[Johnthurston]

I am still experiencing periodic cut outs on the Epsilon 12 circuit of my SOHA II. I mentioned this back in April and received a response from wiatrob:

The only time I have experienced nuisance trips with my production board 'II was when my power was waaaay out of spec - the HV supply was designed to prevent most all of these trips. Do you have anything else on the power circuit with the amp that might be producing surges or noise on the line?

I have very high confidence in my A/C power feed. I have a 200 amp service, with 14ga copper feeding the circuit which feeds only my amp, laptop and Squeezebox. With the exception of my clothes drier, I have no motor or inductive loads on my panel (and the behavior I experience is unrelated to my clothes drier).

So I'm looking for other ideas. What can I measure or observe to try to narrow down the cause of my problem. Back in May, I worked back through the startup proceedure/checklist for the SOHA and everything measured was within the suggested ranges. I really like the way my SOHA sounds, but having it drop out for anywhere from .5 to 2 seconds while I'm listening to it is saddening.

John Thurston
Juneau, Alaska

 

[runeight]

Hi John.

Can we assume that the dropout, since it's not related to any power usage in the house, is totally random? Does it correlate with anything?

 

[Johnthurston]

I am unable to correlate it to:
Input signal - I've tried looping a song I thought was dropping out, but I can't make it fail. For the past three months, I have driven the amplifier with my Squeezebox Receiver but the behavior is about the same as I recall from from using my Rotel Cd player.
Temperature - I thought the power supply might be running hot and unable to deliver the required power. I've been running the amp without a lid for the last month with no noticeable change in behavior.
Output - I only own one pair of headphones (Sennheiser HD530) so don't have a good way to isolate this aspect.

In short, it seems random to me

But I'm absolutely willing to try to identify some positive correlation if you can suggest something to look for

 

[Pars]

One problem that I have found with this circuit is the fact that the ground used for the error opamp (pin 3) is the e12 virtual ground. The e12 virtual ground seems to bounce around depending on relay state. I have cut this away on one of my boards and instead use the amp ground. This has seemed to be much more stable... no more false triggers.

 

[runeight]

Pars, would you be kind enough to explain this better. The e12 in the SOHA II uses the amp ground plane as its reference because it has a split supply available to it. So I think I'm not understanding what you've done.

Thanks.

 

[Pars]

If it is using the amp ground as its ground (and not the virtual ground created by the OPA551 in the standard e12) then you are doing the same thing that I did. I haven't looked at the schematic for the SOHA II, so was assuming a standard e12.

 

[runeight]

OK. Thanks for the help.

In this case the e12 is not virtual, but the actual amp ground.

John, here is something to try. Change R4E to 3k instead of 2k. This will cause the e12 to slow down its response. We just need to see if we can pinpoint what might be happening.

If the dropouts stop with this change we have a data point.

Thanks.

 

[dBel84]

I did the same to the original e12 , my soha ii is pretty much stable ..dB

 

[Johnthurston]

Quote:

Originally Posted by runeight /img/forum/go_quote.gif John, here is something to try. Change R4E to 3k instead of 2k. This will cause the e12 to slow down its response. We just need to see if we can pinpoint what might be happening. If the dropouts stop with this change we have a data point. Thanks.

I didn't have a 3k handy but I did have a 1k. I lifted one leg of the 2k and tacked the 1k in series with it.
In the first five minutes of use since then, it's dropped out once for about 1 second. I'm going to go listen to it for a bit and report back on its behavior.

John Thurston
Juneau, Alaska

 

[runeight]

OK. If this doesn't work then put the resistor back to 2k and change R1E and R2E to 47k. See what happens.

 

[Johnthurston]

I've put several hours on my SOHA II with R4E at 3k (as suggested by runeight) and find the behavior much improved. I had a very nice time listening to The Wall start to finish with no drop outs

As the original problem was intermittent, I'm hesitant to say it has gone away. I am willing to say that changing R4E from 2k to 3k correlates to a greatly decreased drop-out rate. Let's say that it continues to behave well. Where do I go from here?

John Thurston
Juneau, Alaska

 

[runeight]

If it stops dropping out I think you should leave it alone and enjoy the amp.

If there still is a problem I would start wondering if there is some kind of intermittent connection on the board somewhere that is popping one of the channels somehow.

 

[Johnthurston]

I'm resurrecting this ancient thread for some more help with my SOHA II. Since I seldom have an opportunity to listen to my music at home anymore, I've relocated my amp to my desk at work (where it serves a critical service drowning out office chatter). Since doing so, the cut-outs have become unbearable

In this location, I am perfectly ready to admit that the AC power is extremely dirty and is probably under-voltage as well. What can I do in the power-supply department to correct this situation? Can I replace the simple toroid with some sort of external regulated supply?

 

[runeight]

Do you have a computer at the desk? If so, does it have a UPS??

Maybe the protected power from a UPS will be more stable.

 

[Johnthurston]

Sorry for the delay in responding, but I've tried several things and may have a new lead.

* I put the amp on a UPS at my desk which made no difference.
* I put the amp behind a line conditioner (with and without the UPS) which made no difference.
* I monitored the mains voltage with my DMM. While there was a little variation through the course of the day, it was surprisingly stable and I could detect no correlation between input voltage and cut-outs.

*I took the amp back home and experienced very similar cut-outs.
* I pulled the top off the amp and put a 'scope on the power leads.The output of the transformer isn't a perfect sine-wave, but it sure isn't bad. One side is at 15.00V, the other is at 15.17V.
* I pulled the J1 J2 J3 jumpers and went back through the set-up procedure. This is where I have found the only thing that doesn't look spot-on.

When I get down into the "tail current for the input stage", I find the difference between TB+L and TPL on one channel is only about 15V, while on the other channel it is 20.4V. The target is 19V. This measurement moves between the channels when I swap the tubes between channels.

What is the significance of this voltage, and (how) does this relate the periodic cut-out I've been experiencing?

John Thurston
Juneau, Alaska