ICP Signal Conditioner question 2

[Twoballcane]

Hello,

If I had an accelerometer with a sensitivity of 10mV/G, on the ICP Signal Conditioner, when and why would I go to X.1, X1, and X10? Will this change the senstivity of the accelerometer? If I go to X10 will it become more sensitive or less? Is X1 the default setting?

Thanks in advance for your time and thought.



Tobalcane
"If you avoid failure, you also avoid success."

 

[GregLocock]

You fiddle with the range setting to get a good signal out. Each of those is a 20 dB jump. In the olden days, when we were using tape recorders with an S/N of 43 dB, it was not possible to record low level signals directly, so we had to set each gain in the system to maximise the S/N ratio of the entire instrumenation string, while not clipping. A typical gotcha would be to use a bandpass filter on the input to the tape recorder. This would clip, but then would round the clipping off, so anyone who was just looking at vu meters would think they were getting a good signal. (etc etc)

These days with an S/N of 80 dB it is possible to more or less set and forget, for a given test type.

None of them is a default setting, as such. Obviously on x1 you'll know exactly what the output calibration is.

What make of amp is it? The handbook should tell you the output sensitivity, if it isn't written on the faceplate of the amp.





Cheers

Greg Locock

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

 

[Strong]

If you use a X10 gain with a 10 mv/g accelerometer, then the output signal 100 mv/g. It is more sensitive as viewed by the measurement system. You still have to be concerned with overload from high acceleration signals or amplifying low level background noise.

For X0.1 gain then output is 1 mv/g or less sensitive
For X1.0 gain is 1, so output is 10 mv/g or same as accelerometer

Walt

 

[sms]

As Greg said it is really a matter of what your data acquisition system will handle. If you have a default for 100 mv/g in the data acquisition system then put it on X10, to save time and confusion.

"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.

 

[Twoballcane]

Thanks for responding!

Greg: You brought up another question, if your signal is clipped, how would you resolve this? Would you bring down the signal down by X.1?

All: So what is my goal? To have all of my sensitivities at the same level? If I have an accelerometer at 28.5, 5.6, and .48 mV/g do I set the 28.5 to X.1, 5.6 to X1 and .48 to X10 so that the data recorder will receive 2.58, 5.6, and 4.8 mV/g?

Thanks again for your time and effort!



Tobalcane
"If you avoid failure, you also avoid success."

 

[IceStationZebra]

Your "goal" is to record your data with as much fidelity as possible without clipping. What gain you choose will depend on the input range of your particular recording device, your sensor output, and your anticipated max signal. You should allow a little extra headroom so that if your signal is a little higher than expected you won't clip.

There might also be some consideration for the ability to repeat the test. In the case of the "one time only" test you are probably better off catching the max value and loosing some fidelity.

In an ideal world all of your sensors would have the same sensitivity and you could set everything for the same gain. In the non-perfect work it is imperative that you keep detailed notes for each test run. (which sensor s/n is on which channel, amplification values, detailed sample configuration info, etc.)

ISZ

 

[sms]

That is exactly right, you want to use a gain setting that uses as much of your head room in the recording device as possible without clipping. Signal to noise ratio is also key. Unfortunately to figure that out you need to do the test otherwise you won't know for sure what magnitude of signals you are up against.

If it is a one time test be conservative in estimating g levels and set your gain accordingly.




"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.

 

[Twoballcane]

Thanks for responding!

So, if I change the "X" factor, should I be seeing the amplitude of the profile move up and down? For example, if I'm doing a sin sweep, if I change the "X" factor (for the response) to X.1 the amplitude will go down and if I put in X10 the amplitude will go up? Also, if Im signal is clipping, do I switch to X.1 and if the signal is near noise, do go to X10? I guess this will aslo applie to the control.

Thanks!

Tobalcane
"If you avoid failure, you also avoid success."

 

[IceStationZebra]

The peak-to-peak voltage will increase or decrease by a factor of 10. Saying that the amplitude goes "up" or "down" is not a good term as you can also shift the signal up or down with the zero offset. ISZ

 

[Twoballcane]

Yup I see that, but in my simple minde say for example if we have an accelerometer with sensitivity of 5.0 mV/g, if I get 1 mV back from the accel with a factor of X1, the scope would read 1/5=.200gs (right?). If at the next factor is X10, would that be 10/5=2gs? I may be thinking completly wrong, but that is why im asking.

Thanks for bearing with me, Im trying to get this thru my head.

Tobalcane
"If you avoid failure, you also avoid success."

 

[cincibcats]

No it will be 1/50*10=0.200g all the gain does is amplify the signal form the amplifier outward it will not change the actual output from the sensor itself.

 

[sms]

"if we have an accelerometer with sensitivity of 5.0 mV/g, if I get 1 mV back from the accel with a factor of X1, the scope would read 1/5=.200gs (right?)"

Right.

If at the next factor is X10, would that be 10/5=2gs?

Nope, because by applying the X10 your new scale factor to the instrument is 50 mv/g so you would have 10/50 which is still .2 g's (with apologies to the previous poster, I guess that is what you said as well).

All you are doing is taking the transducer output and multipying it by the factor on the amp. The scale factor to your instrument then needs to change appropriatly as well.

Again, typically the headroom in your instrument is fixed. All you are doing is adjusting the signal amplitude to fit in that space without clipping, and without loosing it in the noise.



"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.

 

[Twoballcane]

Thanks guys for your responses.

sms,
I think Im getting it. Another question. How does this work for the control? I have heard stories in the past where people set the wrong X factor for the control and had ten times more Gs then they wanted.

Tobalcane
"If you avoid failure, you also avoid success."

 

[GregLocock]

Best check for clipping is a scope, or headphones. An FFT is not much help.



Cheers

Greg Locock

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

 

[sms]

I guess I would have to know more about the control. What are you controlling? A shaker table? Or a force input for a modal analysis?

Generally, an amp works the same way no matter what. If you set a shaker table up with an accel and tweek it to 1 g at 100 Hz with a controler gain setting of 1, and then shift the controller to 10, the accel will produce 10g. If you then shift the controller to 0.1 the accel will now read .1g.

"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.