Instrumentation Amplifier for Wheatstone Bridge Output

[Micheal_Leo]

TL;DR Summary

Instrumentation Amplifier

I am going to do some experiment on mechanical mechanism , studying stresses

Thing that in my mind

Arduino
Strain gauge
Instrumentation amplifier

I want to know that which Instrumentation amplifier i should use to get good output voltage from wheatstone bridge circuit and also easy to use

Thank you very much

 

[Baluncore]

Welcome to PF.

Will you be buying a module that does the job, or will you be building the amplifier?

https://www.ebay.com.au/itm/175387368293

There are delta-sigma AD converters that do not need an external instrumentation amplifier. They range from 16 to 24 bit resolution, and are designed for use in strain gauge scales.

https://www.ti.com/lit/an/sbaa184/s...26&ref_url=https%3A%2F%2Fwww.google.com.au%2F

https://www.ti.com/lit/ds/sbas426h/sbas426h.pdf?ts=1688296275595

Some CORTEX M0 microcontrollers include a delta-sigma converter.

 

[Micheal_Leo]

Welcome to PF.

Will you be buying a module that does the job, or will you be building the amplifier?

https://www.ebay.com.au/itm/175387368293

There are delta-sigma AD converters that do not need an external instrumentation amplifier. They range from 16 to 24 bit resolution, and are designed for use in strain gauge scales.

https://www.ti.com/lit/an/sbaa184/sbaa184.pdf?ts=1688296127626&ref_url=https%3A%2F%2Fwww.google.com.au%2F

https://www.ti.com/lit/ds/sbas426h/sbas426h.pdf?ts=1688296275595

Some CORTEX M0 microcontrollers include a delta-sigma converter.

Thank you very much for welcoming
i want to buy module

i came across this now also

https://protosupplies.com/product/ad620-instrumentation-amplifier-module/

please can you tell is this also good , too much pins confused me

does this model also do not need an external instrumentation amplifier>

 

[TonyStewart]

Thank you very much for welcoming
i want to buy module

i came across this now also

https://protosupplies.com/product/ad620-instrumentation-amplifier-module/

please can you tell is this also good , too much pins confused me

does this model also do not need an external instrumentation amplifier>

If you cannot recognize the symbol of the inst. amp (IA) in your link schematic, I suggest you do more reading about common mode noise rejection and read how to use it with biasing and sensor cables, DC voltages.

Then how you must be concerned with imbalance in both wires and possibly shield them with STP wire to reduce unbalanced injection of grid E-fields.More importantly, you must learn to define your expectations or specs on overall project plans. Analog? Digital? Packaging? Power Supply or battery? and all the specs for gain, SNR and output format. Single supply or bipolar?

 

[Baluncore]

please can you tell is this also good , too much pins confused me

does this model also do not need an external instrumentation amplifier>

You need to select a strain gauge, then identify the range of the signal expected during the experiment. Then evaluate what resolution ADC is required, over what range. Only then will you know the amplification required from the IA stage, internal or external.

For slow strain gauge signals, the optimum AD converter is now a delta-sigma converter. The operation of an ADC can reject AC supply noise. The IA integrated into the delta-sigma converter block, can have better performance specifications than an external amplifier.

It does not matter what you do, so long as you start to do something. You can then change the parts to get a stepwise improvement where it is needed. In the end, the instrumentation should meet the requirements, and you will finally understand the unexpected problems you met along the way.

 

[TonyStewart]

ditto above and

Tradeoffs for any measurement system that you define are:

- input Signal Noise Ratio, minimum signal and resolution expected thus gain required, maximum signal accuracy, sources of error with a budget for total error, speed of measurement, signal bandwidth, scanning rate, to mention a few.
- analog & digital conversion separate or an integrated card ( preferred) with higher resolution Sigma Delta preferred.

Keep your grounds close, and use analog regulators for best noise reduction. Using AC gnd to 0Vdc is an option or cap to AC gnd to reduce stray hum. is a good option 1~10 nF to shunt CM E-fields.

 

[berkeman]

Quick Quiz Question for those reading this thread... What is the technical error in the circuit for U3.2, and how will that error manifest itself?

 

[Baluncore]

The root of all evil, is connected to pin 5, REF.

 

[jrmichler]

TL;DR Summary: Instrumentation Amplifier

I want to know that which Instrumentation amplifier i should use to get good output voltage from wheatstone bridge circuit and also easy to use

Sparkfun sells a strain gauge signal conditioner for Wheatstone bridges: https://www.sparkfun.com/products/1...64.1575937733.1688417106-434566512.1688417106

I have no experience with Sparkfun, but their products are designed for use with Arduino. And their amplifier costs only $10.95.

You don't say anything about required bandwidth, required linearity, accuracy, allowable range of ambient temperatures, or other constraints, so you will need to check your requirements against the data sheet.

 

[DaveE]

Quick Quiz Question for those reading this thread... What is the technical error in the circuit for U3.2, and how will that error manifest itself?

View attachment 328715

Worst PSRR design ever.
One nearly universal rule in precision analog design is that you don't use the PS voltage as part of the signal path.

Also, most of my EE career was spent getting rid of trim pots in real circuits. Real circuits are the ones that are built by other people and sold for money, over and over again. When I started in DoD work, they were essentially forbidden. Later in the high performance commercial world I inherited products full of them. It was a manufacturing and field service nightmare. Everyone thinks they can make a circuit better by adjusting it, in fact it is probably only the original designer, that knows why that pot is there. We'd get boards back where the only problem was that someone tried to fix it with a screwdriver. Plus, they are crappy resistors, both electrically and especially mechanically. It is the mark of either a one-off demonstration design (like something you might sell to a hobbyist), or a crappy EE.

As an aside, that company used IC sockets in earlier designs. When my generation took over in the late 1980's we determined that the IC sockets were a big reliability problem and eliminated them. You would not believe the grief we got from our field service people who said "we need those sockets, we change ICs all the time to fix things." Our response was, yes, we know, that's exactly the problem we are fixing. It worked great.

 

[Micheal_Leo]

Sparkfun sells a strain gauge signal conditioner for Wheatstone bridges: https://www.sparkfun.com/products/1...64.1575937733.1688417106-434566512.1688417106

I have no experience with Sparkfun, but their products are designed for use with Arduino. And their amplifier costs only $10.95.

You don't say anything about required bandwidth, required linearity, accuracy, allowable range of ambient temperatures, or other constraints, so you will need to check your requirements against the data sheet.

Welcome to PF.

Will you be buying a module that does the job, or will you be building the amplifier?

https://www.ebay.com.au/itm/175387368293

There are delta-sigma AD converters that do not need an external instrumentation amplifier. They range from 16 to 24 bit resolution, and are designed for use in strain gauge scales.

https://www.ti.com/lit/an/sbaa184/sbaa184.pdf?ts=1688296127626&ref_url=https%3A%2F%2Fwww.google.com.au%2F

https://www.ti.com/lit/ds/sbas426h/sbas426h.pdf?ts=1688296275595

Some CORTEX M0 microcontrollers include a delta-sigma converter.

the following image is mechanism static study and project planning .waiting to your kind reply

 

[Micheal_Leo]

Sparkfun sells a strain gauge signal conditioner for Wheatstone bridges: https://www.sparkfun.com/products/1...64.1575937733.1688417106-434566512.1688417106

I have no experience with Sparkfun, but their products are designed for use with Arduino. And their amplifier costs only $10.95.

You don't say anything about required bandwidth, required linearity, accuracy, allowable range of ambient temperatures, or other constraints, so you will need to check your requirements against the data sheet.

the following image is mechanism static study and project planning .waiting to your kind reply

 

[Micheal_Leo]

the strain gauges selected for this project will assist to determine stress, strain all three direction

0-60-120 or 0-45-90 with 120 ohm

Arduino will reading data from a Wheatstone bridge circuit and sending the data to a computer via serial communication. The data sent to the computer includes the voltage values calculated from the analog readings of the Wheatstone bridge circuit.

One big question

Rosette strain guages consist of three guages so should i connect all guages in wheatstone bridge and give one input to Arduino or separate make wheatstone bridge for every guages and connect to three IA , so in this ways 3(wheastone bridge) + 3( IA or ADC) ,please which one is best practice and most efficient way

One more question
To study experimental stress in the given model should we use IA or ADC

 

[Baluncore]

One more question
To study experimental stress in the given model should we use IA or ADC

You need both an IA and an ADC.
Use the Spark Fun Load Cell Amplifier - HX711.
Look at the HX711 datasheet, that shows it is both an IA, and a 24 bit delta-sigma ADC, all in the one package. It is designed to interface directly to a resistive bridge strain gauge.
https://cdn.sparkfun.com/datasheets/Sensors/ForceFlex/hx711_english.pdf

 

[Micheal_Leo]

You need to select a strain gauge, then identify the range of the signal expected during the experiment. Then evaluate what resolution ADC is required, over what range. Only then will you know the amplification required from the IA stage, internal or external.

For slow strain gauge signals, the optimum AD converter is now a delta-sigma converter. The operation of an ADC can reject AC supply noise. The IA integrated into the delta-sigma converter block, can have better performance specifications than an external amplifier.

It does not matter what you do, so long as you start to do something. You can then change the parts to get a stepwise improvement where it is needed. In the end, the instrumentation should meet the requirements, and you will finally understand the unexpected problems you met along the way.

You need both an IA and an ADC.
Use the Spark Fun Load Cell Amplifier - HX711.
Look at the HX711 datasheet, that shows it is both an IA, and a 24 bit delta-sigma ADC, all in the one package. It is designed to interface directly to a resistive bridge strain gauge.
https://cdn.sparkfun.com/datasheets/Sensors/ForceFlex/hx711_english.pdf

thank you for reply so rosette gauges consist of three so i need three hx711 amplifier.i study in google that hx711 used for weight measuring applications . does this amplifier will be applied for my mechanism

please also tell which one will be more preferable or should i buy just chip not module

 

[Baluncore]

One big question
Rosette strain guages consist of three guages so should i connect all guages in wheatstone bridge and give one input to Arduino or separate make wheatstone bridge for every guages and connect to three IA , so in this ways 3(wheastone bridge) + 3( IA or ADC) ,please which one is best practice and most efficient way

A bridge is most sensitive to strain and least sensitive to temperature changes. The way you connect the gauges will depend on what you are trying to measure. You can rotate gauges 90°, or place them on opposite sides of a sheet or structural member. Maybe post a link to the gauges you will use.

The HX711 selects between two pairs of connected inputs; (A+ & A-) or (B+ & B-). So one HX711 module can handle two bridges, one at the time.

please also tell which one will be more preferable or should i buy just chip not module

You do not have time, nor need, to fabricate the circuit boards.
Buy two of the assembled modules, then experiment with the Arduino code, the strain gauge geometry, and the connections to the gauges.

 

[Micheal_Leo]

my first choice will be 0-60-120 configuration i am not sure i can find in china
but second choice will be this BHF 350-3CA ( strains at specific point )

 

[Micheal_Leo]

my first choice will be 0-60-120 configuration i am not sure i can find in china
but second choice will be this BHF 350-3CA ( strains at specific point ) or BHF 120-3CA

 

[Baluncore]

I do not think you need rosettes.

Bridge gauges cost about $2 each.
https://www.aliexpress.com/item/1005001564881864.html

You can also make bridge gauges from four identical rectangular gauges, oriented to measure strain in the required orientation.
https://www.aliexpress.com/item/1005004705591422.html

 

[Micheal_Leo]

I do not think you need rosettes.

Bridge gauges cost about $2 each.
https://www.aliexpress.com/item/1005001564881864.html

You can also make bridge gauges from four identical rectangular gauges, oriented to measure strain in the required orientation.
https://www.aliexpress.com/item/1005004705591422.html

just an example can i make rosette strain gauge by chossing three identical strain guages and attached at angles measured precisely on component

 

[Baluncore]

just an example can i make rosette strain gauge by chossing three identical strain guages and attached at angles measured precisely on component

Yes.
If you know your principal axis, why do you need a rosette?
A bridge is a better solution, because it is balanced, and independent of temperature.

 

[Micheal_Leo]

Yes.
If you know your principal axis, why do you need a rosette?
A bridge is a better solution, because it is balanced, and independent of temperature.

yes will go for https://www.aliexpress.com/item/1005001564881864.html
wheatstone bridge will make and hx711

 

[Averagesupernova]

What is the technical error in the circuit for U3.2, and how will that error manifest itself?

No current being drawn through the wiper. That point of contact eventually becomes intermittent. At least that's one of them.

 

[berkeman]

No current being drawn through the wiper. That point of contact eventually becomes intermittent. At least that's one of them.

Perfect, yes, that's what jumped out at me as well. Pretty amateurish mistake for a schematic that got published on the web, IMO. The opamp input bias current looks to be several orders of magnitude below the minimum wiper current for reliable potentiometer operation.

That circuit will work fine on the workbench and in initial testing, but will fail in the field after a while. Operational reliability testing at 85C/85%RH for 1000 hours (Accelerated Life Testing) of the product prior to First Customer Shipment would hopefully catch this error.

 

[DaveE]

Shout out to EEPOTs. They stay where you put them. A yahoo with a tiny screwdriver can't change them, but your SW calibration routine can (even for the end user, like in your network analyzer). Any calibration adjustment MUST have a well documented, and ideally controlled, procedure.

Still not great resistors in them, but you can't have everything.

 

[Edmonton Sleepless]

Wow - great information - but I am a Mechanical Engineer - so understand only bits and pieces of the discussion! I want to measure what is expected to be very low levels of strain in a test specimen over a 1,000 hr pressure test. Is there such a thing as a plug and play system whereby I can mount the strain gauges on the test specimen (I need to measure axial strain and longitudinal strain - so as I understand one bi-axial strain gauge or 2 - 1/4 bridge strain gauges installed 90 degrees to each other), run the wires from the data logger to the strain gauge(s), to the amplifier/signal conditioner, and back to the data logger. Something that even a Mech Eng could do! Right now I am thinking that I need to hire an instrumentation company that specializes in strain measurement!
Sleepless in Edmonton

 

[DaveE]

Wow - great information - but I am a Mechanical Engineer - so understand only bits and pieces of the discussion! I want to measure what is expected to be very low levels of strain in a test specimen over a 1,000 hr pressure test. Is there such a thing as a plug and play system whereby I can mount the strain gauges on the test specimen (I need to measure axial strain and longitudinal strain - so as I understand one bi-axial strain gauge or 2 - 1/4 bridge strain gauges installed 90 degrees to each other), run the wires from the data logger to the strain gauge(s), to the amplifier/signal conditioner, and back to the data logger. Something that even a Mech Eng could do! Right now I am thinking that I need to hire an instrumentation company that specializes in strain measurement!
Sleepless in Edmonton

google, dude. I'm sure there are a ton of turn key systems out there.
You might start with these instrumentation suppliers:
https://www.ni.com/
https://www.keyence.com/
https://www.dwyeromega.com/

 

[Edmonton Sleepless]

Hi DaveE
Thanks for your reply. I will check out the suppliers you suggest. I have been pounding Google - that is how I happened onto this forum/blog. For a guy that has reviewed a ton of strain vs time plots in my career - I am a complete novice when it comes to getting accurate/repeatable data to make the plots! I find that the vendor sites have info on their devices - but not complete descriptions on how they need to be installed, the end devices required, the signal outputs generated, how to tell if you are getting appropriate inputs into the data logger, etc.... I have a bunch of questions sent out to potential suppliers and I will check these 3 out.
SiE