Bugera 1990 Channel Switching Problem

[Neutronarmy]

Hello,

I had a Bugera 1990 on the bench that randomly channel switched. After having read many accounts of failed regulators on the V22, I replaced the LM7815/LM7915 on the power amp board and the LM7805 on the DSP board. I also touched up any solder joints that seemed questionable on the various board. I ran the amplifier for three hours to ensure that heat wouldn't channels switch and no problems were apparent. Admittedly I didn't run it at full volume with a signal generator, just left it off of standby and played on it every fifteen minutes to ensure it continue to function.

The 1990 was returned to the customer and within 15 minutes of him playing at bedroom levels it began to channel switch again (always from overdrive back to clean). He did note that he could switch it on the front panel back to overdrive and it would function normally before shortly switching back to clean.

I am stumped and the lack of documentation on these amps makes it even more difficult. The only other information I know is that it had a failed 6L6 and was retubed with a quad of Ruby 6L6GCMSTR. I checked the bias with a probe and it was sitting at ~40mA per 6L6. Any recommendations?

Thanks in advance for any help,
Chris

[Jazz P Bass]

Bugera 1990 Schematic

Well, here is the documentation that I have on the 1990 (NON INFINIUM).
The clean channel works in the relay off position.
Lead, they must be on.
The relays work off of the +24 Vdc rail with a MPSA06 transistor that supplies ground to the relays.
I cannot tell how the Lead On command is routed to the A06 without having the amp in front of me.
Other than the front panel switch.
Maybe you have a bad switch.

[Neutronarmy]

Well, here is the documentation that i have on the 1990 (NON iNFINIUM).
The clean channel works in the relay off position.
Lead, they must be on.
The relays work off of the +24 Vdc rail with a MPSA06 transistor that supplies ground to the relays.
I cannot tell how the Lead On command is routed to the A06 without having the amp in front of me.

Thank you so very much for the documentation. Much internet searching was absolutely fruitless on finding any kind of tech document.

If the clean channel is the default state of the relay switching (which makes sense as it is what it defaults to upon power on), is it safe to say the fault is somewhere between the 24V line and the relay? If I'm reading the schematic right and the MPSA06 is being flaky, would that also force the relay back to its default state?

When I get the amp back I'll try to check the 24V rail for ripple and make sure nothing looks too fishy. Any other suggestions? Also, any further reading on relay switching? It isn't something I've had to troubleshoot regularly yet and I'd love to get properly versed.

Thanks again so much!
Chris

[Jazz P Bass]

As a starting point, I would monitor the base leg of the MPSA06.

When the transistor is on (base high) the relay coil gets a ground path for the 24 volts that is always (check it) on the coil.
Which energizes the relay.

You need to find out what is right & what is incorrect when the failure occurs.

[Neutronarmy]

As a starting point, I would monitor the base leg of the MPSA06.

When the transistor is on (base high) the relay coil gets a ground path for the 24 volts that is always (check it) on the coil.
Which energizes the relay.

You need to find out what is right & what is incorrect when the failure occurs.

Excellent. Thank you again. I'll try to document all I can once I get the amp on the bench and let you know the outcome.

[g1]

It's going to be pretty tough if you can't reproduce the problem. You may have to get him to make the problem occur in your shop.
I'm not sure if in the dsp version maybe the switching is run differently. You mentioned a 7805, that reminded me of when the 6100 Marshalls came out. Guys were complaining about switching lock-up but we could never reproduce the problem. Turned out it only occurred in places with poor AC, like clubs. We simulated brown-outs with the variac (down around 100V I think) and were finally able to re-create the problem. I believe it had to do with the 7805 dropping out.
Probably nothing to do with your problem, but just to illustrate you must be able to reproduce the fault before you can fix it.

[audiopete]

The tube heaters are wired up all in series off the +-24VDC supply. Is a wonky tube heater shorting out one side and affecting the switching? Try tapping the tubes and see if there is any reaction.

Surprised there is even a schem of this amp. When your business model is "rip off all your designs from other companies", you tend not to put the proof out in the form of technical documents.

[Neutronarmy]

It's going to be pretty tough if you can't reproduce the problem. You may have to get him to make the problem occur in your shop.
I'm not sure if in the dsp version maybe the switching is run differently. You mentioned a 7805, that reminded me of when the 6100 Marshalls came out. Guys were complaining about switching lock-up but we could never reproduce the problem. Turned out it only occurred in places with poor AC, like clubs. We simulated brown-outs with the variac (down around 100V I think) and were finally able to re-create the problem. I believe it had to do with the 7805 dropping out.
Probably nothing to do with your problem, but just to illustrate you must be able to reproduce the fault before you can fix it.

In relation to your Marshall anecdote: what did you do once the problem was deduced? I think you might be on to something as the customer lives in a rural area with potentially questionable wall power. I have yet to get this amp to act up on the bench. The customer claimed it occurred consistently within 15 minutes of play. 45 minutes in and no issue. I just set up an MP3 player through the customer's exact settings and I'm hoping it'll act up in the next few hours. Thanks again for the insight.

audiopete - Thank you for your idea regarding the filament supply. I tried tapping the tubes with pencils erasers while the amp was running and wiggling within the sockets and no failure. I'll keep that potential failure point in mind.

[g1]

In relation to your Marshall anecdote: what did you do once the problem was deduced? I think you might be on to something as the customer lives in a rural area with potentially questionable wall power. I have yet to get this amp to act up on the bench.

Can you simulate a low power condition with a variac or a lamp limiter/dim bulb tester? In this case I am skeptical about them using a CPU controlled switching as they are probably using the same switching system as the non-dsp version of the amp.
Like JazzP said, you need to observe the +24V line when the problem occurs.
With the Marshall, I think we used low dropout 5V regulators, some tried changing the delay on the reset line to the CPU at power up:
Re: Marshall 6100 5881 > EL34 - AMPAGE Archive

[Neutronarmy]

Can you simulate a low power condition with a variac or a lamp limiter/dim bulb tester? In this case I am skeptical about them using a CPU controlled switching as they are probably using the same switching system as the non-dsp version of the amp.
Like JazzP said, you need to observe the +24V line when the problem occurs.
With the Marshall, I think we used low dropout 5V regulators, some tried changing the delay on the reset line to the CPU at power up:
Re: Marshall 6100 5881 > EL34 - AMPAGE Archive

The amp I have on the bench is the non-Infium 1990 - so it doesn't have the microcontroller bias regulation. However, it has DSP for the reverb on the clean/lead channels. The lead/clean switch is also momentary so the switching logic is occurring somewhere - I'm just not certain where (I may be totally overlooking it on the schematic). The digital board houses an AVR and two Cool Audio ICs if I remember correctly. I assume the switching code is on the AVR.

Several hours of audio signal from an MP3 player at customer specified settings still led to a consistently functioning amplifier. I'll knock the line voltage down with my variac tomorrow and see if I can get it to flake out and channel switch. Thanks for the link discussing the Marshall 6100 and all of the help!

[audiopete]

So the schem posted is not for your version of this amp? Pretty simple relay switching in the schem.

Have you looked for bad solder connections on the board? I've had Fender Hot Rod series amps with phantom switching problems and it turned out the dropping resistors for the low voltage were getting so hot they were desoldering themselves. It's not the same psu in this amp but a bad solder could be producing the problem you are having.

[Jazz P Bass]

On the posted (non Infinium) schematic, the relay control shows a conection to header X55.
Where does X55 go?
(the schematic is a bit vague)

[Jazz P Bass]

I notice that the footswitch jack goes directly to the DSP board.
I can see why that would be for the Reverb on/off but channel switching?
I will try to dig up tyhe DSP circuit.

[Neutronarmy]

On the posted (non Infinium) schematic, the relay control shows a conection to header X55.
Where does X55 go?
(the schematic is a bit vague)

I still have yet to witness the phantom switching. The client says it has happened with and without the footswitch and both without and without new tubes (it had a blown 6L6 when it was first given to me and I retubed it with a fresh quad of Ruby 6L6GCMSTR and biased it). The client does have a mixture of used tubes in the preamp - so I'm still considering that option if it acts up.

X55 is the 20-pin ribbon connector between the front panel board (featuring the pots/input/channel switch/relays) and the Bugera DSP board (Bugera BDM1A). The NC poles of the momentary DPDT channel switch go to pin 8 of the ribbon connector (when read from left to right facing the front panel board). The footswitch jack goes to the DSP board as well through a separate connector. If I am understanding this correctly, the AVR is likely doing the switching of both channels and reverb.

There is always voltage on the relay coil and ground is actively switched. D19 tests fine. There is 25.7V on relay - 145mV ripple. 10.44V at input of 5V regulator - no ripple. 4.97V out - no ripple.

MMBTA06 Voltages
Base
Lead - 739mV
Clean - 5mV

Collector
Lead - 20mv
Clean - 25.8V

Emitter
Lead - 0V
Clean - 0V

I'm going to try again to get it to freak out with time and an input signal. I'm also going to try backing off the line voltage with a variac and see when it causes ruckus on the switching circuit. Super strange.

[Neutronarmy]

I just ran the amplifier for over 30 minutes with the variac at 100VAC and I still witnessed no channel switching. The only time I could get the channel switching to fail it was at ~70VAC and the channel switching failed entirely, not merely switching back to the clean channel. Perplexed is what I am.