Marshall DSL100 Humming

[bajaman]

Somebody here mentioned something about raw bias to pin 8, and cutting traces etc., to achieve this. It may have been in another Marshall DSL100 problem thread. Excuse my ignorance - what exactly is raw bias and how can it be implemented on the DSL 100? Someone else mentioned that there was no pin 8 connection?!
Many thanks for any help
cheers
bajaman

[rf7]

Glad you got it going dude ! fwiw we see probably more failures with those amps than any other on here because of those reasons and the bad part is it was one of their best selling amps so theres tons of them out there. Your right though they are an accident waiting to happen.

Unfortunately, I did not get the problem solved. It still is humming when you depress the deep switch.

[Enzo]

Not sure what you are asking about baja, raw bias is not something you implement, it just refers to the power supply. The negative bias voltage starts out as a basic negative voltage source - I call that the raw voltage - and then it is fed through a couple resistors and a filter cap or two to bring it down to the desired voltage for the circuit.

For example a typical amp might make -70v from a transformer winding, then uses a couple resistors to make -55v for the actual bias to the tubes. I would call the -55 the bias supply, and I would call the -70 the raw bias supply. That is my term, and some others may use it. I wouldn't say it is an official term in electronics.

You'd have to come up with exactly what was discussed elsewhere on pin 8 - assuming we mean a 6L6 or similar based tube. One would not bias pin 8. In a fixed bias amp it will be grounded. In a cathode biased amp, there will be a resistor to ground that develops bias, but then we would not refer to it as a raw bias in any way.

[bajaman]

thanks for the clarification Enzo - I will persevere with this DSL until the runaway bias and hum in the preamp and poweramp etc is sorted.

[AdmiralB]

You'd have to come up with exactly what was discussed elsewhere on pin 8 - assuming we mean a 6L6 or similar based tube. One would not bias pin 8. In a fixed bias amp it will be grounded. In a cathode biased amp, there will be a resistor to ground that develops bias, but then we would not refer to it as a raw bias in any way.

I suspect he's talking about the Traynor-implemented, O'Connor-hyped tying of the suppressor grid of pentodes to the bias supply. So he means pin one, not pin 8.

[bajaman]

I know what I am gonna do now - make sure the suppressor grid is tied to the cathode - never had any problems with amps that have that configuration - do you mean to tell me that this DSL crapper does not have the suppressor connected to the cathode - what were those guys at Marshall smoking when they designed this crap!!
bajaman

[bajaman]

Oh - not to forget the 220k grid stopper resistors - that's right - 220k not 5k6
Must have been strong shit - or did the schematic drawer make a mistake???
Maybe, but the goons at the factory built theis amp with the 220k resistors fitted - wonder what f@#$ups they made with their new JVM 410 model - I can hardly wait (sarcastic laugh).

[Enzo]

You will find pins 1 and 8 all neatly wired together. In the case of an amp like this, I think wiring the suppressors to the bias supply is a solution looking for a problem.

And maybe those 220k look too large, but the amp seems to work with them. Before we crack heads in the engineering department, we might consider what they were trying to accomplish. they don't design these in the basement after the softball game beer bash.

[bajaman]

OKAY -
I resoldered every connection and component on the main power board (both sides).

I hard-wired the bias adjustment board to the main board.

I replaced the 1 ohm cathode resistors.

I snipped C46 - who needs it!?

I fitted a fresh matched quad set of EL34S Groove Tubes.

I replaced both the 22n coupling caps from the anodes of the PI stage.

I replaced all three of the 47uf bias filter capacitors.

I replaced the 4 x 220k control grid stopper resistors - fitted (surprise surprise!!!) 5k6 types.

I screwed the board back in its housing, replaced all the non soldered birds nest of edge connectors, and powered the bitch up.

I measured 42mV on one side, 52mV on the other side.

10 minutes later, I measured again - 86mV and 106mV and still climbing.

I turned it off in total disgust - it's going to the dump!!!!!!
What a totally disgusting piece of shit.
bajaman

[Enzo]

One wonders WHY the cirrent was drifting up. Was the bias voltage at the grids stable? Or drifting down?

After the 10 minute warm up, and expecting about 80mv as a likely target anyway, what happened when adjusting the new readings back down? I never accept bias adjustments made without a good 20 minute warm up on anything.

[bajaman]

to be honest - i have not checked the bias at the grids yet, just monitoring the voltage across the pins or the cathode to ground connections, but please tell me how I am supposed to turn the bias back down when the bias pots ARE both already at minimum or should i say maximum!!.
bajaman

[Alex R]

Bias supply cap might be leaking a bit to earth. Take a look at the bias.

[bajaman]

ALL bias caps have been replaced - bias drops from -42 to -38 as current goes up - I paralleled R77 with another 10k resistor - maybe these things just take a long time to reach full bias??!!
bajaman

[Alex R]

Oh yeah sorry you said you'd done the three bias filter caps. Still it might be worth a look for any AC in the bias supply.

A higher value for R68 would I think lift the bias voltage more effectively that lowering R77 - but it does sound like there's a fault to trace first.

Could anyone clarify how and where the bias voltage originates on these amps? I've always been bamboozled by the schematic with its CON-this and W-that.

[Enzo]

Bias is right there on the page with the power tubes. Look to the right of the OT.

Bridge comprising D7-10 is capacitor fed from the PT. That raw DC line swings under the power tubes on the drawing, is filtered by C42, then heads through R77 to the CON1 connector to the little bias adjust board - pin 1 of CON1.

On the little board - the one with the three test points that sticks out the back of the amp near the power tubes - this raw voltage sits on the high side of the two bias control pots. The other end of the two pots are also wired together and that common returns to the main board via pin 7. That finds its way to ground through R68.

Now on the little board, the wiper of each pot returns to the main board on pins 2 and 6. That is one for each side of the push pull. Pin 2, on the main board CON1 now, is filtered by C36, and 220k R67 feeds it on to the grids of V5,6. Likewise the bias pot wiper on pin 6 of CON1 is filtered by C37, and feeds the grids of V7,8 through R69.