New 5F8A HUMS!

[MWJB]

We're not home & dry yet. We've just established that something is driving your negative feedback positive.

It is most likely something to do with your layout/grounding.

You have a glass fibre board so you don't need to leave slack in the wires running to your tube sockets, therefore shorten your grid wires, you might want to shieldthe wire running from the treble pot to the PI input cap, the grid wires from the PI to the 6L6 grids too, don't thread them behind the board, run directly to the tube grids, avoiding heater wires.

Instead of grounding everything as it falls on the buss wire, run an individual wire from each ground point on the eyelet board to the ground lug of the#1 Normal input jack.

There is a wire running from the ground lug of your ext. speaker jack plug, where does this go?

[ellum68]

I'm gonna start by isolating the main boards grounds in a moment. The ground off the output goes a short a distance to the power tube next to it where the cathode pin gets grounded. I am also including a diagram of my output. It's a multi-tap unit. With nothing pluged in, the jack makes contact with the 2-ohm tap and thus the jack on the right would be for two ohm use. If you plug into the jack on the left then it switches over and hits the 4-ohm tap. If you have both jacks filled, then it should still be a 4-ohm load.

[MWJB]

At what point does the OT common secondary connect to ground, the way you have drawn it the ground lugs on the jacks connect to each other & the powertube cathode ground (not really necessary), but not to the OT common?

Did you adjust (reduce the size of the NFB resistor to restore loop ratio)? By running a 56K NFB into the 2ohm tap you will be reducing the voltage fed back to the PI, you would normally increase it for a lower ohms output tap.

[ellum68]

At what point does the OT common secondary connect to ground, the way you have drawn it the ground lugs on the jacks connect to each other & the powertube cathode ground (not really necessary), but not to the OT common?

Did you adjust (reduce the size of the NFB resistor to restore loop ratio)? By running a 56K NFB into the 2ohm tap you will be reducing the voltage fed back to the PI, you would normally increase it for a lower ohms output tap.

OT common connetcts to ground at the output jack. Wasn't aware of the resistor change. Think its better to ditch my little "changes" and go back to dead stock.

[ellum68]

Got 'er going. Swapped the OT leads on the output tubes and she shut right up. Gotta figure this bias game put now (first time). Running a 51k bias resistor hooked to a 10k pot and am only getting 32ma. I think I'll swap it for a 47k and see what I got. Oh BTW, it makes music too.

[MWJB]

32mA per tube will be fine. Try a 33-39K resistor and a 50KL pot (select a fixed resistor that doesn't let the tubes idle at over 40mA) At the moment your bias sweep is very limited, probably -52 to -53v at pin 5? Your voltage divider only allows adjustment between 80% & 77% of the voltage after the bias diode.

[ellum68]

10K is all Mojo had. Little did I know that how much room I would have to play with. I can't believe how overblown people have made setting a bias. Quite simple and very useful piece of knowledge. All and all I had a great shake run down a moment ago. I need to look at something on one input jack and had a little microphonic tube sound on the normal channel when I turned it up to 10 and then it left after hard playing. Sweet Jesus is this thing loud. I'm only running a 12AY7 in the preamp too! I'll let you know in the comming days how things end up. Thanks so much!

[MWJB]

If you want to keep the 10K bias pot use it, plus a 10K resistor in series, to replace the 15K vertical dropping resistor after the bias diode, run a stock value 56K to ground, in place of your 51K & 10K pot. The bias voltage divider consists of a stock 15K dropper & 56K load resistor, you have the pot in the "load" side right now, but it can be part of the "dropper" instead.

You could just use a reguar 50KL amp pot to replace your 10K, but it won't have the inboard screwdriver slot, you'll need to reach around the outside of the chassis...easier to just put the 10K in a different part of the bias circuit.

[jmaf]

10K is all Mojo had. Little did I know that how much room I would have to play with. I can't believe how overblown people have made setting a bias. Quite simple and very useful piece of knowledge. All and all I had a great shake run down a moment ago. I need to look at something on one input jack and had a little microphonic tube sound on the normal channel when I turned it up to 10 and then it left after hard playing. Sweet Jesus is this thing loud. I'm only running a 12AY7 in the preamp too! I'll let you know in the comming days how things end up. Thanks so much!

10K is fine if you hook it up as a voltage divider between bias and ground, it'll then allow any voltage from -B to zero(I would add a 1k to 4k7 resistor before ground to avoid zero volts bias), thus allowing almost any level of current - including ruining your tubes, but then again with greater control comes greater responsibility for the technician.

If you have it in series with larger resistances it'll allow only a narrow band of voltages to be dialed in, I think that's what MWJB meant.

[MWJB]

Sorry Jmaf, I meant that if ellum68 wanted to use his 10K pot (with in board screwdiver slot, like a BF/SF pot) that he could move the adjust pot from the grounded portion of the circuit & us it (plus a 5-10K fixed resistor in series) to replace the stock 15K bias dropping resistor. Still using the stock 56K value for the bias circuit load resistor to ground.

10K (or even 14.7K) in a stock 5F8A bias circuit will result in...
-70vdc*14.7K/(14.7K+15K) = -34.6vdc at the 6L6 grids...& red plates, even at the coldest setting.

[jmaf]

Sorry Jmaf, I meant that if ellum68 wanted to use his 10K pot (with in board screwdiver slot, like a BF/SF pot) that he could move the adjust pot from the grounded portion of the circuit & us it (plus a 5-10K fixed resistor in series) to replace the stock 15K bias dropping resistor. Still using the stock 56K value for the bias circuit load resistor to ground.

10K (or even 14.7K) in a stock 5F8A bias circuit will result in...
-70vdc*14.7K/(14.7K+15K) = -34.6vdc at the 6L6 grids...& red plates, even at the coldest setting.

Got ya.