Hi friends
I have built a balanced line buffer recently but I am having a 28mV offset (reference to GND) from both of the outputs surprisingly. Attached is the schematics, and the offset labeled.
I tried adjusting the gain, series a caps + shunt a 10kohm into Gnd to the LME49610 non-inverting inputs, ground the lme49990 non-inverting pin, but still resulting the same offset.
I am guessing adding a trimmer at the OCM pin can eliminate it but it's a bit pointless no? Any clue on where is the offset comes from?
PS. The Gnd from every node in PCB remained as 0mV, so there is no "bad/leakage gnd" in PCB drawing.
I have built a balanced line buffer recently but I am having a 28mV offset (reference to GND) from both of the outputs surprisingly. Attached is the schematics, and the offset labeled.
I tried adjusting the gain, series a caps + shunt a 10kohm into Gnd to the LME49610 non-inverting inputs, ground the lme49990 non-inverting pin, but still resulting the same offset.
I am guessing adding a trimmer at the OCM pin can eliminate it but it's a bit pointless no? Any clue on where is the offset comes from?
PS. The Gnd from every node in PCB remained as 0mV, so there is no "bad/leakage gnd" in PCB drawing.
Attachments
Seeing as a quick review of the offset voltages and input bias currents to be expected doesn't account for your measurements, it seems likely to me that you've got some low-level oscillation going on. The LME49610 is introducing some HF phase shift so perhaps your OPA1632 isn't so happy operating with its phase margin considerably reduced. There's no plot of the OL gain for that part in the DS I have so its just speculation.
Hi Andrew,
No additional feedback line back to the lme49610 itself, I did tried to cross the feedback of opa1632 but that make the circuit stop to work. The circuit is working and singing fine but the 28mV offset is annoying.
I did tried swap with a LME49724, resulting a 26mV output offset reference to ground. When i float up the ocm pin the output offset reduced to 13mV reference to gnd. It seems like a trimmer can't be avoided in such case.
No additional feedback line back to the lme49610 itself, I did tried to cross the feedback of opa1632 but that make the circuit stop to work. The circuit is working and singing fine but the 28mV offset is annoying.
I did tried swap with a LME49724, resulting a 26mV output offset reference to ground. When i float up the ocm pin the output offset reduced to 13mV reference to gnd. It seems like a trimmer can't be avoided in such case.
The top input to the opamp is +IN.
The top output from the opamp is +OUT.
The Buffer is non inverting.
You have the feedback from +OUT back to +IN. That is positive feedback.
I think the top output needs to go to the bottom input.
And the bottom output needs to go to the top input.
The 1632 datasheet should confirm the polarity of the 4 IN/OUT
The top output from the opamp is +OUT.
The Buffer is non inverting.
You have the feedback from +OUT back to +IN. That is positive feedback.
I think the top output needs to go to the bottom input.
And the bottom output needs to go to the top input.
The 1632 datasheet should confirm the polarity of the 4 IN/OUT
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.