I'm thankful for any ideas right now, so keep them coming! 👍I'm offering some suggestions purely on instinct.
That makes sense! Some schematics on the internet leave out Rf and Rin. Unfortunately, bridging them didn't make any difference.I would try making Rf=0
Yes, it's oscillating.unity gain stage
I've never used anything different. Also, I tried probing before and after some additional buffers, so technically the probes shouldn't have any effect.P.S. Make sure you're using x10 scope probes.
100u ceramic + 100n film that's all I had at hand.100n ceramic + 10u-100u electrolytic?
The problem seems to be less extrem with a NE5534 instead of the TL071. Since the NE5534 has about 20dB more PSRR it might have to do with the supply rails.
But wouldn't the problem be always there then, regardless of the pot position? 🤔
Well, turns out the caps I've been using aren't that bad. At least that's what the impedance analyzer tells me. (Being an electrical engineering student at a well equipped university has it's benefits.) 😅Could the phase shift caused by the treble capacitor be the culprit? After all it's in the feedback network and might turn the negative feedback into positive feedback that doesn't get attenuated enough depending on the pots position. Are high quality caps important for this circuit? 🤔
Attachments
Progress! Getting rid of the oscillation requires a combination of measures:
Power filtering is always important but can be challenging on a breadboard.
Using a "better" opamp certainly won't hurt and a NE5534 is by no means high tech, nor particularly expensive.
For now I'll leave it at this point. Maybe I'll add a test with the digipot at some point, but that shouldn't make a big difference.
tl;dr -> The Baxandall circuit may be "tried and tested", but implementing it isn't as trivial as it might seem at first glance.
- Clean and stable power rails.
- An opamp that isn't too sensitive to noisy supplies (high PSRR).
- Adding a capacitor from the opamp's output to its inverting input does help but it essentially creates a low pass filter.
Power filtering is always important but can be challenging on a breadboard.
Using a "better" opamp certainly won't hurt and a NE5534 is by no means high tech, nor particularly expensive.
For now I'll leave it at this point. Maybe I'll add a test with the digipot at some point, but that shouldn't make a big difference.
tl;dr -> The Baxandall circuit may be "tried and tested", but implementing it isn't as trivial as it might seem at first glance.
Attachments
From post 22, you connect the opamp as unity gain stage and it oscillates?!!
I presume you tacked a short from opamp output to minus input? Assuming the passives were left in place and R6=0, I believe this connects the output with R2, R3, R5, and CT. These parts and their associated passives certainly present an odd load network, but the impedance is modest and the opaamp should have no problem driving it in a stable manner
I suggest determining what components are provoking the oscillation. I note the oscillation is about 500kHz in photo above. I suspect wiring strays to the pot might involved. Does oscillation frequecy vary with pot position?
Would you post pictures of your setup and breadboard?
I presume you tacked a short from opamp output to minus input? Assuming the passives were left in place and R6=0, I believe this connects the output with R2, R3, R5, and CT. These parts and their associated passives certainly present an odd load network, but the impedance is modest and the opaamp should have no problem driving it in a stable manner
I suggest determining what components are provoking the oscillation. I note the oscillation is about 500kHz in photo above. I suspect wiring strays to the pot might involved. Does oscillation frequecy vary with pot position?
Would you post pictures of your setup and breadboard?
- Home
- Design & Build
- Electronic Design
- Digitally controlled analog shelving filter (low/high shelf eq)