EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: metrologist on March 23, 2021, 04:58:49 pm
-
Hi All,
I made the circuit shown below. R3 is a current shunt and, in my actual circuit I do not have V2 - I am just simulating the voltage across the current shunt for the analysis to work.
However, when I apply 5V power, I measure ~350mV on my output with the input grounded through the shunt. I am expecting less than 20mV. I checked the resistance from the output to several ground points and measure 114k ohms. Why am I getting so much voltage on the output?
Another idea I had was to use the circuit here: https://www.analog.com/en/products/lt1078.html (https://www.analog.com/en/products/lt1078.html) with the shunt between the +/- inputs and lower the 1M feedback resistor to 100k - except the simulation shows a wide low valley. I am not sure the advantage of that circuit vs using just one input.
-
Why am I getting so much voltage on the output?
Perhaps your LT1078 is not a genuine part.
-
It came from expensive equipment, but I did remove it and resolder to perf board. It's an SMT device. I should mention, I am just using the B side and the A inputs are floating, if that might be an issue?
-
You can wire unused section as buffer and connect IN+ to ground or some other fixed level.
-
It would be a bit howyadoin' if one channel meaningfully affected the other at DC, I think even LM358 shouldn't have such problems. But you could try.
There are some crappy-ass opamps which can't really hit the ground while input common mode is close to ground (TLC27x family) due to onset of borderline phase reversal, but it would be surprising to see it from what appears to be an LM358/LT1013-style design, optimized for single supply operation. That being said, the fix is to increase input overdrive, so you could test what's the minimum output voltage in open loop configuration, with IN+ at GND and IN- pulled to 1V or VCC.
The configuration shown on AD website appears to be a differential input amplifier. In your case, the left opamp would be tied to ground and permanently output 0V - rather boring.
-
Since I plan to connect this to a high current shunt, I had considered reverse polarity. Do I just need decoupling caps on the supply?
-
Your two op amp simulation has a bug. You seem to have a short between the inverting input and ground. I see the output waveform is a bit truncated near zero, this may be the issue. But it is not clear what sets the DC operating point of the input pins once the ground short is removed. The two inputs are free to float around. Perhaps a 1 Meg resistor to ground would be more appropriate than the short? Or a bias arrangement with two resistors setting the DC operating point to something appropriate?