Electronics > Projects, Designs, and Technical Stuff
Extremely weird issue with op amps!
(1/3) > >>
OM222O:
Hello
I designed a board for a project which was a switchable constant current source and it could select between 1A and 100mA using a divide by 10 voltage divider. after assembling the PCB I found that it wasn't working. I tried everything I could before giving up and mounting the components on DIP adapters in order to isolate the issue to one component on a breadboard. Everything worked fine, except the op amp! I was using a TLV9002. I didn't have all the right components as through hole so hacked together a divide by 10 circuit by putting 3 1k resistors in parallel and 3 1k resistors in series and instead of the intended 1V referance to the op amp, I used a 3.3V (easily available on an MCU) and a 10ohm resistor instead of the intended 1ohm as the breadboard nor the dupont connectors would not be able to maintain 3.3A going through them. Here is the schematic of the breadboard setup:


after checking the voltage across the 10ohm shunt, it was 2.68 instead of the expected 0.33V!


I searched around and found an MCP6002 which is a direct replacement (pin to pin) for the TLV9002 and after literally just removing the adapter board, I inserted the MCP6002 without moving a single wire on the breadboard and bam! it works just fine! (ignore the fact that it's showing 0.37 ... it's a really cheap multimeter and the ball park is good enough here)


I checked the datasheets again and confirmed that I'm not crazy and the pin outs are the same and that TLV9002 is better in every single aspect compared to the MCP6002  |O I've been scratching my head as to why it might not be working properly  :-// I thought maybe it's a weird thing with the base current, but then again the current going through the 10ohm resistor is being monitored and limited to 33mA which is base current + emitter current and emitter current is base current * beta which in a darlington pair should be > 1000 so I have absolutely no clue as to what might be wrong here. Ironically enough the better op amp is the one that's acting weird and I have no idea why  :wtf: :scared: :palm:
duak:
Do you have an oscilloscope handy? If not, can you set your DMM to check the output pins and +5V for an AC signal?  I will bet it's oscillating for the following reasons:
1.) I don't see a +5V bypass capacitor right beside the opamp - I would try a 100n in parallel with 10 uF.
2.) these opamps are unity gain stable but when arranged like this with the sense voltage amplified X10 I'll bet it is not.

Cheers,
David Hess:
If it worked than it was an accident.  Adding gain and delay like that within the feedback loop is almost always a problem.

The added stage of amplification is not doing anything useful.  Apply the feedback directly and attenuate the 3.3 volt signal instead.
OM222O:
I had a pack of 10 MCP6002s and every single one of them worked fine without decoupling capacitor. This seems to be specific to the TLV9002 that I have. it might be fried for some reason but I don't think that's the case. I'll replace it tomorrow and share the results.
Kleinstein:
The OPs used are relatively slow (1 MHz GBW) so chances are it can work without decoupling. Still the TLV9002 might be a little more sensitive to missing decoupling. The breadboard adds quite some capacitance - this might be just enough to keep it stable by chance. The pinout of the OPs is in a way that capacitance between adjacent pins tends to help with stability.
Navigation
Message Index
Next page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod