How to trim offset of instrumentation amplifier to 0V automatically

Thread Starter

fieryfire

Joined Feb 14, 2017
150
Hello All,

I am trying to figure out how i can digitally calibrate my outtput of an instrumentation amplifier connected to a quarter bridge. My Strain gauge sensor is continually varies with temperature through out the day. And sometimes it drifts outside the measurable range which means i loose data. Is there a way i can automatically calibrate the sensor value to 0V output by controlling it digitally. I do not want to do this manually. As it is drifts through the night. I have attached my schematics. Could someone please advise?
 

Attachments

Analog Ground

Joined Apr 24, 2019
460
How will you know the difference between the strain signal and temperature drift? Without knowing more, it looks like you need to reduce the gain of the instrumentation amplifier (IA) to stay within range at all times. The gain of your IA is fixed. You may need a different part.

Also, are the resistors in the bridge low temperature coefficient? If they drift, your signal drifts.
 

Thread Starter

fieryfire

Joined Feb 14, 2017
150
How will you know the difference between the strain signal and temperature drift? Without knowing more, it looks like you need to reduce the gain of the instrumentation amplifier (IA) to stay within range at all times. The gain of your IA is fixed. You may need a different part.

Also, are the resistors in the bridge low temperature coefficient? If they drift, your signal drifts.
For me, i am only trying to measure dynamic movements from the strain gages and not the static DC measurement. The drift from temperature is very slow, and hence i can neglect it.

The problem with reducing the gain of the IA is that i my signal to noise ratio is reduced as well. i.e. if i double the measurement range, i unfortunately half the signal to noise ratio.

The resistors of the bridge do not drift as much as the plastic material that the strain gauge is glued too. typically the strain moves at about 1000uE with every 10C while the strain gage or resistors move only 50C with 10C.
 

Analog Ground

Joined Apr 24, 2019
460
If your signal is being swamped out by the temperature drift of the sensor, you may be able to use a different bridge configuration containing another sensor which is only there to cancel out the temperature drift and has no signal. Using your nomenclature ("quarter bridge") you might use a "half bridge" with two sensors. One with strain and the other no strain. Of course, the temperature response of the two sensors must match. Here is a link explaining various bridge configurations.

https://www.hbm.com/en/7163/wheatst...MI-vLogpGl4wIVUPDACh2APAOrEAAYASAAEgIILvD_BwE
 

Thread Starter

fieryfire

Joined Feb 14, 2017
150
If your signal is being swamped out by the temperature drift of the sensor, you may be able to use a different bridge configuration containing another sensor which is only there to cancel out the temperature drift and has no signal. Using your nomenclature ("quarter bridge") you might use a "half bridge" with two sensors. One with strain and the other no strain. Of course, the temperature response of the two sensors must match. Here is a link explaining various bridge configurations.

https://www.hbm.com/en/7163/wheatst...MI-vLogpGl4wIVUPDACh2APAOrEAAYASAAEgIILvD_BwE
Hello Analog Ground, I have infact considered this option but this cannot be used in my application. My sensors are embedded inside the plastic material and is seperated from the signal conditioning interface. Would you have an option of offsetting the output of the in-amp back to zero ?
 

Analog Ground

Joined Apr 24, 2019
460
If you can differentiate signal and drift, the REF input of the IA can be used to give a dynamic offset and possibly stay within range. There are app notes on the Analog Devices site with explanations of this technique. Some are analog with very low frequency feedback to the REF pin and some are digital. The key, of course, is differentiating signal from drift.
 

Thread Starter

fieryfire

Joined Feb 14, 2017
150
Would this mean that i would have another sensor just to be able to identify the difference? and in this case be able to change the REF input for all my other sensors? would this work?
 

MrChips

Joined Oct 2, 2009
30,708
To remove DC offset, a common solution is to AC couple the amplifier. This is implemented with a high-pass filter which is also a differentiator. This method is applicable if the information of interest is rapidly changing (or turned into AC by using a chopper). Can you modulate the excitation voltage?

Bridge circuits are used to eliminate environmental effects such as temperature.

If you have an identical strain gauge sensor, you can try taking the difference between the two readings either using a difference amplifier or in software.
 

Thread Starter

fieryfire

Joined Feb 14, 2017
150
If you can differentiate signal and drift, the REF input of the IA can be used to give a dynamic offset and possibly stay within range. There are app notes on the Analog Devices site with explanations of this technique. Some are analog with very low frequency feedback to the REF pin and some are digital. The key, of course, is differentiating signal from drift.
Hello Analog,

I was just thinking about your idea, and i think this would be the best solution. What i was thinking is that my plastic material has 76 strain gauges and 2 temperature sensor. I could use the temperature sensor inside my plastic material to adjust the output value of a Digital to analog converter (which can output the entire measurement range of my system) and then connect this single DAC to the reference pin of 76 strain in-amps. Is this idea feasible? Will there be a problem if i try to connect 1 DAC output to 76 reference points of the in-amp?
 

Thread Starter

fieryfire

Joined Feb 14, 2017
150
To remove DC offset, a common solution is to AC couple the amplifier. This is implemented with a high-pass filter which is also a differentiator. This method is applicable if the information of interest is rapidly changing (or turned into AC by using a chopper). Can you modulate the excitation voltage?

Bridge circuits are used to eliminate environmental effects such as temperature.

If you have an identical strain gauge sensor, you can try taking the difference between the two readings either using a difference amplifier or in software.
Hello Mr. Chips,

I have seen this application in electonicsdesign.com but i hav never understood the theory of how it works. Currently i am powering the bridge via a 2.5V Voltage reference with as less noise as possible. But i dont understand how changing it to an ac, would work and added to it to remove the offset. Could you please elaborate more?
 

MrChips

Joined Oct 2, 2009
30,708
Can you provide us with information and data on your strain gauge sensors and a circuit diagram on the setup with the excitation voltage and instrumentation amplifier?
 

Thread Starter

fieryfire

Joined Feb 14, 2017
150
Can you provide us with information and data on your strain gauge sensors and a circuit diagram on the setup with the excitation voltage and instrumentation amplifier?
Hello MrChips,

Attached the circuit diagram named "Sensor Board".
The excitation voltage used is 2.5V with the ISL21090 voltage reference
The instrumentation amplifier used is the AD8428 with GAIN=2000. This was the best lowest noise amplifier i could find in the market with the required gain of 2000 and i still have noise in my signal :(

The sensor i use is the 350OHM
i used both WK-13-125BT-350 and the J5K-06-S5109U-350/DP
 

Attachments

Analog Ground

Joined Apr 24, 2019
460
Hello Mr. Chips,

I have seen this application in electonicsdesign.com but i hav never understood the theory of how it works. Currently i am powering the bridge via a 2.5V Voltage reference with as less noise as possible. But i dont understand how changing it to an ac, would work and added to it to remove the offset. Could you please elaborate more?
Using AC excitation of the bridge is for moving the frequency range of the signal to a more advantageous frequency. Typically, this is to move the signal out of the 1/f noise range of the preamp or avoid environmental noise such as line power pickup (fundamental and harmonics). It requires demodulation of the signal after amplification to get back the original signal. Search on "synchronous detection" or "lock-in amplifier". In this case, these techniques will not help differentiate the signal from the temperature drift. Both originate at the sensor and both will be modulated and amplified.
 

Thread Starter

fieryfire

Joined Feb 14, 2017
150
Using AC excitation of the bridge is for moving the frequency range of the signal to a more advantageous frequency. Typically, this is to move the signal out of the 1/f noise range of the preamp or avoid environmental noise such as line power pickup (fundamental and harmonics). It requires demodulation of the signal after amplification to get back the original signal. Search on "synchronous detection" or "lock-in amplifier". In this case, these techniques will not help differentiate the signal from the temperature drift. Both originate at the sensor and both will be modulated and amplified.
Thats what i thought. This technique would not be able to help with removing the offset of temperature. Thank you for the explaination.
 

Analog Ground

Joined Apr 24, 2019
460
Hello Analog,

I was just thinking about your idea, and i think this would be the best solution. What i was thinking is that my plastic material has 76 strain gauges and 2 temperature sensor. I could use the temperature sensor inside my plastic material to adjust the output value of a Digital to analog converter (which can output the entire measurement range of my system) and then connect this single DAC to the reference pin of 76 strain in-amps. Is this idea feasible? Will there be a problem if i try to connect 1 DAC output to 76 reference points of the in-amp?
Very interesting idea! you are really slammin' and jammin' now. Some thoughts:

1. If the temperature response of all sensors is predictable, then I think a compensation scheme would work. If they are all the same, then one compensating signal would do it. However, I am thinking there is a temperature gradient through your sensor assembly and they may be a bit different? Perhaps more than one compensating signal is required? Anyway, characterizing the temperature behavior may be a project on it's own.

2. A caution about the DAC output(s). DACs usually have a high output noise level. You will probably need a low pass filter between the DAC and the REF inputs. The REF signal is a sensitive input since there is a gain of only 100 for the signal before it adds to the REF signal. Then, there is a gain of 20 for both. So, any noise on the REF input may be significant.

3. The input impedance of 72 REF inputs in parallel is less than 100 ohms. So you will need to divide them into groups and drive each group with an independent buffer. Or the REF signal source will be overloaded.
 

Analog Ground

Joined Apr 24, 2019
460
Why is that?
DACs (other than PWM types) typically consist of a resistor divider string, so should be fairly quiet.
I am not sure but maybe shot noise through the switches? I think this application is going for lowest, low noise. So, my comment is just a caution. He will need buffers anyway to drive the REF inputs and I would plan double duty as low pass filters.

Edit: As I think about this issue, there are multiple noise sources in the DAC. Noise from the reference voltage, noise from the resistor ladder, noise from the output op amp and noise from the op amp feedback resistor. Adds up. How much the switches add, I don't know.
 
Last edited:

danadak

Joined Mar 10, 2018
4,057

Attachments

Top