EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: Simon on November 05, 2016, 08:46:13 pm

Title: Instrumentation amplifier problems
Post by: Simon on November 05, 2016, 08:46:13 pm
I tried building a circuit with an instrumentation amplifier. The intention was that I would take a signal from a thermistor and remove the DC voltage I did not require and then amplify what is left to make full use of my voltage rail as I am then feeding into an ADC as I don't want to be wasting bits. However despite a voltage of just over 3 V this circuit outputted nothing. Well a very very low voltage practically zero. What did I get so wrong?
Title: Re: Instrumentation amplifier problems
Post by: bktemp on November 05, 2016, 09:35:36 pm
What is the voltage between pin 1 and 4?
In your configuration pin 4 must always be more positive than pin 1, otherwise the output will be stuck at 0V.
Did you swap pin 1+4? The datasheet says pin 4 is the non inverting input.
Title: Re: Instrumentation amplifier problems
Post by: StillTrying on November 06, 2016, 12:08:36 am
Yep, looks like you may have 1,4 reversed.

And just for good measure, if you have Vs=5V, Ref=0V and a CM voltage about 3V, then you might still have another problem, it looks like you have to keep the CMV at 1.9V. A 'normal' op amp might be easier.
Title: Re: Instrumentation amplifier problems
Post by: danadak on November 06, 2016, 12:11:23 am
Page 5 shows a thermistor setup

http://www.ti.com/lit/ug/tidu491/tidu491.pdf (http://www.ti.com/lit/ug/tidu491/tidu491.pdf)


Regards, Dana.
Title: Re: Instrumentation amplifier problems
Post by: Simon on November 06, 2016, 06:50:32 am
ok yes, sounds like I have inverted the inputs, that would completely explain it. Yes I found the input/output graphs a bit confusing, it almost sounds like they say it can do one thing and then say something else.

What would be a good rail to rail opamp to do this with, that's rail to rail on inout and output. Fortunately I didn't want much gain and still get a good reading bypassing the amp for the sake of a prototype.
Title: Re: Instrumentation amplifier problems
Post by: Simon on November 06, 2016, 07:20:06 am
I could run it at 24V if that helps
Title: Re: Instrumentation amplifier problems
Post by: rstofer on November 06, 2016, 02:44:08 pm
This sounds a lot like 'offset and scale' - remove the DC offset and scale the desired signal.  If so, Chapter 4, Section 4.3 of "Op Amps For Everyone" deals with this.  They also discuss it in terms of a single supply rail-to-rail op amp.

http://web.mit.edu/6.101/www/reference/op_amps_everyone.pdf (http://web.mit.edu/6.101/www/reference/op_amps_everyone.pdf)
Title: Re: Instrumentation amplifier problems
Post by: Kalvin on November 06, 2016, 02:48:44 pm
Please, show the schematics how you have connected the thermistor and how you have biased the thermistor.
Title: Re: Instrumentation amplifier problems
Post by: Zero999 on November 06, 2016, 02:55:34 pm
If you want to run off 24V, yet interface with an ADC, then you don't need a rail-to-rail op-amp, just one which will operate at the negative rail. There are plenty of op-amps which will do this and have a low offset voltage, which I presume is important: OPA234, LT1006, LT1013. If all you need is something cheap, then there's LM358 but the offset voltagemay be too high for your application.
Title: Re: Instrumentation amplifier problems
Post by: Simon on November 06, 2016, 03:56:14 pm
If you want to run off 24V, yet interface with an ADC, then you don't need a rail-to-rail op-amp, just one which will operate at the negative rail. There are plenty of op-amps which will do this and have a low offset voltage, which I presume is important: OPA234, LT1006, LT1013. If all you need is something cheap, then there's LM358 but the offset voltagemay be too high for your application.
Indeed that is what I'm starting to think.

Sent from my Moto G (4) using Tapatalk

Title: Re: Instrumentation amplifier problems
Post by: Simon on November 06, 2016, 04:00:00 pm
Please, show the schematics how you have connected the thermistor and how you have biased the thermistor.
The thermistor goes from input to GND with a bypass cap and then pulled up to 5V

Sent from my Moto G (4) using Tapatalk