Help me fix a mistake in a A/D converter project

[Pheetuz]

I posted earlier about an A/D converter project i was working on, which was all going very smoothly untill I realised I had made a mistake.

The project I am doing involves measuring a 15mV change on a 24V signal, which i was doing by using a potential divider to step down the voltage to about 2.4V and then searching for the 15mV change on that signal.

I forgot to take into account that the 15mV change I was looking for would have been attenuated by a factor of 10 and would now be 1.5mV, way below the accuracy of my A/D converter. - Really stupid of me I know...

Basically, I need a way to take 22V off of my signal whilst leaving the other 2.4V as they are, this would be ideal as my code would not need changing!!!

Or:

I have two signal that I am comparing the difference between, one should be exactly 24V and the other should be between 24.015V and 24.2V, it is this difference that I need to measure, I was thinking that i could use an OPAMP to remove the common signal and just leave me with a value between 0.015V and 0.2V, well within the range of my A/D converter.

The only problem is that I have no experience using OPAMPs and am not sure if this is realistic? How accurate would the signal be copied from one side of the OPAMP to the other ???

Also, if anyone has any circuit diagrams of a potential circuit, it would be much appreciated!

Many thanks.

Pete

 

[karesz]

I made a mistake

Hi Pete,
You are right, take an op as i.e. an adder and use a stabile DC signal for compensating the incoming 24VDC... Thes is allday busines in the electronics .
https://en.wikipedia.org/wiki/Op-amp_inverting_amplifier#Inverting_amplifier
Refer pls to summing amp, where your compensating sigmnal has opposite polarity as the 24V & the Op is +/- (bipolar)supplyed....
Another possibilty is; you have to search for an Instrumentations Amplifier with higher commonmode value, i.e. 50Vdc....
Then do amplifyign with it for the signal, but how speed is your changing pls?
If its a quasi DC (very slow) changing, it will work with the compensating/summing amp. circuit...
K.

 

[betwixt]

Re: I made a mistake

Look at in two stages:

1. measure the 24V, this was covered in our previous question,
2. measure the difference between the rails rather than measuring each and subtracting.

You can achieve both by using a suitable op-amp, look at the data sheet for the INA117 for example which can measure differential voltages even when they are 200V above ground.

Brian.

 

[holandija]

I made a mistake

I also made many mistakes becouse I do not understand evryting what's going one heare!!!
But Im looking for all tim answer for my qestion's...

 

[Pheetuz]

Re: I made a mistake

Thank you both very much.

I will have a look at that OPAMP, the only thing that was worrying me was that some of my signal may get lost from one side of the OPAMP to the other which is a problem if the difference is more than about 5mV, hopefully wont be.

Also, I was thinking that the CMRR would have to be very high because if only a tiny part of the common signal remained then it could potentially cause a huge innacuracy at the output of the OPAMP, as I said before I have very limited experience using OPAMPs so any reassurance would be appreciated.

Again, thanks alot to both of you, Brian youve been a real help today!

/Pete

 

[karesz]

I made a mistake

Welcome!
Dont worry, the INA117 is a good selection for your job... To loss the half of signal (diff to undiff out) is better as to attenuate it to a tenth & than amplify the noisy signals...
K.

 

[Pheetuz]

Re: I made a mistake

Ok, got my hands on an INA117-P and had a few questions about its operation, havent had a chance to play with it properly yet.

Basically, on the datasheet it says that it needs a positive and negative supply (power) of between 5 and 15 volts.

The difference between the two input terminals of the OPAMP will never be a negative number, so can i just use 0V as the negative supply and +10V as the positive supply??

I only ask because I do not want the extra circuitry of having to have a negative supply on my board..

Had a few problems with the operation of the opamp today when i had a brief play with it, such as getting 0.8V at the output when the two comparison inputs were shorted together, was wondering if this was due to not having the negative power supply ???

Cheers,

Pete.

 

[karesz]

I made a mistake

Thja, as I remember; you did not writing over so exclusive wishes :-(..
Anyway Im afraiding that you will need both polarities!
Otherwise their are existing some single supply versions too_as INA333,
https://focus.ti.com/lit/ds/symlink/ina333.pdf but so far I know their arent with so extrem comonmode data specified, usually is max input Ucommon a little bit lower as positive supply voltage...
K.

Added after 8 minutes:

Maybe these applications script gives some ideas for you :
https://focus.ti.com/lit/an/slyt311/slyt311.pdf

 

[Pheetuz]

Re: I made a mistake

Drats! Was afraid that would be the case! Really should have looked at the datasheet a bit harder lol ... Hmmmm, could probably find a way of getting both polarities onto the board, just a bit of a pain.

Ermmm, had a look at the opamp that you reccomended but it seems to have very low input voltages.

Am looking for somewhere in the region of about 25V into the inverting and non inverting inputs, whilst idealy supply 0V and 5V to the power connections, whilst still retaining accuaracy and CMRR... Wouldnt happen to know of any opamps that would do the trick ??

Thanks, youve been really helpful =)

/Pete

 

[karesz]

I made a mistake

Pete,
Read the text pls ; "INA170, it is powered by a single supply voltage of +2.7 to +40 V, while the input common-mode voltage (CMV) may be any voltage between +2.7 and +60 V."
K.

 

[jacky_shine]

I made a mistake

good

Added after 1 minutes:

你们好