The NE5532 may be better but are sometimes sensitive to power supply bypass. If there are small capacitors near the op-amps for the power supply, there shouldn't be a problem.
Would the opa1642 be a better match? Seems like it, gbw is wider i see..
Any thoughts there Perry?
Ps you rock.
Any thoughts there Perry?
Ps you rock.
I'm sorry but I have no experience with that op-amp. If they're easy to get to/replace, try one type in one channel and the other in a second channel and compare.
Does 300mV of crap on the power rails seem excessive to you? I wonder if i have a ground issue or a bad large electrolytic.
Were you using the secondary ground as the reference? Just a small difference in the ground can show a significant difference in noise.
i noticed. it was volts of crap grounded to the chassis. there are ground test points scattered on the pcb, i picked one pretty close to the net to ground my probe.
If you have a multimeter with a wide bandwidth, measure the AC voltage across those points. Then connect the scope ground and see if the noise level changes on the multimeter. Introducing a new ground could cause the noise to increase.
For finding noise, I find that a battery powered audible signal tracer is better than virtually anything else.
For finding noise, I find that a battery powered audible signal tracer is better than virtually anything else.
If i try the op1642 swap, I would think the most sensitive (and potentially most rewarding) parts would be the amps on the gain pot stage.
In total there are 12 tl072's and 4 tl074's on the board. Multiple in the hp/lp and what seems to be 1 tl074 and 2 tl072's on gain pot stages.
The middle section opamps before D section have no gain.
I'm thinking about ordering a handful of each to start instead of shooting the works.....
Am I correct in thinking the gain pot stage would be the best place to start?
thanks
In total there are 12 tl072's and 4 tl074's on the board. Multiple in the hp/lp and what seems to be 1 tl074 and 2 tl072's on gain pot stages.
The middle section opamps before D section have no gain.
I'm thinking about ordering a handful of each to start instead of shooting the works.....
Am I correct in thinking the gain pot stage would be the best place to start?
thanks
I don't know enough about the circuit to help with this.
Are you experienced with replacing SMD ICs? If so, it shouldn't take much time to replace them.
Are you experienced with replacing SMD ICs? If so, it shouldn't take much time to replace them.
I am, yeah piece of cake. We have some vssop versions of the part at work I'll probably kludge in first to see if they work and then order the soic8/14 versions if I have any luck.
I hacked in our vssop(? small) versions of the amp into the so8 and 14 parts on ch5-6. Did the gain pot amp then compared, the buffer amp before D then compared, then one other stage on the input that i could see signal on with the scope. None made an appreciable difference in noise on the fft or on headphones.
Figured I'd use these before buying actual soic parts since we had a ton on hand..
Given, putting two of them on a soic14 spot was a bit of a challenge. But got it on cleanly and it worked.. checked various frequency points and response vs original parts..
Was no help in noise. 🙁
Sorry for stringing this one out....
Figured I'd use these before buying actual soic parts since we had a ton on hand..
Given, putting two of them on a soic14 spot was a bit of a challenge. But got it on cleanly and it worked.. checked various frequency points and response vs original parts..
Was no help in noise. 🙁
Sorry for stringing this one out....
The output inductors for channel 1-4 get ungodly hot under normal operation. I measured them at over 100C.
Brad at arc said they do get hot, but that seems excessive to me..
I might play around with re-enforcing ground network next, since I'm running thin on ideas.
🙂
Brad at arc said they do get hot, but that seems excessive to me..
I might play around with re-enforcing ground network next, since I'm running thin on ideas.
🙂
100C is enough to damage the board over time.
Are all of the class D sections operating from a single clocked signal?
Is the operating frequency of the class D section what it should be (would have to confirm with ARC if it wasn't locked to a crystal oscillator with the frequency printed on it).
Are all of the class D sections operating from a single clocked signal?
Is the operating frequency of the class D section what it should be (would have to confirm with ARC if it wasn't locked to a crystal oscillator with the frequency printed on it).
Trying to remember. I know ch5-6 (on 2092's) was different from ch1-4 (on a 2093). It's microcontroller controlled.
Believe it was like 340khz for ch1-4 and 320kz for ch5-6.
Believe it was like 340khz for ch1-4 and 320kz for ch5-6.
Looks like i missed the summing amp on the very front end, it is on the main board. They are balanced inputs.. Didn't swap a low noise part in there. Would be a pita..
I monkeyed with feedback caps on the main gain amp. A .047u shut it up nicely, and killed my levels. 😀 duh.. Food for thought i guess.. I could reduce gain and lower the noise floor if i really wanted too... just a few resistor changes.
I monkeyed with feedback caps on the main gain amp. A .047u shut it up nicely, and killed my levels. 😀 duh.. Food for thought i guess.. I could reduce gain and lower the noise floor if i really wanted too... just a few resistor changes.
- Status
- Not open for further replies.