC would you recomend,
> > I usualy have problems with ADC's I can never get a stable reading,
> > I was told it was /bc I was breadboarding it, and it needed a
> > properly designed PCB with good ground planes etc....
>
>
>
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>
I agree with Ryan about the eval board. But if you want to mount several of
the 202J series, I found a surface-to-dip board (0.72"x0.45") as part of the
Surf Board series at my local supply shop (also at Digikey). Use of a fine
soldering tip and a magnifying class helped solder the 202 to the surf
board.
Denis
Original Message
From: ryan [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=_5TwHmAFI88YUKCRCrwHSkrbOOA6mnf8US1nRw6kXkSIa0nmQmStf4KIr2fKlJOOvIajfkk_]rpogge@3...[/url
Sent: Friday, May 11, 2001 12:35 PM
To: basicstamps@yahoogroups.com
Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
check out the 202EB .. its an eval board.. that 'looks' to be about the size
of a quarter... very small, I think it is 29 dollars from analog's webpage.
it is basicly a carrier board for the chip and 2 capacitors and a resistor.
I think it can also be plugged directly into a breadboard.
-Im ordering one this afternoon.
-Ryan
> I've been following this thread with a lot of interest. I've also have a
> couple of the '202 accelerometers from Analog Devices that I would like
> to use as a tilt/rotation sensor. They gave me a couple free samples, but
> they are surface mount chips. Are there any sockets made that can be used
> to mount these devices to a circuit board? If not is there a different
> version that has pins that can be easily soldered?
>
> Thanks
> Cory
>
>
>
> >>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
>
> On 5/11/01, 12:45:17 PM, "ryan" <rpogge@3...> wrote regarding Re:
> [noparse][[/noparse]basicstamps] accelerometer:
>
>
> > would you recomend a PIC16F870 ?
>
>
> > >
Original Message
> > >
> > >
> > > > if you have a pic or stamp or clone with a builtin a/d then use it
(I
> > > have
> > > > had fine results using a 10-bit a/d.
> > > >
> > > > if not (like a stampII) then use a pwm one so that you can read in
> > > directly
> > > > to pin input
> > > >
> > > > shouldnt be too hard to get stable reading of +/- 1 or 2 bits an
A/D
> > > >
> > > > In past I used something like a adc1034 for a/d worked fine with
only
> > > > 10-bits
> > > >
> > > > 5 volts for 2g at 10 bits should get you 0.005g resolution 12-bit
> > will
> > > > get you 0.001g resolution
> > > >
> > > > refer to pwm specs for thoses chip resolutions
> > > >
> > > >
> > > >
> > > > richard
> > > >
> > > >
> > > >
> > > >
Original Message
> > > > From: "ryan" <rpogge@3...>
> > > > To: <basicstamps@yahoogroups.com>
> > > > Sent: Friday, May 11, 2001 12:44 PM
> > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > >
> > > >
> > > > > Yes i see now, I wasn't thinking.
> > > > > What ADC would you recomend,
> > > > > I usualy have problems with ADC's I can never get a stable
reading, I
> > > was
> > > > > told it was /bc I was breadboarding it, and it needed a properly
> > > designed
> > > > > PCB with good ground planes etc....
> > > > >
> > > > > I have a couple of ADXL chips, the ADXL 150, 250, and 202. You
refer
> > to
> > > a
> > > > > +-2g chip the only one I see that is +-2 is the 202 which is duty
> > cycle
> > > > > output. - I belive I could just use
> > > > > pulsin to read it. What is the advantage of useing one of the
others
> > > over
> > > > a
> > > > > PWM output chip? Analog sells a nice little eval board for the
202 ..
> > > > might
> > > > > check it out.
> > > > >
> > > > > Thanks
> > > > > -Ryan
> > > > >
> > > > >
> > > > >
Original Message
> > > > >
> > > > >
> > > > > > 1% error due to angle is 1% error - does not have anything to
do
> > with
> > > > > > range - of course total error is based on range
> > > > > >
> > > > > > the 1% I had reference to was 1% of the acceleration value - not
> > time
> > > > > >
> > > > > > I dont know how the 1% could ever be related to time -
presumably
> > > your
> > > > > time
> > > > > > measurement is based on the accuracy of your clock or crystal -
If
> > it
> > > is
> > > > a
> > > > > > microprocessor then a thousandth of a second is easly achieved
and
> > > maybe
> > > > > > even could get millionth of a second.
> > > > > >
> > > > > > if you want to know acceleration then accuracy is 1%
> > > > > > (a=a)
> > > > > > if you integrate once to get velocity the accuracy loss due to
> > > > > > accelerometer data is still 1% (v=a*t)
> > > > > > if you integrate twice to get distance the accuracy loss due to
> > > > > > accelerometer data is still 1%
> > > > > > (x=0.5*a*t^2)
> > > > > >
> > > > > > for example, If you skid your car to stop and you measured 0.85
g
> > > then
> > > > > you
> > > > > > could calculate the velocity change with respect to time
> > > (v=0.85*32.2*t)
> > > > > if
> > > > > > it took 4 seconds for the skid then the starting velocity would
have
> > > > been
> > > > > > 109 ft/sec (74 mph).
> > > > > >
> > > > > > The distance traveled would have been (x=0.5*0.85*32.2*4^2) 219
> > feet.
> > > > > >
> > > > > > The computer would measure the time to something like 3.9999987
sec.
> > > > > >
> > > > > > The 1% error on 74 mph would amount to 73.3 mph
> > > > > > The 1% error on 219 feet would amount to 217 feet
> > > > > >
> > > > > > Your program would do a similar calculation 100 times a second
> > using
> > > a
> > > > > > delta-t rather than total time
> > > > > >
> > > > > > performance computer manufacturers like gforce and vc-200, 10
years
> > > ago,
> > > > > > just used single accelerometer to measure acceleration and then
made
> > a
> > > > > > correction to the data to account for "dive" on braking or
"lift"
> > on
> > > > > > acceleration - this brute force correction attempt was somewhat
> > > > mitigated
> > > > > by
> > > > > > their software asking if the vehicle had stiff suspension (less
> > dive)
> > > > or
> > > > > > luxury car suspension (more dive). Then their program added or
> > > > subtracted
> > > > > 1%
> > > > > > or 1/2% to the measured data.
> > > > > >
> > > > > > The right way to do it (and maybe they do it that way now) is
to
> > use
> > > > two
> > > > > > accelerometers one measures the acceleration and the others
> > basicallly
> > > > > > monitors the tilt of the vehicle. The program could then
correctly
> > > > compute
> > > > > > the actual acceleration of the vehicle regardless of the lift or
> > dive.
> > > > In
> > > > > > all cases it would be good to to get the best accelerometer you
can
> > > > afford
> > > > > > with the lowest coss-axis sensitivity possible. The $20 Analog
> > Devices
> > > > > brand
> > > > > > work great - I have used them.
> > > > > >
> > > > > > For the accuracy you need I would ignore all of the other
effects
> > > > > (assuming
> > > > > > 12 bit a/d making 1000 samples per second, using 0-2 g
> > accelerometer)
> > > > > >
> > > > > > richard
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
Original Message
> > > > > > From: "ryan" <rpogge@3...>
> > > > > > To: <basicstamps@yahoogroups.com>
> > > > > > Sent: Friday, May 11, 2001 12:23 PM
> > > > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > > > >
> > > > > >
> > > > > > > Ah but you are assuming that the car can produce 1 G or more
of
> > > > > > > acceleration. What if the car only produces only 0.5G's(i have
no
> > > > idea)
> > > > > > then
> > > > > > > you have more like 2% error.
> > > > > > > beacause you will be adding (1deg/90deg * 1G(gravity)) +
0.5G(from
> > > > car).
> > > > > > >
> > > > > > > And that is assuming that you only have an inch of lift. Then
you
> > > > have
> > > > > > to
> > > > > > > take into consideration the error from accelerometer
measurments,
> > > time
> > > > > > > measurments, the math(averaging) , the error that I will get
when
> > > the
> > > > > > wheels
> > > > > > > slip and create acceleration 'spikes', vibration, sun spots,
women
> > > > > drivers
> > > > > > >
> > > > > > >
> > > > > > > and a 2% error on a 5 second run is 0.1 seconds. and I want
> > acuracy
> > > to
> > > > > the
> > > > > > > 0.1 at least, 0.01 would be ideal.
> > > > > > > -i.e. being able to distinguish a 4.1 second time, from a 4.2
> > second
> > > > > time
> > > > > > is
> > > > > > > minimal requirment, being able to distinguish between 4.11 and
> > 4.19
> > > > > > second
> > > > > > > runs is ideal.
> > > > > > >
> > > > > > > I guess it is imposible to get that kind of accuracy from
> > something
> > > > > > onboard.
> > > > > > > need to set up lasers on the racetrack if i want accuracy.
*shrug*
> > > > > > >
> > > > > > > thanks for all the input
> > > > > > > Ciao
> > > > > > > -Ryan
> > > > > > >
> > > > > > > > one inch lift on a 8 foot wheelbase is about 0.9 degree
change
> > > > > > > > the sine of 0.9 degrees is 0.01
> > > > > > > > It looks like less than 1% error
> > > > > > > >
> > > > > > > >
> > > > > > > > richard
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
Original Message
> > > > > > > > From: "ryan" <rpogge@3...>
> > > > > > > > To: <basicstamps@yahoogroups.com>
> > > > > > > > Sent: Friday, May 11, 2001 11:39 AM
> > > > > > > > Subject: Re: [noparse][[/noparse]basicstamps] accelerometer
> > > > > > > >
> > > > > > > >
> > > > > > > > > Chilton-
> > > > > > > > > Its a Dodge Viper, there is VERY little lift.
> > > > > > > > > but I wonder if even an inch or two would effect the
reading
> > > > enough
> > > > > to
> > > > > > > > throw
> > > > > > > > > the results. yes the rear wheels slip durring
acceleration,
> > now
> > > > that
> > > > > I
> > > > > > > > think
> > > > > > > > > about it, this
> > > > > > > > > also causes a 'jerk' of forward acceleration when they
catch
> > > > again.
> > > > > > > > > ....another problem to overcome.
> > > > > > > > >
> > > > > > > > > I mean maybee its not even an issue, ...unless I decide to
use
> > > it
> > > > > with
> > > > > > > the
> > > > > > > > > 66 vett.
> > > > > > > > > front end of that lifts so much I loose sight of the road
(=
> > > > > > > > >
> > > > > > > > > I wanted an elegant solution, however I may resort to just
> > > taking
> > > > > > > readings
> > > > > > > > > from
> > > > > > > > > the speedometer sensor instead of an accelerometer, this
would
> > > > > > simplify
> > > > > > > > > things
> > > > > > > > > a great deal.
> > > > > > > > >
> > > > > > > > > Thanks for all the input guys and gals.
> > > > > > > > >
> > > > > > > > > -Ryan
> > > > > > > > >
> > > > > > > > > > Hey there SpeedRacer Ryan,
> > > > > > > > > >
> > > > > > > > > > Volkswagon employed special suspension devices in their
> > > Corrado
> > > > > line
> > > > > > > > > > (and hopefully we'll see 'em again in the Steppenwolf)
to
> > > > decrease
> > > > > > the
> > > > > > > > > > effects of this. It didn't hurt that it also featured a
> > > variable
> > > > > > > spoiler
> > > > > > > > > > on the back, but I don't think it affected the lift
during
> > > > initial
> > > > > > > > > > acceleration (I think I remember it shifting twice, once
at
> > 45
> > > > and
> > > > > > > again
> > > > > > > > > > at 140).
> > > > > > > > > >
> > > > > > > > > > My first question would be, where is the engine seated?
If
> > > it's
> > > > a
> > > > > > > front
> > > > > > > > > > mounted engine and you're having lift problems, it might
be
> > a
> > > > > > totally
> > > > > > > > > > different scenario than if your engine is sitting above
the
> > > rear
> > > > > > > wheels.
> > > > > > > > > > Also, are your rear wheels skidding at all during
> > > acceleration?
> > > > > And
> > > > > > > > > > lastly, is this a front wheel or rear wheel drive
vehicle?
> > If
> > > > it's
> > > > > a
> > > > > > > > > > front wheel drive vehicle and you're seeing lift, eh,
you
> > may
> > > > have
> > > > > a
> > > > > > > > > > whole other series of problems :-)
> > > > > > > > > >
> > > > > > > > > > (A hotrod junkie friend of mine tells me that the
easiest
> > way
> > > to
> > > > > > > > > > overcome this is to use two different engines, one for
the
> > > front
> > > > > > > wheels
> > > > > > > > > > and one for the back. Then he started talking about gear
> > > ratios
> > > > > and
> > > > > > > > > > efficient torque balancing or something and I fell
asleep.)
> > > > > > > > > >
> > > > > > > > > > If you *really* need a professional answer to this, I
live 4
> > > > miles
> > > > > > > away
> > > > > > > > > > from a race track, and I'm sure I can find someone out
there
> > > who
> > > > > > would
> > > > > > > > > > know. Next month we have the open Ferrari races, so I'm
> > always
> > > > > > looking
> > > > > > > > > > for a reason to tell someone, "It'd be easier if I just
show
> > > you
> > > > > > what
> > > > > > > I
> > > > > > > > > > mean. Can I have your keys?"
> > > > > > > > > >
> > > > > > > > > > -Chilton
> > > > > > > > > >
> > > > > > > > > > > My concern is that the front end of the car rises
> > > > > > > > > > > slightly during 'take off' (=
> > > > > > > > > > > Any ideas on overcomming this?
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > Your use of Yahoo! Groups is subject to
> > > > > > http://docs.yahoo.com/info/terms/
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > Your use of Yahoo! Groups is subject to
> > > > > http://docs.yahoo.com/info/terms/
> > > > > > >
> > > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > Your use of Yahoo! Groups is subject to
> > > > http://docs.yahoo.com/info/terms/
> > > > > >
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > > > >
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> > > >
> > > >
> > > >
> > >
> > >
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
>
>
>
>
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
thanks for the tip. If my project works I'll be making more then one.
Each one will require 2 '202's so I'd hat to spend $60.00 every time I
need to build one.
Cory
>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
On 5/14/01, 4:47:25 PM, "Dennis P. O'Leary" <doleary@h...> wrote
regarding RE: [noparse][[/noparse]basicstamps] accelerometer:
> Cory -
> I agree with Ryan about the eval board. But if you want to mount several
of
> the 202J series, I found a surface-to-dip board (0.72"x0.45") as part of
the
> Surf Board series at my local supply shop (also at Digikey). Use of a
fine
> soldering tip and a magnifying class helped solder the 202 to the surf
> board.
> Denis
it read anything? What I'm asking is how sensitive are these devices? Is there
a minimum acceleration rate before it starts detecting motion?
I'm working on an application where I need to know if a slow-moving robot has
gotten stuck (i.e. no acceleration when motors are turning). It has to be a
small solution because the robot I'm building is going to be used to explore
some very cramped spaces.
Thanks,
Barry
[noparse][[/noparse]Non-text portions of this message have been removed]
will it read anything? What I'm asking is how sensitive are these devices?
Is there a minimum acceleration rate before it starts detecting motion?
>
>I'm working on an application where I need to know if a slow-moving robot
has gotten stuck (i.e. no acceleration when motors are turning). It has to
be a small solution because the robot I'm building is going to be used to
explore some very cramped spaces.
>
Only a change in speed or direction would be registered. Once it is moving,
at speed everything becomes steady-state. Think about floating in a canoe
down a peaceful river with your eyes closed. Can you tell that you're moving?