Issue with Analogue signals

Thread Starter

Ismael O

Joined Sep 19, 2020
5
I have an issue with a analogue signal.

This signal is a 5Vdc, it is an output from a PCB which is a source to many A/D PCB. its output is connected using wire of length around 3 meters to a breakout terminal. This point distributes the signal to other 4 places (A/D PCB) using long wire ( 10 meters) . the first stage of the A/D converter is a low-noise high-precision amplifier (TLE2027AID), the output of the A/D goes to a digital system which can be accessed using a software interface to read the digital value.

The issue I have is that when I inject 5 Volts at the source (before cabling), I read 5.1 VDC using the software, How is that possible to have more voltage than the input? capacitance effect is not there because it is DC.
Somebody can tell what could be the issue so I can do some more investigation?


Thanks
 

ericgibbs

Joined Jan 29, 2010
10,809
Hi Ismael,
Welcome to AAC.
It sounds like a Common/ground wiring problem.
Do you have a block diagram of the project to post.?
E

Just show the ADC Voltage wiring and 0V wiring, with lengths and connection points
 

ericgibbs

Joined Jan 29, 2010
10,809
hi Ismael,
You need some serious noise filtering on the10 mtr cable run from cabinet 1 to cabinets 2/3.
Use an oscilloscope to check across the incoming lines to the ADC amp inputs.
Check the effective of the groundings of the cabinet .
E
 

BobaMosfet

Joined Jul 1, 2009
1,208
Measure voltage difference between different ground points. If you have any significant continual difference, you may have a ground loop. If so... how much? I'm only providing this as a means to include or exclude a possible issue mentioned above, by another poster.
 

Thread Starter

Ismael O

Joined Sep 19, 2020
5
hi Ismael,
You need some serious noise filtering on the10 mtr cable run from cabinet 1 to cabinets 2/3.
Use an oscilloscope to check across the incoming lines to the ADC amp inputs.
Check the effective of the groundings of the cabinet .
E
Thanks guys for your replays, very interesting.

noise filtering sounds costly (there are 100 analogue signals ). any other idea? I heard impedance matching,.. grounding loop, crosstalk..., any reference to those concepts in practical application?
 

ci139

Joined Jul 11, 2016
1,696
i'am sorry to note that you have not provided any circuit diagrams specifying your signal/voltage source (A/D input conditioning if any) nor the power supplies . . . it's been mostly random guessing from our side -- just a note.
---
Edit :
I read 5.1 VDC using the software
 
Last edited:

andrewmm

Joined Feb 25, 2011
574
"The issue I have is that when I inject 5 Volts at the source (before cabling), I read 5.1 VDC using the software "

If you are injecting 5v, direct into the input, then filtering etc is not the problem.
its either

a) calibration of the 5v source
b) calibration of the ADC / receiver
c) bad connections causing a raised / lowered ground level at a point, 100mV is not a lot.
d) the software is wrong
e) resolution of the ADC
 
Top