NFB & PI tail resistor 5F6A vs 6G6A questions

[rp]

Hello again, I was reading up around the net and found some posts claiming the wonders of changing the 10K tail resistor (I think that's the correct name) of the PI to 6K8 and changing the Presence Control wiring to a rheostat type with a 4K7 to ground, White Bassman style. Supposedly, this is the actually wiring of the last Tweed Bassmans that most agree are best, but that never made it to schematic.

My amp is an old project, pre-internet, pre-mojotone, pre-turret boards galore, just AES, Triode, and Tom Bremer for parts and the Mouser catalog wasn't an inch thick. So it's all built w/ terminal strips in a narrow bud box, OT is from a Montreal pawnshop Traynor. I avoided wires since I'm an idiot-savant and was told by a helpfull tech long ago that the best way to stay out of trouble is to go PTP and avoid wires as much as possible. Great advice, my amps are dead silent, but makes for amps that's a bear to mod, tweak, repair.

So before I do these changes I'd like to hear some advice. The 10K tail resistor is easy for me to get to but I'd have to trash a nice old 1W Allen Bradley and I've so few left. What will I hear going from 10K to 6K8? Why did Fender choose to do this? If it makes the amp brighter I don't want to know about it. Is it best to change the PI cathode resistor from 470 to 820 too? This one's hard to get to.

Will rewiring the Presence pot to G6G style noticeably change the action/sound, or did fender just do this to cure the DC scratchiness? Is is best to go with the 25K pot too or can I leave the 5k, change the ground and cap to a rheostat, and add the 4K7? I have only one 4K7 and it's from my pulls box, it's damn short - can I leave the whole thing 5F6A non-rheostat style and just add the 4K7 across the pot as it will just span the lugs? I like to run everything to star ground points / bus bar and don't solder to the back of the pots.

I noticed Marshall changed to the 6G6 presence wiring in the 1987 onward, but always kept the 10K / 470 in the PI.

Much thanks

[martin manning]

Comparing the two circuits 6G6A vs. 5F6A I see the following:

The total resistance from cathode to ground is a little lower 11.7k vs. 15.5k. In theory the 5F6A's higher resistance would give a little better balance to the PI output, but reduce its headroom. Parallel a 22k across the existing 10k to get 6.9k and see what happens.

The PI bias resistor difference of 2x will probably make a noticeable change in the tone contribution of the PI, since it sets the bias point of the two triodes. The 470 is hotter, of course, and I'd expect more distortion with that value since the 820 should bias nearer to the center of the plate curves. Perhaps you can lift one end of the existing resistor and tack in an 820 to try it out?

The bottom of the tail resistance is essentially the same at 4.7k vs. 5k. The feedback resistor is much different, though, so the voltage divider on the 6G6A is 56k and 4.7k giving 7.7% of Vout, where it is 27k and 5k on the 5F6A, giving 15.6%, or twice as much. However, the 6G6A has two speakers in parallel for a 4 ohm load, where the 5F6A has four speakers and a 2 ohm load. Since the voltage will be ~1.4x as high at the speaker as on the 6G6A (plus the effect of the 38V or +9% B+ difference), the feedback voltage at the PI is probably about 70% of that on the 5F6A.

The presence pot on the 6G6A is 5x the tail resistor value so that it won't appreciably affect the PI's operation when it is turned to max resistance.

In both cases, when the presence control is turned all the way up, you have a ~5K resistance in parallel with a 0.1uF capacitor to ground. The series resistance is different by ~2x, but this makes little difference in the -3 dB knee frequency of the resulting low-pass filter (3.67 vs. 3.77 kHz), so I don't see much difference there. If the scratchiness is troublesome, switch to the 6G6A circuit with 4.7k fixed and a 25k pot.

MPM

[rp]

The total resistance from cathode to ground is a little lower 11.7k vs. 15.5k. In theory the 5F6A's higher resistance would give a little better balance to the PI output, but reduce its headroom. Parallel a 22k across the existing 10k to get 6.9k and see what happens.

Shame on me, should have thought of that. But was also hoping for - by all means do it, best thing since ever

The bottom of the tail resistance is essentially the same at 4.7k vs. 5k. The feedback resistor is much different, though, so the voltage divider on the 6G6A is 56k and 4.7k giving 7.7% of Vout, where it is 27k and 5k on the 5F6A, giving 15.6%, or twice as much. However, the 6G6A has two speakers in parallel for a 4 ohm load, where the 5F6A has four speakers and a 2 ohm load. Since the voltage will be twice as high at the speaker on the 6G6A (plus the effect of the 38V or +9% B+ difference), the feedback voltage at the PI is probably about the same as on the 5F6A.

Forgot: NFB is Marshall style off 16 Ohm tap

The presence pot on the 6G6A is 5x the tail resistor value so that it won't appreciably affect the PI's operation when it is turned to max resistance. In both cases, when the presence control is turned all the way up, you have a ~5K resistance in parallel with a 0.1uF capacitor to ground. The series resistance is different by ~2x, but this makes little difference in the -3 dB knee frequency of the resulting low-pass filter (3.67 vs. 3.77 kHz), so I don't see much difference there. If the scratchiness is troublesome, switch to the 6G6A circuit with 4.7k fixed and a 25k pot.

Understand, so it's basically just a cure for scratchiness, no big change in the presence sound. But does this still hold if I'm coming off the 16 ohm tap? If so I will leave sleeping dogs lie.

Thanks, Martin

[PCollen]

If you want to have some fun, put a 22K or 25K trim-pot in series with your 10K PI tail resistor, and incrementally increase it from 0 upward. Doing so will incrementally decrease the headroom of your PI and you might find you like some of the resultant tones. I believe that Nik (Ceriatone) incorporated such a mod to the Expression amp he did for Andy Baylor in The Gear Page forum .

see YouTube demo http://www.youtube.com/watch?v=472K5yGVKeQ

You can see the trim-pot on the tail of the PI in the video ( @ 1:06, 2:12 and 2:26) if you watch close.

[martin manning]

Understand, so it's basically just a cure for scratchiness, no big change in the presence sound. But does this still hold if I'm coming off the 16 ohm tap? If so I will leave sleeping dogs lie.

Yes, I believe it's essentially the same either way. Note I fixed a math error above; the 6G6A has a bit less NFB re the 5F6A, and you have much more (like the JTM45).

MPM

[PCollen]

Understand, so it's basically just a cure for scratchiness, no big change in the presence sound. But does this still hold if I'm coming off the 16 ohm tap? If so I will leave sleeping dogs lie.

Yes, I believe it's essentially the same either way. Note I fixed a math error above; the 6G6A has a bit less NFB re the 5F6A, and you have much more (like the JTM45).

MPM

The 'scratchy-ness' is due to the DC current flowing from ground, through the Presence Pot, the PI tail resistor, cathode resistor, and subsequently the tubes. It's this DC current that causes the voltage drop across the presence pot and resulting scratchy-ness. It has nothing to do with the AC feedback path from the OT.

[martin manning]

The question was does the choice of 5F6A vs. 6G6A presence control circuit depend on which tap is being used as the source. Answer, no.