Milli-volt measurement problem

[garg29]

Hi,
I hope that this is not funny only with me. I'm designing a circuit to measure some mV (milli-volts). But my multimeter is showing different voltages between same mV point & at different ground points (even when there is only one ground in whole circuit). With a small drawing I'm trying to explain my problem, as can be seen electrically the points are same but measurements are different at different points. What can be the reason of this & how can I solve this problem?.

with best regards,
garg

 

[engr.waqas]

check continuity of all ground points to make it sure they are all common

 

[karesz]

Hi,
I would check it with oscilloscope!
I assume taht your signal is in endeffect the supply voltage?
Than you have noth enough/sufficient bypassings...
Did you take the attached values with "active/supplyed Board", then are here some board-current efects in game ?
K.

 

[garg29]

I think I'm not getting what you are trying to say

 

[betwixt]

I can't see anything unusual in this. If you are measuring AC mV it is just circuit noise imposed across the supply wires. If it's DC, you are seeing resistive drops from current flowing in the connecting wires. The only puzzling thing is the reverse polarity of one of the measurements.

The voltages are extremely low and the power supply only appears to be putting out 1.2mV. What powers the remainder of the circuit and are the power supplies isolated? You may have more than one ground or supply line from different power sources with slightly different offsets on them.

Brian

 

[garg29]

Thanks for replying Brian. I'm having only one ground. I'm having homemade power supply which provides me +12v -12v +5v & a variable 0-100mV source. So, what should I do to solve the problem??? :-( I'm puzzeled

Anyways,
Thanks a lot once again

 

[betwixt]

Are you sure there is only one ground in the whole circuit.
Is there a ground on your home made supply and another ground on the mV supply?

Brian.

 

[albbg]

I think you should measure the voltages at the same points but with no connections between the power supply (I mean the 0-100 mV output) and the circuit. In this way you can check the no load voltage at the generator output and also if there is a reverse biasing on your circuit. Take into account that you are handling very low voltage signals, so even the bias of op-amps could interfere with the measurement.
In particular I suspect that you have a reverse (in polarity) bias at you circuit input that sums over the wires resistance with the generator output.
How do you measure the voltages ?

Regards.

 

[garg29]

betwixt: Thanks for replying. There is only one Ground in the whole circuit. my home made supply contains all +12v -12v +5v and the mV supplies together running on same ground.

albbg: Thanks for replying. Under no load conditions if my mV source shows 0.223 millivolts then under loaded condiotions it shows 1.452 mV. I made my ground stronger my ground problem solved somewhat but this voltage change problem under no load condition is still there, what about this?

Actually, what I'm trying to do is, I'm simulating the millivolt source as a thermocouple output under different temperature conditions. I'm attaching my first stage circuit diagram here, please have a look.

Thanks Once again,

With best regards,
garg

 

[FvM]

Seems like your mV source has an output impedance of 10 ohm, it would be a reasonable value in my opinion. Unfortunately
you're driving a huge current of 120 µA into the thermocouple, which is inappropriate for any thermocouple type, I think.
Small thermcouples can have up to 50 or 100 ohm resistance, so you get a temperature error of up to 200 degree C.
You may want to use at least 10 M resistors for the open input detection circuit.

Regarding ground potential differences: They are expectable, if a DC current is flowing through the ground net. Because you
didn't tell design details, there can't be said much more. As a special problem, has you're mV source an isolated supply? If not,
you should be aware that part of the supply current is flowing through your measurement ground.

 

[garg29]

Fvm Thanks for your reply. I think you are right. Should I use 10M resistors in place of both 100k resistors(R5 & R14). You said "you didn't tell design details" can you please tell me what more details I can provide to you?
Regarding isolated mV source : if I make seperate ground for my mV source will it be better?

Thanks Once again,

With best regards,
garg

 

[FvM]

if I make seperate ground for my mV source will it be better?

I was refering to you measurement that shows different voltages at both ends of the connection from mV source to the PCB. If this
measurement is correct, it means, that a considerable current must be flowing through this connection. The only plausible
explanation would be a ground loop.

You should build your instrument in a way, that no mV connection can be affected by interfering currents.

 

[garg29]

FvM Thanks a lot. You always help me out. last time also you only helped me out from my RTD problem.

One last thing before I face any second problem hehehe.... ;-)
Should I use 10M resistors in place of both 100k resistors(R5 & R14).

 

[FvM]

Should I use 10M resistors in place of both 100k resistors(R5 & R14).

Yes. You still get a small (about 1/4 degree C) temperature error.