How do accelerometers remove gravity from vertical component?? 4

[davidd31415]

I've been using shear-type PCB ICP accelerometers and out of curiosity, I'm wondering how the vertical component is removed? I've read that some accelerometers read 1G when positioned vertically but these always read 0...

 

[davidd31415]

I should have said the "vertical component from gravity," above.

I should also note these are single-axis accelerometers; I can understand how a tri-axis could handle this.

 

[IRstuff]

see:

http://www.sensotec.com/accelerometer_faq.asp?category=All

The dual-mass configuration cancels out gravity in the vertical orientation

TTFN

FAQ731-376

 

[sms]

Piezo accels don't measure acceleration, they measure changes in acceleration.

The way I look at it is that the piezo crystal is like a wet sponge. If you place a wet sponge on the counter and squeeze it by pushing down on it with your hand, water comes out, but if you release the sponge it sucks the water back in. leave it alone and the water stays where it is.

A piezo crystal does the same thing except with electrons, and in a static gravity field it quickly finds equilibrium. You have to vibrate it for the electrons to move in and out of the crystal.





"Why don't you knock it off with them negative waves? Why don't you dig how beautiful it is out here? Why don't you say something righteous and hopeful for a change?" Oddball, "Kelly's Heros" 1970

Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[davidd31415]

TTFN: Thanks for the link, I had skimmed that page as well as PCB's but did not see anything about gravity or vertical positioning; I'll give it a second read.

sms: Oh and I thought I figured it out a few hours ago when I realized my software has a "view DC value" function. When I activated it I saw different signals when the accelerometer was in the vertical / non-vertical position. The signal seemed abnormally noisy and it was very slow to stabilize (5-10 seconds); after seeing that I thought it was a ceramic and added capacitance was causing it.

When I changed to the vertical orientation the voltage change must have indicated the change to 1G and the accelerometer was then finding equilibrium. Thanks!

 

[davidd31415]

If pizeos only measure change, I'm curious about velocity output accelerometers... I've seen these made with pizeo materials, any idea how they work?

 

[sms]

A piezo velocity sensor is simply an accel with an integration circuit on board. An integration circuit is basically a preamp, and is not to complicated.

"Why don't you knock it off with them negative waves? Why don't you dig how beautiful it is out here? Why don't you say something righteous and hopeful for a change?" Oddball, "Kelly's Heros" 1970

Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[davidd31415]

sms- Would it be correct to state that velocity accelerometers do not measure constant linear velocity then?

I've read that they are used on rotating machinery so this is making much more sense to me now, would just like to double check.

Thanks for the assistance, everyone.

 

[rob768]

this is a usefull little link

http://www.sensorland.com/HowPage003.html

 

[sms]

Again, if the velocity is constant, there is no acceleration, and if the accelation is constant, there is no changing acceleration to measure, so yeah your statement is correct.

A piezo velocity sensor in a car would not measure its constant speed, it would read zero (unless you hit some bumps)

No vibration, no reading...

By the way, I use both accels and piezo velocity sensors for rotating machinery all the time, but again the point is to measure vibration. Velocity is a nice parameter for this as it is not dependant on frequency, so 0.5 in/sec is high weather the machine is running 300 rpm or 3000 rpm

"Why don't you knock it off with them negative waves? Why don't you dig how beautiful it is out here? Why don't you say something righteous and hopeful for a change?" Oddball, "Kelly's Heros" 1970

Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[Strong]

A dynamic motion with a steady/constant velocity at a specific frequency does indeed have an acceleratoion and a displacement that are mathematically related! Be careful with the automobile analogies. An ICP accelerometer has no response at DC or zero Hertz, and this has nothing to do with the readout instrument being AC or DC coupled. If you need an acclerometer with DC or 0-Hertz response, then use a capacitance type or a MEMS type.

Walt

 

[rob768]

I think all piezo meters measure accelration. Only the output is in velocity..

 

[GregLocock]

Piezo types have poor low frequency response, so are not sensitive to gravity after a while. If you really want an accelerometer to measure gravity, they are available. They use a mass on a cantilever with a strain gauge. They are fun to calibrate - just flip them over to get a 2g offset.



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[ccw]

DC coupled accelerometers do measure the gravity component. It was a handy way to do a three point field calibrate check, back in the day. On their side = 0 g, on their bottom = 1 g, on their top = -1 g. These were the precision servo controlled, force-balance units that were used in inertial platforms, etc.

Piezio units are typically outputting AC coupled signals. The electrical charge packets are generated by changes in semiconductor strains, not static strain. Think of them more like microphone diaphrams.

 

[MikeyP]

There seems to be some confusion (mainly caused by people not being careful enough with their words).

Forget about "AC or DC coupled", just concentrate on the sensor itself

Piezo-electric accelerometers produce a *charge* which is proportional to the acceleration

Piezo-resistive accelerometers produce a *resistance* which is proportional to the acceleration

You can measure a static *resistance* with a bridge circuit, just like you would do for a strain gauge. Therefore these sensors will measure static acceleration.

The only way to measure a static *charge* is to count the electrons. To do that, you have to pull the electrons down a wire to a charge amplifier which "counts the electrons" (i.e. integrates the charge) and produces a voltage proportional to the charge (the charge amplifier is often built into the sensor packaging itself these days). By taking the electrons from the sensor to the amplifier you reduce the charge in the sensor. Therefore these sensors will not measure a static acceleration.

M

--
Dr Michael F Platten

 

[davidd31415]

Thank you MikeyP. I think I understood you up until the very end...

You said charge amplifiers, which integrate the charge, are often built into the sensor packaging and that with a charge amplifier static charge can be measured using a pizeo-electric accelerometer.

You then concluded that these sensors will not measure a static accleration because the charge in the sensor is reduced when electrons are taken by the amplifier.

Were you talking about the case of a pizeo-electric sensor without a charge amplifier built in?

Your discussion of charge amplifiers lead me to believe a sensor consisting of a pizeo-electric element and charge amplifier would indeed measure static charge.

Thanks again,

David

 

[davidd31415]

I WAS confident there was an option to edit such that I could fix those embarrassing spelling errors after posting.

 

[MikeyP]

Perhaps my sentence structure was a bit tortured (like usual). It doesn't matter if the charge amplifier is inside or outside the sensor packaging. You still can't measure a static charge.

A piezo-electric material generates a charge across it when it is squeezed or strained (i.e. when a force is applied to it). The charge produced is proportional to the force.

Now, an accelerometer has a reference mass attached to one side of the peizo material and your test structure is attached to the other side.

(see

http://www.bksv.com/pdf/Measuring_Vibration.pdf)

When you accelerate your structure the reference mass accelerates with it and applies a force (F=ma) to the piezo material and a charge is produced across it (i.e some electrons move to one side of the material).

Charge is proportional to the force which is proportional to the acceleration.

So if we accelerate the transducer from rest at some constant acceleration, then there is a force F = ma applied to the piezo material and (assuming a linear stiffness k) it is squeezed by some amount x = F/k and produces a charge C.

Say that charge is 100 electrons. Now to measure the charge we pull a proportion (say 10%) of the electrons down a wire to the charge amp and count them.

So in the first instance we count 10 electrons.

Now some time later we measure the charge again, but now there are only 90 electrons left, so when count 10% of the electrons we only count 9. There are now only 81 electrons left.

So each time we take a measurement, the acceleration that we measure decreases exponentially, even though we know that the transducer is still accelerating at a constant rate.

Therefore you can't measure a constant acceleration with a piezo-electric accelerometer!

Phew, I think I have succeeded in making that a lot more complex than it actually is!

M

--
Dr Michael F Platten