Any ideas on tilt switches...

deno

I have been Googling around the internet for tilt switchs.· There is alot out there, but some are very expensive, and others are only $.95, but the company has a $20 minimum order.· I have tried to homebrew a couple of different types, using a rolling ball in a tube to activate a snap switch, but there is just to much friction on the ball to detect low angles of tilt.· (Plus or minus 10 to 15 degrees of tilt)

Does anyone have a source for good tilt switchs that are reasonally priced and I don't have to buy 20 of them to meet the minimum order.·

Or...perhaps, someone has a better idea then I had for a homebrew tilt switch.· All I want to do is detect slope to either apply braking action, or more power to go up hill using two tilt switches.

The Mimsic 2125 that is sold by Parallax is a good device, but upon acceleration, the device shows to much G force, thus producing a false slope reading.· It's just to senseitive to be usful here.

Deno

Phil Pilgrim (PhiPi)

Physics is physics, deno. There's an equivalence between gravity and non-inertial (i.e. accelerating) frames of reference. Anything that registers tilt, like the Memsic or a rolling ball, will also be sensitive to the forces of motion-related acceleration. If you're accelerating downhill faster than G, for example, any "tilt" sensor will tell you that you're pointed uphill. You really can't know your orientation for sure until you stop — especially on rough terrain.

-Phil

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Still some PropSTICK Kit bare PCBs left!

deno

I'm not really going that fast.· This is for my Robo Caddy.· Which follows quit well, but when I go downhill, even though the wheels are driven by the HB25, will gather speed and "loose sight of me" by going left or right.·· My plan is to install a "slight braking" or drag system using door lock actuators to slow down Robo (a brake shoe, if you will) when ever an incline of 10 degrees or more is incountered.

On the other hand, or going up hill, of 10 degrees or more, I would like to "shift gears" and apply more power (PWM) to the HB25.

As mentioned in the above post, the 2125 is just to sensitive.· By the time I increase the offset to a point where the 2125 won't take action (speed ramping up by the HB25), I am up to about 20 degrees of tilt. A mercury switch would probably work well because the mercury itself is heavy, but I just wanted to get some input form some others who might have experience the same problem in Robotics.

Deno


Post Edited (deno) : 8/22/2008 8:19:38 PM GMT

Beau Schwabe

deno,

"A mercury switch would probably work well because the mercury itself is heavy" - You will still have the inertia that Phil was describing acting on the mercury or whatever... It doesn't really matter that it's heavy or not.


To determine tilt from an accelerometer, you also need to know how fast you are going.· So some sort of wheel encoder to monitor ground·travel.

speed = delta Distance / delta time
acceleration = delta speed / delta time


Tilt = accelerometer·- acceleration







▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

metron9

Mercury Free
http://www.scmstore.com/english/sensor/highQuality/data/CW/1300/

However I had to buy 100 of them about a buck each.

PM me for my address and you can send me 4 bucks and a sase for 2 of them.

If you put 4 on a board you can get a nice tilt sensor. by putting the angle of each a tad higher in the center you can make it more sensitive. I have used them for on off switches for battery operated devices where you turn it upside down to activate it or shake it to make it turn on etc... lots of fun things you can do with them. I still have 94 of them in the bag.



▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Think Inside the box first and if that doesn't work..
Re-arrange what's inside the box then...
Think outside the BOX!

Post Edited (metron9) : 8/23/2008 3:37:07 AM GMT

Tim-M

In trailer brake controllers, incline is sensed so that the amount of trailer brake applied varies depending on whether you are climbing or descending hills. The sensing method is simple and effective. A small pendulum hangs inside the housing that always hangs plumb because the bottom end is weighted. The top end of the pendulum is in the shape of a small arc with slots in it that act like an encoder of sorts. The slotted arc swings through an optical (led) emitter/detector pair to sense incline. This is why it's so important to set the controller to level, while parked on a level surface.

I would think that a method like this could work quite well for you and be fairly easy to home-brew.

Hope this helps,

Tim

SRLM

I was thinking along the same lines with the pendulum, but you might want to use a pot at the axis of rotation. That way you don't need to keep constant watch and count how many spaces over the pendulum is (with encoders), rather you just RCTIME it.

SRLM

I was thinking along the same lines with the pendulum, but you might want to use a pot at the axis of rotation. That way you don't need to keep constant watch and count how many spaces over the pendulum is (with encoders), rather you just RCTIME it.

Phil Pilgrim (PhiPi)

Besides the usual inertial considerations noted above, pendula have the additional disadvantage of oscillating after all motion has ceased. You may be able to damp the motion to reduce this tendency, but you still have to take inertial and acceleration issues into account.

-Phil

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Still some PropSTICK Kit bare PCBs left!

GICU812

Maybe you would be better off reading speed. If at a given power level, the speed slows down or is lower than expected, increase power, if the speed is faster than normal, decrease power \ apply brake.

I dont see a pendulum working well at all, it would give you the same problems as the accelleromoeter, and every time you hit a bump it would move, and the allready noted swinging when you stop.

I dont know what scale you are working at,Maybe by usuing to pieces of wire and a heavier ball (like a pinball)·you might overcome the friction problems of using a tube type track, the wire would have less friction and less surface area.

deno

Wow, thanks for the replys.· I have PM'ed Metron9 for some of his tilt switches.

Deno

sam_sam_sam

deno

http://www.goldmine-elec-products.com/prodinfo.asp?number=G16165

Here is one more to look I bought some these they work but you have go little more than 0* to 110* degrees to tilt them

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
··Thanks for any··that you may have and all of your time finding them

·
·
·
·
Sam