piezo vibration
Archiver
Posts: 46,084
Hello,
I'm using a piezo vibration sensor, the ldt xxxx from digikey, and
an LM386 amplifier circuit...
I think the thing is working, now, but it's not quite what I wanted or
maybe it won't even do what I want.
I have the sensor stuck vertically into a hobby board, and the output of
the amplifier circuit is being read by the pulsin command, in a loop,
when I flick the sensor it definitely generates a number, but it takes a
flick,
and I wanted a sensor that would be more sensitive, to something like
tapping on
the table, or reading the vibration of an electric motor...
[noparse][[/noparse]does this require something like a gyro?]
SO...how do I tweak this? can this sensor be made sensitive?
any help is appreciated
if I understand this thing correctly, the piezo sensor generates a current
when it is
compressed by bending, then the current is amplified by the LM386 circuit,
then the
stamp uses the pulsin command which is really an a to d converter, and this
is
the number I see in my debug window
[noparse][[/noparse]is this true]??
program
start:
pulsin 'into variable
debug 'result
goto start
this is the essence of my program
thanks
I'm using a piezo vibration sensor, the ldt xxxx from digikey, and
an LM386 amplifier circuit...
I think the thing is working, now, but it's not quite what I wanted or
maybe it won't even do what I want.
I have the sensor stuck vertically into a hobby board, and the output of
the amplifier circuit is being read by the pulsin command, in a loop,
when I flick the sensor it definitely generates a number, but it takes a
flick,
and I wanted a sensor that would be more sensitive, to something like
tapping on
the table, or reading the vibration of an electric motor...
[noparse][[/noparse]does this require something like a gyro?]
SO...how do I tweak this? can this sensor be made sensitive?
any help is appreciated
if I understand this thing correctly, the piezo sensor generates a current
when it is
compressed by bending, then the current is amplified by the LM386 circuit,
then the
stamp uses the pulsin command which is really an a to d converter, and this
is
the number I see in my debug window
[noparse][[/noparse]is this true]??
program
start:
pulsin 'into variable
debug 'result
goto start
this is the essence of my program
thanks
Comments
case measure the width of the pulse when it exceeds the threshold of the
stamp input that is why you have to flick it, you will need an A/D
converter to get the vibration data.
Craig
> if I understand this thing correctly, the piezo sensor generates a current
> when it is
> compressed by bending, then the current is amplified by the LM386 circuit,
> then the
> stamp uses the pulsin command which is really an a to d converter, and this
> is
> the number I see in my debug window
> [noparse][[/noparse]is this true]??
>an LM386 amplifier circuit...
>I think the thing is working, now, but it's not quite what I wanted
>or maybe it won't even do what I want.I have the sensor stuck
>vertically into a hobby board, and the output of the amplifier
>circuit is being read by the pulsin command, in a loop, when I flick
>the sensor it definitely generates a number, but it takes a
>flick,and I wanted a sensor that would be more sensitive, to
>something like tapping onthe table, or reading the vibration of an
>electric motor...
Hi rad0,
It is possible to read the output of the LDT vibration sensor with
RCtime, using an op-amp circuit:
http://www.emesystems.com/BS2rct.htm#SmallV
The vibration sensor is near the end of the article. Basically, the
vibration sensor is the input to a voltage to current converter,
which also rectifies the AC signal. The current is integrated by a
capacitor, and the voltage across the capacitor is sensed by the
Stamp input pin running RCtime.
The LM386 is an audio amplifier, not an op-amp, so it is not suitable
for this purpose. An LM358 single supply op-amp or equiv. should
work okay. Put a resistor of ~1 megaohm in parallel with the sensor
at the input to the amplifier.
As to pulsin, it is possible to get a reading from the stamp in
response to big "flicks" without any amplifier at all. The sensor
generates more than one volt into the high impedance input of the
Stamp. (Use a 10 Mohm parallel resistor) The trouble is, the "flick"
has to start within the 0.13 second window of time that the PULSIN
command waits before giving up. Smaller vibrations definitely
require an amplifier of some sort.
>if I understand this thing correctly, the piezo sensor generates a
>current when it is compressed by bending,
That is right, I think. The current charges up the capacitance of
the plastic film and that also gives a voltage. (The sensor is a
sandwich of piezoelectric kynar film between metalization layers--a
capacitor) A high impedance input like the Stamp or an op-amp can
read that voltage without disturbing it or leaking off the charge.
But you do need to put a resistor (~1 to 10 megaohm) in parallel with
the film to leak off the charge on purpose so that the film won't
develop an overriding DC offset voltage. It is also possible to
sense the current produced by the sensor directly into a short
circuit, but that is a different circuit.
-- regards,
Tracy Allen
electronically monitored ecosystems
mailto:tracy@e...
http://www.emesystems.com
alarm used for vibration sensors.
> when I flick the sensor it definitely generates a number, but it takes a
> flick, and I wanted a sensor that would be more sensitive, to something
> like tapping on the table, or reading the vibration of an electric
> motor... [noparse][[/noparse]does this require something like a gyro?]
>
> SO...how do I tweak this? can this sensor be made sensitive?
>
> any help is appreciated
Actually generates a voltage which can be quite high if given enough force.
You'll
probably want a clipper circuit to protect your amp.
> if I understand this thing correctly, the piezo sensor generates a current
> when it is compressed by bending, then the current is amplified by the
> LM386 circuit, then the stamp uses the pulsin command which is really an a
> to d converter, and this is the number I see in my debug window [noparse][[/noparse]is this
> true]??
Pulsin measures the width of a pulse, which in this case would only give you an
indication that you got some type of input assuming it had enough amplitude to
trigger the STAMP's input..
Original Message
From: "Mike DeMetz" <miked@e...>
To: <basicstamps@yahoogroups.com>
Sent: Thursday, February 21, 2002 3:23 PM
Subject: [noparse][[/noparse]basicstamps] Re: piezo vibration
> A standard piezo element should work. These are what my old Radio Shack
car
> alarm used for vibration sensors.
> > when I flick the sensor it definitely generates a number, but it takes a
> > flick, and I wanted a sensor that would be more sensitive, to something
> > like tapping on the table, or reading the vibration of an electric
> > motor... [noparse][[/noparse]does this require something like a gyro?]
> >
> > SO...how do I tweak this? can this sensor be made sensitive?
> >
> > any help is appreciated
> Actually generates a voltage which can be quite high if given enough
force. You'll
> probably want a clipper circuit to protect your amp.
> > if I understand this thing correctly, the piezo sensor generates a
current
> > when it is compressed by bending, then the current is amplified by the
> > LM386 circuit, then the stamp uses the pulsin command which is really an
a
> > to d converter, and this is the number I see in my debug window [noparse][[/noparse]is this
> > true]??
> Pulsin measures the width of a pulse, which in this case would only give
you an
> indication that you got some type of input assuming it had enough
amplitude to
> trigger the STAMP's input..
>
Yeah, thanks everyone, I was giving pulsin mystical properties...I have the
flu,
but that's no excuse...