Fender 5F8A Bias

[ViperDoc]

I'm warming up a new 5F8A build on my bulb limiter and, in true form, I've got blinding light with the power tubes in.

Here's a power supply schematic of my build. The circuit voltages are target voltages on the original 5F8A schematic, not my build.

VIPERDOC 5F8A POWER SUPPLY.pdf

My PT is a Mercury Magnetics FTHPT-P with a measured 47 VAC on the bias tap off the limiter and - 39 VDC to -42 VDC on the distal end of the 15K dropping resistor with no power tubes in. My bias supply is stock with the exception of a 47K + 10K-pot for bias adjustment (stock bias resistor is 56K) and at full resistance it measures 57.5K. I tacked a 2K2 dropping resistor between the bias tap and the bias diode to enable a -41 VDC measurement as per the schematic, but that's without power tubes in; perhaps that is unnecessary, the bias tap voltage will reduce with a full load? ~10%? I installed a LARMAR2 type PPIMV which measures about 210K across each side with the safety resistors in place (that's about as close as I could get to 220K). It's left full open for now.

When I put the GZ34 and preamp tubes in only, I get a dim limiter bulb and reasonably higher, unloaded voltages everywhere; off the limiter and no power tubes, the A-node voltage is 480 VDC! When I put in the power tubes on the limiter, the bulb is white hot and my bias tap AC voltage is incredibly low, like 21 VAC. My negative bias voltage is also low. I thought of lighting it up off the limiter to check the loaded voltages, but being as power tubes are a hot commodity, I'd love not to burn these to the ground!

I do not detect any problems with the power tube socket wiring. I chose not to tie pins 1 and 8 because 5881 tubes do not have a pin #1 and I used two of them to support the two gridstopper resistors on the outer tubes. I will bypass the 2K2 resistor and check it again, later.

[martin manning]

The bias voltage on the schematic should be considered a suggestion. I would try jumping that 2k2 with the bias trimmer centered to get lower (more negative) bias voltage in range that you need.

[sluckey]

I would replace the 10K pot with a 50K pot. This will give you more negative voltage and also allow a wider range of adjustment.

[Stevem]

If your using a pot Once you get it to where you have a good range of bias voltage I would replace that pot with the same value cermet type trimmer .

[pdf64]

For any schematics that show a nominal bias voltage, it is incorrect the view it as an absolute value. Rather its value must be seen in the context of the voltage also supplied to the screen grid node.
If the magnitude of the latter is x% greater than the schematic, the nominal bias voltage magnitude should be similarly greater.
Of course the nominal bias voltage will only be fully suitable for valves that conform to nominal; many won’t, and will benefit from a tweak.

[martin manning]

I'm guessing you want to keep the 10k pot, as it's probably the standard Fender type with screwdriver slot adjust. I'm also guessing you want to keep the fail-safe wiring (pot wired as a variable resistor), which limits the bias range. You're also challenged a bit by ripple filtering by keeping the low (8uF) filter cap values. If I take a wild guess and assume you need another 10V negative bias, the following circuit would work. In any case you can use these values to get close. The range is only 4V, which gives you a fine adjustment, but you will have to tweak the 220Ω resistor value to get the range centered.

[martin manning]

For any schematics that show a nominal bias voltage, it is incorrect the view it as an absolute value. Rather its value must be seen in the context of the voltage also supplied to the screen grid node...

True, and secondarily the anode voltage, which is basically the same as the screen voltage in most amps. Those voltages depend upon the HT current draw, which in turn depends upon the bias voltage. You don't know where things will settle out until you get the bias set where you want it.

[ViperDoc]

I’m not a purist, I’m open to any and all recommendations to improve upon the original. The 10K trimmer is a small plastic pot on a bias board, easily changeable. If 8 uF bias e-caps are low, what would you recommend?

I’m comfortable with math, but I don’t understand how Ohm’s Law applies to negative voltage. I understand that the more input voltage that is rectified will be a larger negative VDC, but why do I want less resistance on the end of the bias circuit? Relative to the -51V new target?

Always appreciated!

[martin manning]

The lower R next to the pot is so that the 10k is a larger fraction of the total to keep the adjustment range up. The other resistors are smaller to reduce the voltage rise from the peak negative voltage you have to work with (around -60V). If you can swap the trimmer easily, then you can use larger resistors all around and get more ripple reduction with the 8u caps by lowering the corner frequency, which is proportional to 1/RC. For the same -10VDC bias you could go with 220, 10k, 22k, 50k pot. The range will be ~13V. It's a good idea to have some resistance to limit the peak current going into the first cap.

[xtian]

What wattage bulb is in your limiter?

[ViperDoc]

What wattage bulb is in your limiter?

40W

[martin manning]

Well that's pretty low for a 4x 5881 amp. Did you get any voltage or current readings?

[xtian]

Right--I would expect the amp to draw around 80 watts when operating, so yes, a 40w bulb would be full brightness with the amp...attempting...to run properly.

[ViperDoc]

Well, that just makes too much sense. 120W then? The bulb is still alive and well, and the voltages were much lower than expected, so the limiter worked! Perhaps harder than necessary.

Regarding "corner frequency". I'm a deer in the headlights. I'll try and look that up. I'd like to know more about bias supply design relative to "what you all know" vs. vintage schematics and the old data that they possess. Some day. I've read Blencoe's bias supply chapter several times and I'm sure there are more than a few concepts I'm reading at about 20 grit.

But initially, I plan to:

1) Get a higher power light bulb, although incandescent bulbs are all but non-existent now.
2) Rearrange the bias dropping values to target -50VDC on the bias supply.
3) Look at the bias adjustment range and allow for a wider sweep if needed.
4) I see 100W Marshalls with 10 uF bias capacitors. Is 25% larger value anything to look at?
5) Learn what corner frequency is.

[xtian]

There's no point to the bulb limiter at this point. It's function is just to prevent short from FUBARing your amp.