Unstable Bias, blowing fuse

[OC75]

I have a late 60's Sunn Spectrum II, blowing it's 3a fuse after a few minutes of playing. The bias moves around quite a bit, never really settles. I am about ready to open it up and take a look, but looking for opinions on what to look for and what to check.

Thanks in advance,
OC

[MWJB]

What idle plate current do you have? How are you measuring it?

Does the negative voltage at pin 5 fluctuate after a while?

Check for leaky power tube coupling caps (0.25uf feeding pin 5).

[R.G.]

1. Pull the tubes out and leave it powered on. Does the fuse blow after a similar time? If yes, you have a problem with the power supply, likely filter caps or rectifier. Low-likelihood causes are problems inside the PT or OT leaking current to ground.

2. If the answer in (1) is no, then the fault is likely in the coupling caps to the output tubes or the bias circuit to the output tubes. Has the amp been recapped within the last decade? If not, do it. And while you're at it, replace all the parts in the bias voltage generation section, the bias pot, the grid leak resistors and the coupling caps from the PI. Remelt the solder joints on all of the power tube circuits.

[Gtr_tech]

If the filter caps are original, I guarantee they are wasted. Replace all the lytics in the plate and bias supply. Series stack for the first 2 sections so they won't see more voltage than they can handle, which is prob'ly the case as it sits. They didn't leave much wiggle room, and with todays higher line voltages its just cheap insurance. And then you'll have the option of using a SS rectifier if you want to.

[Chuck H]

All the above considerations are up in front. Just to add, considering the age of the amp there are other things to look at too. How old are the power tubes. Tubes generally just become weak or go up in smoke but sometimes they just go a little bad and do wierd things. Also, when was the last time the amp had a good cleaning. If, for example, you have oxidized tube pins or sockets you could be losing bias intermittantly.

[Rhodesplyr]

Ah, you have a Dynaco MkIII power amp with an extra 12AX7 for guitar :-)

http://forums.klipsch.com/forums/sto...09/markIII.jpg

I don't know if Sunn used these or not, but Dynaco used selenium rectifiers for the bias supply, and they do give out. If the bias supply rectifier is a little green cube, replace it.

I've been working on Dynas for at least 15 years--my brother's stereo consists of two MkIIIs I rebuilt with improved driver circuits. These amps are *notororious* for blowing their power supply capacitors, especially with current wall voltages. I've been in a room with one that blew and filled the room with green smoke as we ran for cover. Especially with the first stage, don't try to replace it with a single capacitor. You need two in series with divider resistors, and don't go over 40uF if you want maximum life from your 5AR4 rectifier. I'd recommend two 68uF caps in series--or two 82uF at most.

The original caps were rated 525V, but collective experience has shown that this simply isn't enough voltage headroom for the first filter stage--and possibly for the second.

The main difference I notice in the Sunn schematic is that Sunn left out the MkIII's current-sensing resistor in series with the cathodes.

For longer 6AN8 life, put a positive DC bias on the heaters to lower heater/cathode voltage in the triode section. Heater cathode leakage is a frequent failure mode.

[Gtr_tech]

Silicon diodes in the bias supply for sunns of that vintage. Yea it *is* a dynaco....even had Dyna transformers. in the early days they would use actual Dyna amps. Later they just copied them and used thier xfmrs.

[Rhodesplyr]

Yea it *is* a dynaco....even had Dyna transformers.

...which is the important point since David Hafler's output transformers were/are very nice. The 6AN8 driver tube . . . not so much.

[OC75]

What idle plate current do you have? How are you measuring it?

Does the negative voltage at pin 5 fluctuate after a while?

Check for leaky power tube coupling caps (0.25uf feeding pin 5).

Measured at pin 3 is 522v.
Neg.voltage hold steady, -70v at max setting, also hold steady at lower values.

1. Pull the tubes out and leave it powered on. Does the fuse blow after a similar time? If yes, you have a problem with the power supply, likely filter caps or rectifier. Low-likelihood causes are problems inside the PT or OT leaking current to ground.

2. If the answer in (1) is no, then the fault is likely in the coupling caps to the output tubes or the bias circuit to the output tubes. Has the amp been recapped within the last decade? If not, do it. And while you're at it, replace all the parts in the bias voltage generation section, the bias pot, the grid leak resistors and the coupling caps from the PI. Remelt the solder joints on all of the power tube circuits.

Fuse doesn't blow in standby, with tubes pulled or at idle. Just when playing. Some caps have been changed, but I have trouble identifying the function of each cap. Would pics help?

All the above considerations are up in front. Just to add, considering the age of the amp there are other things to look at too. How old are the power tubes. Tubes generally just become weak or go up in smoke but sometimes they just go a little bad and do wierd things. Also, when was the last time the amp had a good cleaning. If, for example, you have oxidized tube pins or sockets you could be losing bias intermittantly.

Was cleaned by last tech, who also changed the filter cap for the PO. He got it back still having the same problem.

I am hearing the ocean sounds and velcro sound coming from the speaker, while powered up and volume on zero.
The lowest bias setting I can achieve is 47.7mv. Bias calculator states that 70% should be at 46.9, 60% at about 40mv. It appears that the resistor between pin 8 and the bias pot has been changed, although I can't read the value.

[OC75]

Appears to be 18k?

[MWJB]

If the coldest current you can get is 47.7mA then you need to change the fixed resistor from bias pot to ground, as you don't have a useful sweep. As this resistor goes to ground just lift the end that connects to the pot & measure it to ground, you want a larger value, ideally the value that you now have, plus pot value.

522vdc isn't too bad, if you have some old sovtek 5881WXT lying around, try them around 30mA each, see if the fuse still blows.

[OC75]

Ok, just to avoid my confusion, that is the grey flameproof resistor in the middle of the attached pic, correct? It measures about 18k, which is what the circuit calls for...So maybe bump it up to around 25k?

I don't have a set of wxt's handy, but I think a buddy uses those, so he might.

Thanks to Everybody for their help.

OC

[MWJB]

As your pot at it's coldest setting delivers 0.97 of the voltage out from the bias diode, increasing the fixed resistor value to 25K or 27K isn't going to have a big, or any real impact, it's only going to give you another -0.5vdc....I would try 56K, or more, to ground. We want the bigger load resistor to maximise negative voltage available, but you really don't have much room to move in that respect at the minute. You would seem to have -72vdc after the bias diode (google bias voltage divider). You could rewire the bias supply as a voltage doubler, but that means more than changing a couple of values here & there.

Does your amp have a GZ34 as in the schem? If so, try a 5U4GB. This will drop plate current slightly. I'm surprised that the GZ34 has survived with 2xKT88, The 5U4GB will be more rugged.

...& yes, sorry, the grey flameproof 18K is the one I recommend you sub for >56K.

[MWJB]

Check that the 5VAC rectifier voltage doesn't drop with the 3A 5U4, compared to with the 2A GZ34.

[OC75]

I just check the rectifier tube, it is a new looking 5AR4.

I have a couple 5u4gb's from my Hammond, is there anything I should do before I plug it into the Sunn?

Where do I check the rectifier voltage?

THANKS FOR ALL OF YOUR HELP!!