## Re: cheap accelerometer?

Expand Messages
• hmmm, now that s interesting. I assume that the AC is due to strain in different directions so that +V is from one direction and -V from the other. Is the
Message 1 of 18 , Aug 1, 2004
• 0 Attachment
hmmm, now that's interesting. I assume that the AC is due to strain
in different directions so that +V is from one direction and -V from
the other. Is the amplitude linear wrt strain force? I'll have to
experiment to see if there is enough to feed into an ADC. Probably
would have to bias it to Vcc/2 and an opamp to get something usable.

might be a bit too much for a product, though, but thanks!

Phil

--- In Electronics_101@yahoogroups.com, smartdim@a... wrote:
> I am coming into this very late.......
>
> You can construct a cheap analog accelerometer fairly easy.
>
> Take a piezo buzzer and hack it for the "wafer" portion. At one end
glue a
> small mass to it (nut).
>
> It will produce an ac voltage when it bends. If you hold on to it
opposite
> the end with the mass and shake it ac voltage will appear across
>
> Ken
• Take a pendulum (small) or a derivative of it. When accelerated the bob will make an angle theta with respect to equilibrium. Then use the pendulum and a rigid
Message 2 of 18 , Aug 1, 2004
• 0 Attachment
Take a pendulum (small) or a derivative of it. When
accelerated the bob will make an angle theta with
respect to equilibrium. Then use the pendulum and
a rigid structure around it as a variable capacitor.
Capacitance will be more or less proportional
to acceleration. This is a method used in many
accelerometer chips on the market.

You could also just buy one of these chips, they
are not too expensive.

Albert

>
> van: "Phil" <phil1960us@...>
> datum: 2004/08/01 zo AM 07:24:20 CEST
> aan: Electronics_101@yahoogroups.com
> onderwerp: [Electronics_101] Re: cheap accelerometer?
>
> thanks bruce. That is definitely good food for thought.
>
> I'm being a bit circumspect about the application but its basically
> to determine what behavior the system should have. At a minimum, I
> want to know [hi, lo, 0, -lo, -hi]. I can do something interesting
> with [hi, 0 -hi] but much more interesting things with more
> resolution.
>
> The mercury switches might work though they wont allow me to get more
> than 3 states. But I could see how it could be used to create
> something that I could use - a double throw switch, in effect. I
> could probably make something fairly easily.
>
> Do you have a source of the impact dectors? Or at least some
> keywords... I tried a google search but wallowed in a sea of
> irrelevant results.
>
> --- In Electronics_101@yahoogroups.com, "Bruce" <bruceorama@y...>
> wrote:
> > Phil,
> >
> > I think knowing the application is important. In a sense a mecury
> > switch could be considered an acell if you only need Z axis
> alignment
> > for your project (1 G in the Z axis) which could also be done with
> an
> > expensive acellerometer.
> >
> > If you are doing impact detection, maybe you could use one of
> those
> > disposable systems for shipping containers (a ball bearing captured
> > between springs, aligned along one or more axes) set up to open a
> > circuit after a shock.
> >
> > I believe there are RC airplane multi axis accelerometers sometimes
> > avaiable on ebay intended for use as a miniature gyros. (I have
> > always thought of rotating mass gyros as a sort of acellerometer
> that
> > gives you turn rate), which you could also derive from a quality
> > acell.
> >
> > Just food for thought.
> > Bruce
> >
>
>
>
>
>
>
>
>
>
>
>
• under \$1? That s my goal. I haven t seen anything in that range. very interesting idea. I played around with a couple of small pieces of metal and quickly
Message 3 of 18 , Aug 1, 2004
• 0 Attachment
under \$1? That's my goal. I haven't seen anything in that range.

very interesting idea. I played around with a couple of small pieces
of metal and quickly realized I need a larger surface area. Using a
2 1x2 pieces of copper clad PCB material seperated by a piece of
paper I got about 145 pFs at max overlap and 60 pFs with no overlap.
I think I could just use the time it takes to charge to 63% of V (RC
constant) to get a rough approximation of the C. This could be done
from the PIC. definitely worth more experimentation and could lend
itself to reasonably simple fabrication directly on a PCB.

--- In Electronics_101@yahoogroups.com, <a.mil@c...> wrote:
> Take a pendulum (small) or a derivative of it. When
> accelerated the bob will make an angle theta with
> respect to equilibrium. Then use the pendulum and
> a rigid structure around it as a variable capacitor.
> Capacitance will be more or less proportional
> to acceleration. This is a method used in many
> accelerometer chips on the market.
>
> You could also just buy one of these chips, they
> are not too expensive.
>
> Albert
>
> >
> > van: "Phil" <phil1960us@y...>
> > datum: 2004/08/01 zo AM 07:24:20 CEST
> > aan: Electronics_101@yahoogroups.com
> > onderwerp: [Electronics_101] Re: cheap accelerometer?
> >
> > thanks bruce. That is definitely good food for thought.
> >
> > I'm being a bit circumspect about the application but its
basically
> > to determine what behavior the system should have. At a minimum,
I
> > want to know [hi, lo, 0, -lo, -hi]. I can do something
interesting
> > with [hi, 0 -hi] but much more interesting things with more
> > resolution.
> >
> > The mercury switches might work though they wont allow me to get
more
> > than 3 states. But I could see how it could be used to create
> > something that I could use - a double throw switch, in effect. I
> > could probably make something fairly easily.
> >
> > Do you have a source of the impact dectors? Or at least some
> > keywords... I tried a google search but wallowed in a sea of
> > irrelevant results.
> >
> > --- In Electronics_101@yahoogroups.com, "Bruce" <bruceorama@y...>
> > wrote:
> > > Phil,
> > >
> > > I think knowing the application is important. In a sense a
mecury
> > > switch could be considered an acell if you only need Z axis
> > alignment
> > > for your project (1 G in the Z axis) which could also be done
with
> > an
> > > expensive acellerometer.
> > >
> > > If you are doing impact detection, maybe you could use one of
> > those
> > > disposable systems for shipping containers (a ball bearing
captured
> > > between springs, aligned along one or more axes) set up to open
a
> > > circuit after a shock.
> > >
> > > I believe there are RC airplane multi axis accelerometers
sometimes
> > > avaiable on ebay intended for use as a miniature gyros. (I
have
> > > always thought of rotating mass gyros as a sort of
acellerometer
> > that
> > > gives you turn rate), which you could also derive from a
quality
> > > acell.
> > >
> > > Just food for thought.
> > > Bruce
> > >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
• Phil, Here are some links to shipment shock monitors: http://www.isthq.com/main.asp?a=2&b=0&pageid=86&view=active
Message 4 of 18 , Aug 1, 2004
• 0 Attachment
Phil,

Here are some links to shipment shock monitors:

http://www.isthq.com/main.asp?a=2&b=0&pageid=86&view=active

www.shockwatch.com/pdfs/productspecpdfs/Tiltwatch.pdf

http://www.agmcontainer.com/shockmaster/smFeatures.htm

Get ready, the marketing folks make surfing for the information a very
circular experience.

--- In Electronics_101@yahoogroups.com, "Phil" <phil1960us@y...>
wrote:
> thanks bruce. That is definitely good food for thought.
>
> I'm being a bit circumspect about the application but its basically
> to determine what behavior the system should have. At a minimum, I
> want to know [hi, lo, 0, -lo, -hi]. I can do something interesting
> with [hi, 0 -hi] but much more interesting things with more
> resolution.
>
> The mercury switches might work though they wont allow me to get
more
> than 3 states. But I could see how it could be used to create
> something that I could use - a double throw switch, in effect. I
> could probably make something fairly easily.
>
> Do you have a source of the impact dectors? Or at least some
> keywords... I tried a google search but wallowed in a sea of
> irrelevant results.
>
> --- In Electronics_101@yahoogroups.com, "Bruce" <bruceorama@y...>
> wrote:
> > Phil,
> >
> > I think knowing the application is important. In a sense a mecury
> > switch could be considered an acell if you only need Z axis
> alignment
> > for your project (1 G in the Z axis) which could also be done with
> an
> > expensive acellerometer.
> >
> > If you are doing impact detection, maybe you could use one of
> those
> > disposable systems for shipping containers (a ball bearing
captured
> > between springs, aligned along one or more axes) set up to open a
> > circuit after a shock.
> >
> > I believe there are RC airplane multi axis accelerometers
sometimes
> > avaiable on ebay intended for use as a miniature gyros. (I have
> > always thought of rotating mass gyros as a sort of acellerometer
> that
> > gives you turn rate), which you could also derive from a quality
> > acell.
> >
> > Just food for thought.
> > Bruce
> >
• This is interesting. One consideration when choosing sensors is deciding whether to use static measurements or a dynamic measurements. That is to say,
Message 5 of 18 , Aug 1, 2004
• 0 Attachment
This is interesting. One consideration when choosing sensors is
deciding whether to use static measurements or a dynamic measurements.
That is to say, measuring what the acceleration force is (ex. 1 G
aligned to Z axis) vs measuring the change in acceleration. I think
the peizo accelerometer Ken described would generate a signal when the
acceleration is changing. At rest, I would expect no signal.

Another alternative, connect a spring to a mass on a sliding pot.

--- In Electronics_101@yahoogroups.com, "Phil" <phil1960us@y...>
wrote:
> hmmm, now that's interesting. I assume that the AC is due to strain
> in different directions so that +V is from one direction and -V from
> the other. Is the amplitude linear wrt strain force? I'll have to
> experiment to see if there is enough to feed into an ADC. Probably
> would have to bias it to Vcc/2 and an opamp to get something usable.
>
> might be a bit too much for a product, though, but thanks!
>
> Phil
>
> --- In Electronics_101@yahoogroups.com, smartdim@a... wrote:
> > I am coming into this very late.......
> >
> > You can construct a cheap analog accelerometer fairly easy.
> >
> > Take a piezo buzzer and hack it for the "wafer" portion. At one
end
> glue a
> > small mass to it (nut).
> >
> > It will produce an ac voltage when it bends. If you hold on to it
> opposite
> > the end with the mass and shake it ac voltage will appear across
> >
> > Ken
• In a message dated 8/1/2004 5:09:15 AM Pacific Standard Time, bruceorama@yahoo.com writes: I think the peizo accelerometer Ken described would generate a
Message 6 of 18 , Aug 1, 2004
• 0 Attachment
In a message dated 8/1/2004 5:09:15 AM Pacific Standard Time, bruceorama@... writes:
I think
the peizo accelerometer Ken described would generate a signal when the
acceleration is changing.  At rest, I would expect no signal.
----------------------------------------------

Your statement above is correct, at rest no signal. Only when the piezo is under strain do you get a voltage. I cannot remember amplitude values from the one I saw a friend build, it was many years ago.

ken
• ... At rest...or, freefall. Now that could be a problem. A piezo used this way does not measure acceleration; it measures the change in acceleration (jerk).
Message 7 of 18 , Aug 1, 2004
• 0 Attachment
--- In Electronics_101@yahoogroups.com, "Bruce" <bruceorama@y...> wrote:
> This is interesting. One consideration when choosing sensors is
> deciding whether to use static measurements or a dynamic measurements.
> That is to say, measuring what the acceleration force is (ex. 1 G
> aligned to Z axis) vs measuring the change in acceleration. I think
> the peizo accelerometer Ken described would generate a signal when the
> acceleration is changing. At rest, I would expect no signal.

At rest...or, freefall. Now that could be a problem. A piezo used this
way does not measure acceleration; it measures the change in
acceleration (jerk). No, I didn't just call you a jerk, that's just
the term for acceleration change. ;-)
Your message has been successfully submitted and would be delivered to recipients shortly.