I've built 6 channels of Symasym. The last four were made with boards bought from a member of this thread, and are working fine. The first two, I built on perf. board. I want to go back and tune these up, also try the 0302/0281 devices instead of the original 1302/3201. I'd also like to optimize my grounding scheme a bit more.
Here's the problem: The first unit I built has 3-4 mV DC offset and about 0.3 mVAC residual. The second unit (ironically, a neater layout), has about 0.6mVAC and about 10mV DC offset. However, though the DC offset sits mostly around 10mV, it randomly bounces as high as 15mV or as low as 3mV, before coming back to 10-11mV. If I listen closely to the speaker, the hum also meanders up and down some, though not necessarily in sinc. with the DC. Any Ideas on where to probe?
Sheldon
edit: Another interesting note: The second unit is microphonic, but I can't locate any particular area on the board by gently tapping. The first unit is not microphonic.
Here's the problem: The first unit I built has 3-4 mV DC offset and about 0.3 mVAC residual. The second unit (ironically, a neater layout), has about 0.6mVAC and about 10mV DC offset. However, though the DC offset sits mostly around 10mV, it randomly bounces as high as 15mV or as low as 3mV, before coming back to 10-11mV. If I listen closely to the speaker, the hum also meanders up and down some, though not necessarily in sinc. with the DC. Any Ideas on where to probe?
Sheldon
edit: Another interesting note: The second unit is microphonic, but I can't locate any particular area on the board by gently tapping. The first unit is not microphonic.
Hi Sheldon,
This sounds like a bad transistor. Try installing another matched pair for each position, one pair at a time. I'd start with the diff pair. Also, use a little heatsink grease and heatshrink tubing to put the matched pairs together. This will help with thermal tracking. It will not do any harm for sure.
The only other thing I can think of would be the feedback cap to ground.
-Chris
This sounds like a bad transistor. Try installing another matched pair for each position, one pair at a time. I'd start with the diff pair. Also, use a little heatsink grease and heatshrink tubing to put the matched pairs together. This will help with thermal tracking. It will not do any harm for sure.
The only other thing I can think of would be the feedback cap to ground.
-Chris
Hi Sheldon,
Mike reports higher distortion levels, but a nicer sound with a differential fet diff pair. I've noted the same thing over many years servicing different amplifiers.
I'd say, fix your problem, then give fet inputs a whirl and let us know your opinion. Fet inputs would also allow you to increase the 22K input to ground resistor to 47K or 100K. You may have to recalculate the feedback total resistance, but then you could use a lower value feedback capacitor (to ground). The 22, 10, 5 pF (whichever you are using, output device dependant) upper pole feedback cap would need to be adjusted too. Watch for osciillation (only if you increase the 22K input resistance and feedback resistors).
-Chris
Mike reports higher distortion levels, but a nicer sound with a differential fet diff pair. I've noted the same thing over many years servicing different amplifiers.
I'd say, fix your problem, then give fet inputs a whirl and let us know your opinion. Fet inputs would also allow you to increase the 22K input to ground resistor to 47K or 100K. You may have to recalculate the feedback total resistance, but then you could use a lower value feedback capacitor (to ground). The 22, 10, 5 pF (whichever you are using, output device dependant) upper pole feedback cap would need to be adjusted too. Watch for osciillation (only if you increase the 22K input resistance and feedback resistors).
-Chris
Hi Sheldon, i second Chris's suggestion.
Fet input are always interesting, but beware of low max voltage of jfets.
An interesting mod might be using the wilson current-mirror, as shown in the "symasym - the sequel" thread. (replacing q3/q9) This improves balance in 2nd stage and reduces DC-offset.
Also, close matching of the outputdevices is helpful, especially the drivers.
There are some mods, but i can't suggest them without having tried them myself.
3-4 mV DC offset and about 0.3 mVAC residual is approximately the same i have, but VAC residual can be improved by proper wiring.
Mike
Fet input are always interesting, but beware of low max voltage of jfets.
An interesting mod might be using the wilson current-mirror, as shown in the "symasym - the sequel" thread. (replacing q3/q9) This improves balance in 2nd stage and reduces DC-offset.
Also, close matching of the outputdevices is helpful, especially the drivers.
There are some mods, but i can't suggest them without having tried them myself.
3-4 mV DC offset and about 0.3 mVAC residual is approximately the same i have, but VAC residual can be improved by proper wiring.
Mike
Here's another interesting tidbit. If I leave the amp on for an hour and everything gets nice and warm, the DC offset is stable at about 11-12mV, but the AC drops down to less than 0.1-0.2mV. And, it's no longer microphonic when I tap on the amp or the board. But, if I take the cover off and things cool a bit, the wobbly DC offset comes back, as does the AC, as does the microphonic effect.
Sheldon
Sheldon
anatech said:
Thanks Chris. I tried that, expecting to be able to find something that gave a larger response. Couldn't seem to narrow it down. Whatever it is, it does seem to be temperature dependent. I'll play around if I get a chance this week. Several other projects on the docket, and no great urgency here, so as time permits.
Sheldon
Edit: Hmm, gave it another shot with a piece of wire insulation. Seems like it might be the current source transistors or perhaps the feedback cap. A place to start anyway. Another note. I get some scratchy noises as it warms up.
Ok, some interesting findings. Disconnected the feedback cap (100uf electro), but left the bypass in place (0.12uf film). DC offset is now about 11mV, but rock stable. AC residual is gone - less than 0.1mV. So this means? I'll try a different cap.
Sheldon
Edit: Some more. Tried a 220uf cap. DC stable, but AC returns when I connect the cap to ground. It sounds like line frequency. I'll play with grounding points for a bit.
Sheldon
Edit: Some more. Tried a 220uf cap. DC stable, but AC returns when I connect the cap to ground. It sounds like line frequency. I'll play with grounding points for a bit.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.