Scratch built Bassman AB165 bias/hum questions

[1rebmem]

Built a Bassman to the AB165 schematic with very minor deviations like dual bias adjust pots. But the overall circuit is
per the schematic (below).

Problem/question:  When I have a 12AT7 for the PI tube and I match the 6L6 (TAD 6L6WGC-STR) cathode currents, I get 60 Hz hum. If I mismatch by 9 - 10 mA's then it's quiet .  This holds true for a NOS RCA 12AT7 and 2 EH 12AT7's I bought with matched triodes.

When I put a 12AX7 in the PI slot the hum goes to zero with tube currents within 1 mA of each other.  What is going on??
 
I have 1 ohm precision resistors cathode to ground to read 6L6 currents. I validated currents by shunting across OT windings and they are about 6 mA lower than across one ohm resistors.



 

[eleventeen]

Curious (to me/us) and irritating (to you) problem. As a wild guess and idea toss-out, it may be that having the dual bias pots imposes a far more stringent requirement upon the cleanness of the DC bias supply. Perhaps you should supply just a snippet of a schematic that shows exactly how you implemented the dual bias pots. You have the e-cap filtering both legs? Got a scope you can look at it/them on?


I also notice that the schematic you posted is more in the nature of a balance type versus a vanilla AB763 type....and....it's unbalanced! I must confess I never noticed that before.



Or, tack in a much larger e-cap for the bias filter cap, in parallel with the existing one. Got a 100ufd or 150 ufd sitting around of appropriate voltage rating?



And do you have a 6.3 VAC center tap or the dual 100-ohm resistors? When I have built amps and left those out, before I threw those resistors in the amp hummed huge. I doubt this would be your current issue, just asking.

[1rebmem]

This is the dual bias circuit I used, with the following changes:
1) 180K resistor replaced with 470 because I have a 50 volt bias winding
2) 10uF caps replaced with 15 uf
3) Pots are 50K instead of 25K

The 50K pot wipers feed the 220 ohm bias resistors shown on the AB165 schematic

[1rebmem]

eleventeen...thanks for the reply.

My heater circuit is elevated. I tapped off the B+ and dropped it with resistor and cap, I have 70 volts dc elevation on my 6.3v CT. 

I do have a scope and can look at the Bias ripple. I can also parallel a bigger cap across one of my 15uf caps.....tomorrow.

[eleventeen]

1: Measure one bias feed, tweak the other. I think one is going to talk to the other one. I would imagine you would not want that. At minimum, I believe you would need separate 33K "ground side" resistors. The two bias supplies will still talk to each other a bit, but less.


2: It's possible you may need more filtering/smoothing. It is true that bias pulls minimal current, but you only have a half-wave rectifier (which is normal) and 30 ufd (2x 15 ufd in parallel) is barely there, IMHO.


I would tack a bigger cap across the second cap.

[1rebmem]

I will parallel a big cap in tomorrow.

I have used this dual bias circuit on probably 4 other amps and it always worked good. Very little interaction between pots/bias volts.
All I am doing is reading mV across my 1 ohm cathode resistors and adjusting my bias pots for 50mV. I don't even look at the negative bias voltages.

The thing that confuses me is that a 12AX7 PI tube will allow the power tubes to bias up normal/balanced but the 12AT7's won't.

[1rebmem]

I paralleled in a 64uf cap.....same result.

[SILVERGUN]

The thing that confuses me is that a 12AX7 PI tube will allow the power tubes to bias up normal/balanced but the 12AT7's won't.

I would have to try removing the 2 (220k) feedback resistors from 6L6 plate to 12AT7 plate just to see if the problem goes away.
 
12AT7 deals in higher current than AX7 so that could be contributing to your mystery.
 
 
 

[1rebmem]

The thing that confuses me is that a 12AX7 PI tube will allow the power tubes to bias up normal/balanced but the 12AT7's won't.

I would have to try removing the 2 (220k) feedback resistors from 6L6 plate to 12AT7 plate just to see if the problem goes away.
 
12AT7 deals in higher current than AX7 so that could be contributing to your mystery.

I removed those plate to plate 220K resistors and problem solved!  .............. Thanks Silvergun !!
Without the 220K's installed, I can mismatch the power tubes by about 20 mA before any hum is introduced.
 
WITH the 220K's and the 12AX7 installed, hum was eliminated with the power tube currents matched but a 1mA or more mismatch would start introducing hum.  WITH the 220K's and 12AT7 tube, I had no hum with  a 9mA mismatch but hum would increase when I increased the mismatch  to about 10 or more mA's.  So with either tube in the PI slot AND the 220K's installed, there is a narrow bias band where hum is eliminated.

What is the purpose of these 220k's?
I am building this amp for bass use primarily and I read this circuit is better for bass than the earlier Blackface circuits.

The original schematic has a hum balance pot and note #4 says to adjust for minimum hum. Based on my findings this AB165 circuit may have required quite a power tube idle current mismatch to eliminate hum, depending on the 6L6's installed of course. 

[1rebmem]

Reading up some on those 220K resistors........
They act as local feedback on the power tubes?
This would give a little more clean headroom in theory?

[eleventeen]

I never noticed those 220K resistors. Strange how you can look at a schematic 500 times and not see something. Those are...quite odd.