Sallen key low pass and RC time constants

Thread Starter

Gibson486

Joined Jul 20, 2012
355
I made a Sallen Key Filter to block oscillation from a strain gauge. Oscillation was around 20 Hz, so I made the break frequency 1.6 Hz in the filter since those are passive parts that I have on hand. It did it's job. Circuit/pic is attached.

The problem is that when I extract data to an ADC (24 bits), I get a slower oscillation of around 0.1 Hz. Well, turns out that is the RC time constant of resistor and capacitor pair. If had to guess, this is do to the feedback cap? I imagine that since the strain gauge is DC, that the feedback cap, at some point stops charging and begins to discharge (at the rate of the RC time constant) while op amp adjusts. Does that sound right?

If so, how do I fix this? For some reason, something is telling me to put a resistor in parallel to the feedback cap. Howver, I have no idea what is making me think that. Any ideas? Maybe equating R1 to R2 and C1 to C2 was a naive thing to do?
 

Attachments

crutschow

Joined Mar 14, 2008
34,285
That filter should not be causing any problem, so the 0.1Hz appears to be generated by the strain gage.
What is the strain gage measuring?
What was causing the original 20Hz oscillation?
What is the output with the filter input grounded?
What is the op amp power?
 
Last edited:

Thread Starter

Gibson486

Joined Jul 20, 2012
355
What is the strain gage measuring?

-A plate that measures various items the weight less than 1 gram

What was causing the original 20Hz oscillation?

- Just pressed down on the load cell and let it oscillate until it stopped. Scope measured it at around 20 Hz

What is the output with the filter input grounded?

-Not sure what you mean..

What is the op amp power?

- What is it being powered by? 5V bench supply.

Also, I just realized that the time constant is not .1 Hz. So yeah, my theory was wrong.
 

crutschow

Joined Mar 14, 2008
34,285
What is the output with the filter input grounded?

-Not sure what you mean..
Ground the input to the filter. What is the output?
What is the op amp power?

- What is it being powered by? 5V bench supply.
That filter won't work properly unless the signal is biased above ground so no part of the AC signal goes below ground.
Otherwise it needs to be powered by a plus and minus supply.

What is the op amp model number?
 

Thread Starter

Gibson486

Joined Jul 20, 2012
355
Ground the input to the filter. What is the output?
That filter won't work properly unless the signal is biased above ground so no part of the AC signal goes below ground.
Otherwise it needs to be powered by a plus and minus supply.

What is the op amp model number?

I see....

I will have to test that later today.

about the biasing... Isn't there a DC path from the load cell? This is not ac coupled. Also, probably should have stated it in the beginning, the op amp that amplifies the signal is an AD8222, which has a reference pin that you use to bias the signal.

The op amp doing the filtering is a LTC2054...or something in that series (LTC205X).
 

Thread Starter

Gibson486

Joined Jul 20, 2012
355
Thanks...no more oscillation when I put the input to the reference.

On an unrelated note, this is going to a 24 bit ADC. Does regulator need to be a low noise one? All the ones I found seem are not only expensive, but cannot really handle too much of a drop out from the input to output and if it can, the current output drops a lot. Since the first op amp stage would be differential, wouldn't that take care of half the issues? The only thing that would matter would be that the source powering the load cell and the reference should be the same (so you can put those on a reference)?
 

crutschow

Joined Mar 14, 2008
34,285
If you post your complete circuit diagram including the strain gage circuit, and the regulator and power, we can better answer you questions.
 
Top