Hi, I am trying to make a current transducer with full differential output. It will be used for data acquisition.
The transducer uses a isolation amplifier optocoupler (U1) and a full differential instrumentation amplifier with a integrated ADC driver (IC1).

Everything is OK to the isolation amplifier, when I put 8 A on the input I get the correct output on the TP2 and TP3. But, when I connect the isolation amplifier output (U1) to the input of the differential amplifier (IC1) the output TP2 and TP3 goes down to the half of the value.

Need some help to understand whats going on. The voltage supplies are ok, the board is ok, I can't figure out whats happening.

Thanks.

 

Do the ICs share power supplies and grounds? Do you have the specs for the input impedance of the amplifier and the output of the opto?

 

Do the ICs share power supplies and grounds? Do you have the specs for the input impedance of the amplifier and the output of the opto?

Hi, thanks for helping me.
Yes, they share the same supply and GND. The output impedance of the opto is 21 ohms.
The input impedance of the of the amplifier Im not sure, I could not find it very clear in the datasheet (I don't have much experience) here is the datasheet (http://www.intersil.com/content/dam/Intersil/documents/isl2/isl28617.pdf).

I put a 1Mohm resistor on the inputs of the amplifier and even so the output of the opto drops to the half. As I know the current on the input of the amplifier was supposed to be insignificant (correct?) why could this be happening?

 

I put a 1Mohm resistor on the inputs of the amplifier and even so the output of the opto drops to the half.

Was the resistor in series with the opto output or in parallel across the amp inputs? Assuming series, your result is odd. A picture may help. Look for something silly like using the wrong pin out for the amplifier. Test the amp separately from your opto.

 

Was the resistor in series with the opto output or in parallel across the amp inputs? Assuming series, your result is odd. A picture may help. Look for something silly like using the wrong pin out for the amplifier. Test the amp separately from your opto.

Yes, the resistor was in series. I checked the pins and they are ok. Here is a table with results on both cases, with opto and amp isolated and connected.

 

The voltage supplies are ok, the board is ok, I can't figure out whats happening.

Out of curiosity, what supplies are you using? Do you have three separate supplies or is 5Vcc the same as (or derived from) +VDD2?

I also agree with Wayne that pics might help in case there's a simple wiring mistake.

 

Out of curiosity, what supplies are you using? Do you have three separate supplies or is 5Vcc the same as (or derived from) +VDD2?

I also agree with Wayne that pics might help in case there's a simple wiring mistake.

Ok, thanks. Here is the full schematic and PCB, it also show how are the supplies.
On the RC Filter marked area I am leaving those capacitor open until now and shorting the resistors.

 

I don't see anything obvious that would explain your problem. Is it possible you've got a solder bridge or some other short between the two signals going to the amp? Traces get awfully close there in a few spots, particularly threading the gap between the pads on C13.

On an unrelated note, I don't see a second via anywhere near R5 to complete the connection from R10 to R5. Am I missing something? I can't imagine leaving those disconnected would explain your current problem anyway, but I thought I'd check just in case.

 

I don't see anything obvious that would explain your problem. Is it possible you've got a solder bridge or some other short between the two signals going to the amp? Traces get awfully close there in a few spots, particularly threading the gap between the pads on C13.

On an unrelated note, I don't see a second via anywhere near R5 to complete the connection from R10 to R5. Am I missing something? I can't imagine leaving those disconnected would explain your current problem anyway, but I thought I'd check just in case.

Thanks. I solved the problem, it was a short circuit on a conector pin to the plane of the board.

 

Thanks for coming back with the final result. We had suspicion of a simple/stupid mistake - because we've all made them and know what they look like - and those suspicions were confirmed. It can be really hard to find the simple stuff. It's easy to get distracted by the more complex ideas.

Glad you got it going.