Marshall DSL 401 biasing question

[JZRepair]

I've got this amp on my bench.  New tubes installed and bias is still a bit high.  Forgive me now.  My brain has been so fried and foggy lately that I'm struggling with all that's going on in this schematic.  I've attached it for referency, and I believe all I need to do is tweak R36's value which will ultimately give me a bit more swing on the bias voltage.  But like I said, my brain apparently doesn't want to work anymore and I just need to verify this is the correct resistor to adjust.  I think the stress is finally getting to me.

[dwinstonwood]

I have this page in my bookmarks since I tend to forget which resistor varies the range of the pot's sweep:
https://el34world.com/charts/Biascircuits.htm

[JZRepair]

I have this page in my bookmarks since I tend to forget which resistor varies the range of the pot's sweep:
https://el34world.com/charts/Biascircuits.htm

There are differences, but I think that page completely helped a LOT!  It should be R36 that will have to go down in value to allow for less current through the tubes, or more negative voltage.  Thank you!

I kept thinking it's the one closest to ground connected to the bias pot, but was going thrown off in this Marshall, because the wiper is tied to the third leg of the pot and to a 15K that get's tied to -15V.  Now I'm thinking that 15K is simply just a pull-down resistor, so -15V is the highest this particular bias circuit can go, voltage wise.  Correct me if I'm wrong.  Like I said, my head is incredibly foggy right now and stress has really been getting to me, to a point that I just have not been able to think straight.

[sluckey]

It should be R36 that will have to go down in value to allow for less current through the tubes, or more negative voltage.

That's wrong. You need to INCREASE R36 to allow less current through the tubes. Alternately, you can DECREASE R31 to allow less current through the tubes.

[JZRepair]

It should be R36 that will have to go down in value to allow for less current through the tubes, or more negative voltage.

That's wrong. You need to INCREASE R36 to allow less current through the tubes. Alternately, you can DECREASE R31 to allow less current through the tubes.

Thanks sluckey.  I'm sure I would have realized that quickly by adjusting it one way and seeing the current rise rather than drop.  I've got the pot maxed out full CW right now and I'm still a bit high.  Originally I tried to pay attention to the recommended way of checking the bias with the sense resistor using CON5, but the bias probes proved these to be incredibly inaccurate.  So, unless there's a tube problem (i just returned another quad set that came in the same shipment because the bias current was fluctuating like crazy), so I'm not sure about these tubes still.  For being apex burned in and matched, there's some differences, but, moving the tubes around a bit, seemed to fix the issue with that and get them closer to being the same current on all 4 tubes.

Can you explain to me the purpose of R31?  I've never seen a configuration like this before.

[sluckey]

Can you explain to me the purpose of R31?  I've never seen a configuration like this before.

Looks just like all the other Marshall bias circuits. R31, the bias pot, and R36 are connected in series between the negative voltage source and ground. This forms a three resistor voltage divider. Change any of those resistors and you change the voltage sent to the grids.

[JZRepair]

Looks just like all the other Marshall bias circuits. R31, the bias pot, and R36 are connected in series between the negative voltage source and ground. This forms a three resistor voltage divider. Change any of those resistors and you change the voltage sent to the grids.

Thanks, that actually brings up something else I should check.  I have not checked the 15V rails yet and should verify that ZD1/ZD2 are properly conducting. 

Like I said, I've not seen this configuration on any of the other Marshall's I've worked on.  With my fogged mind, none of it was making any sense and I wasn't making out the whole picture.  I'll give another crack at it before I just start swapping things out without checking other things first.  Thanks again, sluckey.

[SILVERGUN]

JZ, with all due respect:
I have a hard time believing that Marshall would put out a modern product without an acceptable bias range adjustment.
You haven't told us what bias measurement you are seeing that seems high.
How are you measuring and what voltages and currents are you seeing?

[JZRepair]

JZ, with all due respect:
I have a hard time believing that Marshall would put out a modern product without an acceptable bias range adjustment.
You haven't told us what bias measurement you are seeing that seems high.
How are you measuring and what voltages and currents are you seeing?

No disrespect taken.  If you go buy their biasing instructions with the current sense resistor for this model (10 Ohm Resistor), 1.3V across that resistor will create a bias current of 0.0325A per tube.  However, using my bias probes, and having to rearrange the tubes a bit to get a closer matched set, I am reading 390V at the plates and V5=19.5mA, V6=21.1mA, V7=20.9mA, V8=19.4mA.  And yes, the current is lower, but dissipation wise, they have it about 55% where I'm about 65%-68% currently.  I'd like to get that closer to 55%-60% to prolong the life of the tubes.  I got the tubes from CE Distro and they are rated on the High end for current, so I expected them to run a little hotter.

[sluckey]

Using 390V and 21.1mA (your highest) gives 8.2W which is 68.5%. I'd call that perfect! Those tubes will last much longer than the typical cathode biased quad such as found in the Vox AC-30.

Is your bias pot set to one extreme to achieve these current readings? If so, I suggest setting the bias pot to half rotation and changing R36 or R31 to get your desired bias point. Then you can adjust cooler or hotter than your desired bias point.

[SILVERGUN]

JZ, with all due respect:
I have a hard time believing that Marshall would put out a modern product without an acceptable bias range adjustment.
You haven't told us what bias measurement you are seeing that seems high.
How are you measuring and what voltages and currents are you seeing?

No disrespect taken.  If you go buy their biasing instructions with the current sense resistor for this model (10 Ohm Resistor), 1.3V across that resistor will create a bias current of 0.0325A per tube.  However, using my bias probes, and having to rearrange the tubes a bit to get a closer matched set, I am reading 390V at the plates and V5=19.5mA, V6=21.1mA, V7=20.9mA, V8=19.4mA.  And yes, the current is lower, but dissipation wise, they have it about 55% where I'm about 65%-68% currently.  I'd like to get that closer to 55%-60% to prolong the life of the tubes.  I got the tubes from CE Distro and they are rated on the High end for current, so I expected them to run a little hotter.

I'm not sure I'm understanding this "If you go buy their biasing instructions with the current sense resistor for this model (10 Ohm Resistor), 1.3V across that resistor"
With 80mA total current (which you have) I'd expect to see .8V across that resistor.
What voltage are YOU measuring across that 10ohm resistor?

[JZRepair]

Using 390V and 21.1mA (your highest) gives 8.2W which is 68.5%. I'd call that perfect! Those tubes will last much longer than the typical cathode biased quad such as found in the Vox AC-30.

Is your bias pot set to one extreme to achieve these current readings? If so, I suggest setting the bias pot to half rotation and changing R36 or R31 to get your desired bias point. Then you can adjust cooler or hotter than your desired bias point.

That's exactly what I was planning on.  I'm maxed to the extreme CW rotation which means I'm getting the lowest current I can at this point.  I wanted to adjust R36 so I can get closer to my desired and have swing on both ends of the bias pot.

[JZRepair]

I'm not sure I'm understanding this "If you go buy their biasing instructions with the current sense resistor for this model (10 Ohm Resistor), 1.3V across that resistor"
With 80mA total current (which you have) I'd expect to see .8V across that resistor.
What voltage are YOU measuring across that 10ohm resistor?

There is a separate document that details their biasing instructions using CON5 to tap into the sense resistor.  They state the 1.3V.  However, you are correct, I'm only getting 0.8V across that resistor.  I've measured it and it's pretty much dead nuts at 10 Ohms.  I figured it would be much better to go off of what I'm actually getting across the tubes.  Sorry that I didn't clarify that more initially.