Fender 5F8A Bias

[ViperDoc]

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

Roger that, so not of any use with power tubes in? Always wondered that.

[martin manning]

I would say the limiter still has some value, limiting the max current and the potential damage. A rule of thumb might be to select a bulb with wattage equal to the amp's full power output to have some consistency in the observed brightness. I've settled on two 60W bulbs, switchable for a low (500 mA) and a high (1A) limit. Incandescents are getting scarce, but 40W oven bulbs are still easily found. Two in parallel for 80W/667 mA would be a good option.

Re corner frequency, that is another name for cut-off frequency for a LP or HP filter, 1/(2πRC), above which there is attenuation of 10 dB/decade. Larger R and/or C will lower the corner frequency and attenuate more of the ripple in the bias voltage. Keeping the current low, say 2 mA, will help too; again, larger resistors. The values suggested above (220, 10k, 22k, 50k pot wired as a variable resistor) will probably work. Note that Marshall amps with small (10u) bias filters typically use 25k bias pots.

[ViperDoc]

I would say the limiter still has some value, limiting the max current and the potential damage. A rule of thumb might be to select a bulb with wattage equal to the amp's full power output to have some consistency in the observed brightness. I've settled on two 60W bulbs, switchable for a low (500 mA) and a high (1A) limit. Incandescents are getting scarce, but 40W oven bulbs are still easily found. Two in parallel for 80W/667 mA would be a good option.

Re corner frequency, that is another name for cut-off frequency for a LP or HP filter, 1/(2πRC), above which there is attenuation of 10 dB/decade. Larger R and/or C will lower the corner frequency and attenuate more of the ripple in the bias voltage. Keeping the current low, say 2 mA, will help too; again, larger resistors. The values suggested above (220, 10k, 22k, 50k pot wired as a variable resistor) will probably work. Note that Marshall amps with small (10u) bias filters typically use 25k bias pots.

Very good! I read a bit of Blencoe on lunch today and found that corner/cut-off discussion, makes a lot of sense. SQUIRREL!

Ok, so just to clarify, you discussed the smaller adjustment fraction as being valuable earlier. Would a fixed 47K into a 25KL adjustable pot be any better/different?

Like this?

ViperDoc 5F8A PS 2.0.pdf

[martin manning]

That’s it. Same as the marked up schematic except for the resistances.

[Stevem]

If you have not let the smoke out yet, you're safe!

[martin manning]

Ok, so just to clarify, you discussed the smaller adjustment fraction as being valuable earlier. Would a fixed 47K into a 25KL adjustable pot be any better/different?

I would try 27k with a 25k pot, which (per simulation) centers at -49V and has 8V range. The 220Ω can be a 1/2W part. Bias circuits look simple, but there is a lot going on. Another consideration is settling time, which is very quick using two small filter caps, rather than one large one, as Fender typically did. Here it's well under 1 sec.

[ViperDoc]

I would try 27k with a 25k pot, which (per simulation) centers at -49V and has 8V range. The 220Ω can be a 1/2W part. Bias circuits look simple, but there is a lot going on. Another consideration is settling time, which is very quick using two small filter caps, rather than one large one, as Fender typically did. Here it's well under 1 sec.

Thanks for that, I'll give that a go.

So cut-off voltage is V(s-k)/mu(triode) which for the 5881 at say 395 V(s-k) is 395/8 = 49. Does that mean the above bias sweep will be cut off on half the pot?

[martin manning]

Depends on where your voltages fall out with all tubes in and biased where you want them. I've only seen one measured voltage quoted so far, 480 unloaded. You'll probably want to do some tweaking in the end.

[ViperDoc]

I put the following in on the bias feed:

BIAS TAP--> 200R-->1N4007-->10K-->22K-->25KL POT-->GROUND

Full up, no limiter and power tubes in, I get:

WALL AC = 121.7 VAC
HTS = 343 VAC
B+ = 430 VDC
B+2 = 429 VDC
BIAS VOLTAGE RANGE = -41 TO -51 VDC

I can knock the B+ down to around 400 VDC with a Brown Box to do testing, but I may need to knock that B+ down a bit with perhaps some zeners, yes?

[martin manning]

What do you have for power tube current? Can you dial in a plate dissipation around 60% of maximum (0.6 x 26W)? If the plate voltage stays at 430, that would be (0.6 x 26)/430 = 36mA.

[ViperDoc]

What do you have for power tube current? Can you dial in a plate dissipation around 60% of maximum (0.6 x 26W)? If the plate voltage stays at 430, that would be (0.6 x 26)/430 = 36mA.

Indeed I can, no problem! All four 5881 tubes I have biased at 37 mA cathode current initially. I've been playing around with the attenuated input AC voltage to see how things shake out when I get just below 400 VDC on the filter cap node. I was testing it slightly below that to start out and the preamp voltages all come in about 5%+ higher, which I'm OK with. It's burning in right now. I did crank it up and play with a Nash HSS stratocaster and HOLY SH!ZB4LL5 does this thing ROCK!!!! WOW!!!!

And it's DEAD QUIET when I'm not playing. Another win for xtian's headphone trick.

EDIT: I have my attention on the 5881 grid voltages, they average about -38 VDC. Looking at the Tung Sol 5881 datasheet, the Class AB1 operating grid voltage is listed at -22.5 VDC for a 360 VDC plate voltage. My PPIMV resistance on each half of the dual pot comes out to about 210K with the safety resistors in place. Am I correct that if I increase the PPIMV (PT grid leak) resistance that the negative grid voltages will reduce and, therefore, allow more output power? I notice on the datasheet that it seems as the negative grid voltage goes up, the power output decreases.

The 5F8A schematic lists the PT plates at 397 VDC, which is roughly 10% higher than the datasheet value. So at 397 VDC on the plates, the operating grid voltage with datasheet operating character would be 1.1(-22.5 VDC) = -24.75 or about -25V. Is it as simple as (38/25)210K = ~ 319K PPIMV pot resistance to enable the datasheet operating output? So if I switched to a dual 500K pot with a 1M safety resistor LarMar2 PPIMV arrangement, that would give me a nominal 330K master volume pot resistance. Is this right?

[pdf64]

… Looking at the Tung Sol 5881 datasheet, the Class AB1 operating grid voltage is listed at -22.5 VDC for a 360 VDC plate voltage. ….

You may not be taking account of the g2 voltage there being 270V?
https://tubedata.altanatubes.com.br/she ... 5/5881.pdf

[martin manning]

HOLY SH!ZB4LL5 does this thing ROCK!!!! WOW!!!!

What you have there is a 5F6-A Bassman with two extra output tubes, so, yeah. ;^)

So if I switched to a dual 500K pot with a 1M safety resistor LarMar2 PPIMV arrangement, that would give me a nominal 330K master volume pot resistance. Is this right?

The grid voltage will be the same for the chosen DC bias point. Larger grid leak resistance won't change that, but you will get a bit more drive, and a shift in the frequency response out of the PI (a bit more bass, and a bit more bass at low volume settings). The higher grid leak resistance will also increase the risk of the power tubes red-plating (at high volume settings).

[ViperDoc]

… Looking at the Tung Sol 5881 datasheet, the Class AB1 operating grid voltage is listed at -22.5 VDC for a 360 VDC plate voltage. ….

You may not be taking account of the g2 voltage there being 270V?
https://tubedata.altanatubes.com.br/she ... 5/5881.pdf

I noticed that, but wasn't sure of the difference. That's significantly lower than the 5F8A schematic values, and all I've heard discussed of screen resistor values (EDIT: and therefore lower screen voltages) seems to involve distortion onset character. I've got 470Rs on there right now. See voltages below.

[ViperDoc]

After a moderate burn in, here are where my voltages are at present. This amp is loud and quiet when desired, sounds really good except for perhaps a bit too much negative feedback. I may play with a 250KL pot in the extension jack position, since I'm never going to use it.

Here you go:

Wall AC = 117 (dropped with Brown Box)
HTS = 330-0-330 VAC
HEATER = 3.03-0-3.03 VAC
All VDC readings below:
B+ = 398
B+2 = 397
B+3 = 357
B+4 = 299

V1: pin 1 = 153
pin 3 = 2.3
pin 6 = 156
pin 8 = 2.3

V2: pin 1 = 183
pin 3 = 0.9
pin 6 = 299
pin 8 = 183

V3: pin 1 = 253
pin 2 = 23.5
pin 3 = 38
pin 6 = 231
pin 7 = 24.4
pin 8 = 38

V4-7: pin 3 = 396
pin 4 = 395
pin 5 = -38
pin 8 = 37 mA