Looking for ideas on how to make NFB vs no NFB infinatly variable

[daz]

Ok, i know how obviously....use a pot who's max resistance yields no audible difference between the sound with NFB in or switched out. the problem is, it takes a 1 meg pot to dial NFB out to the degree where it's not noticable. But then when you want to dial a bit of it in it all happens in the last tiny bit of the pot's throw. I realize there is no way i'm going to find a pot with a taper that will make the throw remotely useable. So aside from a rotary switch with various set resistors in series with the pot, does anyone have any ideas on other ways to do this? I found NFB is often too compressed and dull sounding compared to the amp w/o NFB. But w/o NFB it also gets too bright and aggressive. So i found doing this allows me to really dial in a nice balance between the 2. The only problem is the touchiness of the pot with most of it's throw all in the last 1/10th of it's rotation. unfortunatly i like settings everywhere from almost no NFB to a lot. So i do need the full 1M range, or at least close. (tried 500k but wouldn't let me get quite as far towards the no NFB sound as i wanted)

[defaced]

Cascade two pots in series. One would be your 1M pot, which would act like a "course" control, the second would be a smaller pot that would act like a "fine" control. Basically the 1M pot would get you in the region of where you want to adjust, the smaller pot would let you explore that area with more control.

[daz]

Yeah, i thought about that but not my ideal solution, but definatly on the back burner. If i don't find a better way i can put the 1 meg on the back. I'm also wondering about those multi turn pots, tho then you have no real reference where the sweet spots are.

[Merlinb]

Use a log pot?

Arrange a 1Meg pot as a potentiometer instead of a series (variable) resistor, with a smaller value resistor from wiper to end, so the resistance falls much faster as you turn?

[daz]

Use a log pot?

Arrange a 1Meg pot as a potentiometer instead of a series (variable) resistor, with a smaller value resistor from wiper to end, so the resistance falls much faster as you turn?

I'm trying to visualize how this would affect NFB and what you mean by "with a smaller value resistor from wiper to end". you mean a resistor from in to out of the pot, and if so what value? But i just don't know enough to visualize what that would do. How it would affect the presence, the tone in general. I guess i could try it but doesn't seem right somehow. Then again you know worlds more than I, (who doesn't? )so i suppose i should.

[Gaz]

Use a switch that puts another resistor in series with the low value resistor (where the NFB is not noticeable), so when it's switched it dials in the 'end of the rotation' amount of NFB.

My other suggestion would be getting rid of the NFB and adding a 'hi-cut' pot across the phase inverter like the Vox AC30.

[black_labb]

what about a rotary switch with a number of settings (one being completely off).
whats wrong with a ~100k and a switch? or a pot with a push/pull switch (like they have for guitars with dpdt switches on the back).

[daz]

My other suggestion would be getting rid of the NFB and adding a 'hi-cut' pot across the phase inverter like the Vox AC30.

You know, i actually tried that several times and almost settled on that. I eventually opted to keep NFB because with a presence control along with the treble control i had the ability to change the center freq of the amp and really get a few very different tones. But i REALLY liked the cut control a lot. In fact, thanks for reminding me because i think i'll give it another shot, or maybe even make it switchable as another option. Another thing that turned me off to it tho was even more than NFB it really limited the amp's volume a lot.

[Gaz]

Yeah the hi-cut changed my f'ing life. For years I was always trying to dial out the harshness by turning down the presence and treble, which only muddied things up, and somehow the harshness persisted. I eventually blamed it on 'cone cry', and decided that that's what one gets for playing so loud. I'd put one in on all my amps if it didn't kill their 'value'.

Right now I'm using a presence and hi cut, and although it seems like they might cancel each other out when they're both turned up (presence up and highs cut), they don't!

[daz]

I think i tried it while i had the NFB in before. But the amp sounds so much more alive, richer, more organic without NFB. So today i'm putting a cut in and i will probably remove NFB, but i will give it time to evaluate it and then i may go back with NFB and see how i feel then. But i know what you mean when you say it changed your life. First time i tried it i was amazed too, but i felt the amp lost too much volume. But thinking about it now, i seem to lose quite a bit with NFB too so cut may be the better method.

[daz]

Heres a related question....the matchless chieftain which my power amp is closley modeled after has no NFB and the PI cathode and tail resistors are very different. If i recall 1.2k/47k, while mine is 470R and 10k as a JCM 800 with NFB. Are these values likely so different because of NFB in the JCM vesrses the matchless' lack of NFB, or is it just pure preference of the designer? I wonder if i should use those values. I've tried them with NFB and didn't like it, but i wonder if now with NFB disconnected and a cut in place instead if i should go with the higher values of the matchless. I'd just do it, but i wanted to get some thoughts first.

[pdf64]

I guess that the values of the components for the phase splitter in the matchless were optimised for driving EL84s, which need much less input signal compared to EL34s in the JCM, as well as to keep the overall gain down with being open loop.
Both the phase splitter gain and max output signal will be affected by both the bias and tail resistors. Also the tail resistor increases, the phase splitter becomes more accurate in term of how well the 2 outputs are matched. That's why vox/matchless can have both plate resistors at 100k, rather than the 82k/100k that fender/marshall use to try and get better balance (as their tail resistor is only 10 or 22k). Peter.

[daz]

I'm talking about the matchless chieftain which is EL34's, not 84's.

[tedmich]

to follow up on Merlin's post, the smaller resistor from the leg to wiper (to paraphrase) is a way of changing the taper of a potentiometer, the tie leg to wiper wiring is the rheostat/variable resistor use of a pot. Having the wiper NOT tied to a leg gives you a true voltage divider, and the added resistor will change the taper. See "the secret life of pots" and Aiken Amps for some discussion of changing pot tapers.

With the right resistor you can make log, neg-log etc. tapers and maybe get one perfect for your NFB control.

Whats this high cut stuff? I am looking to add the PV Windsor Presence/ Resonance control to my PP EL84's NFB very soon, check it out its pretty cool (even has a patent...)

[daz]

Whats this high cut stuff?

You know what a cross like master is? The cut control is the same thing but you put a cap in series with it. the value is up to you, but i use a .022uf. The matchless uses a .0047. I found the .022 does just what i want better than any other value i tried. anyway, just wire a pot (i find 50k about right) as a variable resistor from one power tube grid to the other just before the grid stoppers, and put the cap in series. works very similar to NFB except i like the smoothing effect better. There are trade-offs tho.