Any thoughts on how to "soften" attack on a Fender Bassman ?

[pdf64]

Can you verify that the feedback loop on the power amp is negative? eg does the output increase or decrease if the loop is opened?
Positive feedback doesn't necessarily result in free running oscillation.
Pete

[Chevy]

Can you verify that the feedback loop on the power amp is negative? eg does the output increase or decrease if the loop is opened?
Positive feedback doesn't necessarily result in free running oscillation.
Pete

Yep, tried to reverse the OT wiring, and the amp went apeshit... so it's wired correctly. Can run at full power right across the frequency band, and no oscillation.

[pdf64]

Have you got control grid stoppers on the power tubes (5F6A doesn't have them)?
The inductive load from a real speaker can result in oscillation, which isn't there with a resistive load.
Pete

[Chevy]

Have you got control grid stoppers on the power tubes (5F6A doesn't have them)?
The inductive load from a real speaker can result in oscillation, which isn't there with a resistive load.
Pete

Yes, they are there... 1.5K if I remember correctly.

[olddawg]

Old trick: 2.2uf bipolar cap and 8 ohm (or whatever the load is) resistor in series with it from across the speaker terminal.

[Chuck H]

If this phenomenon is most obvious at higher volumes it's possible a beam blocker could help.

After that, I have a circuit on one of my designs that amounts to a "bleeder" sort of circuit, but it doesn't seem to affect the tone as severely as others I've tried. It's just a small pf cap from grid to plate on the input stage. I'm told that the actual effectiveness is plate voltage dependent. So when the plate voltage is highest the effect is greatest. This means that the spikey treble of the attack is diminished more than the treble of the decaying note. I'm using a 4.7pf on my personal proto of this amp and that's the value I'd suggest starting with. If it bleeds off too much chime (I don't think it will) go to a 3.3pf. If you want more effect try a 6pf. Any higher than that and it becomes too audible IMHO.

[frus]

Yes, presence control is there; 27K feedback R, 4.7K pot, 0.1uF cap.

you've put 27k feedback resistor, but you probably don't have 2 ohm secondary OT, more like 4 ohm
so to replicate 5F6a's behaviour you'd need to increase it to 39k

it probably won't be a dramatic change though

[cbarrow7625]

Chuck's response reminds me of a Silver face Fender Twin Reverb I modified for a friend a long time ago. He wanted a high gain channel added. Not a great match with that amp. I have him a 4 stage gain monster that was pretty well behaved in the end. To tame the high frequencies without really affecting the overall tone I used quite a few pf caps from grid - plate, plate - cathode and grid -cathode. They were all kept in the sub 100pf range but we're essential to kill oscillation and "squeallies". I think some stages may have had more than one of these caps between different pin pairs. In the end, you really couldn't tell that there were caps in there at all. Limiting the folk off to very high frequencies allowed the full harmonics to come through fairly faithfully. The trick was to put these caps on all of the stages so that they cascaded to form a steeper overall low pass at a higher than normal frequency.

Sounds like a version of that could be helpful here. Each pin pairing for the cap has a little bit different effect on the sound and dynamic so you may have to play around a bit. My gut tells me that for the issue you describe in this "clean" amp you may have to go higher than 100pf in some location, although I doubt you will need/want to venture above 250pf. I liked the sound of these local feedback caps around the tubes much better than caps to ground or around the plate resistors, although both may be useful, or potentially better (who knows) for this amp.

I am not a fan of the sound of using any filtering on the first stage but any other subsequent stage, including PI and power tubes, are fair game.

[Chevy]

If this phenomenon is most obvious at higher volumes it's possible a beam blocker could help.

After that, I have a circuit on one of my designs that amounts to a "bleeder" sort of circuit, but it doesn't seem to affect the tone as severely as others I've tried. It's just a small pf cap from grid to plate on the input stage. I'm told that the actual effectiveness is plate voltage dependent. So when the plate voltage is highest the effect is greatest. This means that the spikey treble of the attack is diminished more than the treble of the decaying note. I'm using a 4.7pf on my personal proto of this amp and that's the value I'd suggest starting with. If it bleeds off too much chime (I don't think it will) go to a 3.3pf. If you want more effect try a 6pf. Any higher than that and it becomes too audible IMHO.

Ahhhh..... this is good stuff... you guys are great !! Yes, this is most definitely worse at higher volumes, and the beam blocker idea did occur to me. Haven't tried it, (as I thought it was just an electronics issue), but apparently even putting some strategically located patches of duct tape or something on the grille cloth could do the trick. However... if used for recording, the blocker may mess things up for the mic.

Chuck, this bleeder sounds like a bit of "frequency selective compression", which may be just what the doctor ordered. Will have to give it a try ! Don't know if I have any caps that small, may have to order and test later. I recall playing with *umble mods on my previous amps, and one schematic I saw used a 22M resistor in series with a .05uF cap, from grid to plate of the second stage, right after the tone stack. That was something else I was going to mess with; seems if you use an appropriately small enough cap, you can limit the compression to the top end. Don't know how much chime this will rob yet, but if it's what the famous D-man used in his amps, and how great Robben Ford sounds...well, it's certainly good enough for me. Back to the test bench.

[Chevy]

Chuck's response reminds me of a Silver face Fender Twin Reverb I modified for a friend a long time ago. He wanted a high gain channel added. Not a great match with that amp. I have him a 4 stage gain monster that was pretty well behaved in the end. To tame the high frequencies without really affecting the overall tone I used quite a few pf caps from grid - plate, plate - cathode and grid -cathode. They were all kept in the sub 100pf range but we're essential to kill oscillation and "squeallies". I think some stages may have had more than one of these caps between different pin pairs. In the end, you really couldn't tell that there were caps in there at all. Limiting the folk off to very high frequencies allowed the full harmonics to come through fairly faithfully. The trick was to put these caps on all of the stages so that they cascaded to form a steeper overall low pass at a higher than normal frequency.

Sounds like a version of that could be helpful here. Each pin pairing for the cap has a little bit different effect on the sound and dynamic so you may have to play around a bit. My gut tells me that for the issue you describe in this "clean" amp you may have to go higher than 100pf in some location, although I doubt you will need/want to venture above 250pf. I liked the sound of these local feedback caps around the tubes much better than caps to ground or around the plate resistors, although both may be useful, or potentially better (who knows) for this amp.

I am not a fan of the sound of using any filtering on the first stage but any other subsequent stage, including PI and power tubes, are fair game.

Ah... another good one ! On the 5F6A design, it's easiest to start at the mixer stage, rather than each input stage, so you would not have to mod each input stage separately.... but who knows, only bench testing will prove this out one way or another. Thanks for the tip.

[Chevy]

Old trick: 2.2uf bipolar cap and 8 ohm (or whatever the load is) resister in series with it from across the speaker terminal.

This one I dont' quite get as far as load impedance goes... you're putting this R-C circuit in parallel with the speaker load, right?... if you put a 4 ohm R on that circuit, with a 4 ohm speaker, at higher frequencies you will present a 2 ohm load to the amp. At high volumes you could be jeopardizing output tubes and perhaps more, no..... ?

[eschertron]

This one I dont' quite get as far as load impedance goes... you're putting this R-C circuit in parallel with the speaker load, right?... if you put a 4 ohm R on that circuit, with a 4 ohm speaker, at higher frequencies you will present a 2 ohm load to the amp. At high volumes you could be jeopardizing output tubes and perhaps more, no..... ?

The half-power point for the 2.2u plus 8R network is just under 10kHz. How much energy is your OT capable of supplying above that freq? The audible effect will be subtle; the power draw negligible.

edit: Also, the Z of the network increases as frequency drops. At 500Hz, for example, the cap and resistor are > 150R. Not really affecting OT load or speaker response. In theory.

[Chris - CMW amps]

some things to try:
- warmer sounding preamp-tubes
- using a smaller (let's say 47p) ceramic bright-cap
- lower the treble cap somewhat to (let's say) 150p (also ceramic)
- lowering the preamp-voltages somewhat by enlarging the dropping resistors in the powersupply
- adding mo' neg. feedback in the output-section
- warmer sounding speakers

good luck!

[Chevy]

The half-power point for the 2.2u plus 8R network is just under 10kHz. How much energy is your OT capable of supplying above that freq? The audible effect will be subtle; the power draw negligible.

edit: Also, the Z of the network increases as frequency drops. At 500Hz, for example, the cap and resistor are > 150R. Not really affecting OT load or speaker response. In theory.

Oh, right... this sounds like another one to try... thanks!

[olddawg]

The half-power point for the 2.2u plus 8R network is just under 10kHz. How much energy is your OT capable of supplying above that freq? The audible effect will be subtle; the power draw negligible.

edit: Also, the Z of the network increases as frequency drops. At 500Hz, for example, the cap and resistor are > 150R. Not really affecting OT load or speaker response. In theory.

Yep. Just dulls the ice pick a bit. A high pass filter to ground basically. If it was a a cheap hi if speaker you would have a tweeter where the resistor is. It's a cheap, quick, and dirty fix, that won't damage anything.