Peavey 5150 Combo Adjustable Bias

[dorrisant]

Schematic...

[roberto]

I'd remove C65 and connect the bias to the winding dedicated to B+ through a large resistor (or even capacito, but smaller, to have an hogher impedance at 50/60 Hz) in the 220k range.

[martin manning]

A resistor won’t work with a FWB rectifier; it would have to be capacitively coupled.

[roberto]

Right, connect the cap to the PT winding that goes to give you B+.
Just use a smaller value capacitor to increase its impedance at 50/60 Hz.
This way you'll have plenty of negative voltage on the bias.

[martin manning]

The coupling cap should be class XY, and I think the other caps at 220u would then be much too large, making the bias voltage very slow ito establish itself. I would reduce them to 10u, and lose the doubler (see JCM900). With all that, it might be easier to just move the doubler grounds to the -25V supply.

[dorrisant]

Ok, I tried Merlin's bias supply (Power Supplies For Valve Amplifiers pp 178). This did not produce any voltage. There was AC at the bias voltage diode but no negative DC on the other side of it. I Have built this circuit many times and see no errors.

I didn't realize I would have this much trouble with this project, now it seems to be draining the life from me.

I had noticed that in the schematic I posted above that there are three different grounds shown. I've never seen the chassis, signal and power supply grounds separated by 47Ω resistors before. Regardless, I tied the ground of my bias supply to the junction of C59 and R104. No dice. Moved the tap to the chassis... Again, no dice. Am I Missing something here? Any reason that Merlin's supply wouldn't work in this instance? The pic below is pasted over the original in the book. It was included in the errata. I just drew it in.

I know that I'm doing something other than what was suggested above, but have you seen these things? They are ridiculous to work on. I was trying to save a lot of time and effort by installing a small bias supply daughterboard. That way, I could tweak the voltage limiting resistor while the rest of the circuit boards were installed. As it is... you basically have to take everything out of the chassis, change a few parts and reinstall everything. It really has me questioning the designers... Were they stupid, or is it just a sick joke? Sure seems like one of the two.

I really need a simple spice program that could simulate to get the stock circuit into the range I'm looking for. That would also require a learning curve. I need a break from those right now though.

[dorrisant]

... I think I'm gonna pull off of the HT and build a supply like one seen in Merlin's book, swap the pot for a 10k and call it good. I know how that one works.

Whatever idiot!!

[sluckey]

Merlin's circuit works just fine. But you have to supply the raw negative voltage. Your D2 and R4 cannot do that. When you try to pick off voltage in an amp that uses a full wave bridge, you must use a capacitor/diode/resistor.

Look at page 6 of this pdf for an explanation of what you must do when you have a bridge rectifier involved. You must have C1/R1 between the bridge and the bias diode.

http://sluckeyamps.com/misc/Amp_Scrapbook.pdf

[dorrisant]

Merlin's circuit works just fine. But you have to supply the raw negative voltage. Your D2 and R4 cannot do that. When you try to pick off voltage in an amp that uses a full wave bridge, you must use a capacitor/diode/resistor.

That is it! That's what I was missing. Much appreciated!!

So, if I'm not mistaken, this layout should work. Right?

Is there any reason to ever need to change the values of the 56k and the .047µF?

[sluckey]

Is there any reason to ever need to change the values of the 56k and the .047µF?

Those values should be fine. You should not use a standard coupling cap though. The cap needs to be rated "Class X" or "Class Y". You may need to juggle the values of some other resistors to get the desired bias range though.

Why such a complicated bias supply? I really like to keep that circuit simple and bullet proof. A lot of expensive stuff depends on the reliability of the bias circuit. Look at the JCM900 bias circuit on page 5 of that same pdf. That's a simple reliable circuit that has served Marshall well for a long time.

[dorrisant]

Thanks for the direction. One more question... Are there rules for the voltage rating, etc.? Am I looking to cover the AC from the PT secondary?