your thoughts on NFB cap

[daz]

At the risk of taking this forum over completely, (sorry, but i'm rather inquisitive lately ) what are your thoughts on this. i decided to try a larger cap in the NFB circuit. My novice reasoning was that a smaller cap makes the entire amp sound brighter/thinner, so maybe a BIGGER cap would do the opposite. But i never see that in other amps so i figured my reasoning must be flawed and it wouldn't work. but to my very happy surprise, it worked great. Of most every amp i ever owned and the few i've built, i've always felt a need for a thicker tone. But every thing i've ever tried in the design of my builds or in modding production amps or even just trying to accomplish this with pedals or such, has never worked. If i got the thickness it always sacrificed elsewhere, usually by losing the top end in some way, often the sparkle would be lost and it would sound dead.

Anyways, i tried a few values and ended up with a 2.2uf. of course i had to use a electrolytic, but it worked fine. the tone is now just thick and ballsy like never before, yet i haven't lost a thing except the tinniness. It seems to have accomplished something i've tried to get forever in my tone. I guess you could say it's kinda like the difference between a bright JCM800 and the proverbial plexi tone. I did this on a 20 watt 6v6 i have, but wanted to do it to my main amp, a 2xEL34 build. I tried it on a fluke today while near my 20 watter, but i very anxious to try it on my main amp tonite.

I posted this just to get some thoughts from you guys as to why this worked so well yet i never see anything bigger than a .1uf in any schematic i ever looked at. Both amps are high gain preamps with cathode biased PA's and of course NFB.

[defaced]

It worked so well because electrolytics don't pass high frequency well so you lost the tinniness and by raising the value so much you dropped the frequency cutoff point of the cap and the series resistor to where you liked the sound. I would guess that if you put a film cap in parallel with the electrolytic cap you'd get some high end back.

As to why it's not typically done, I don't know. I prefer having more than one NFB control so my build has a depth control in addition to the presence. I can get some serious chug out of that setup.

[daz]

I wondered about that. that is, a second control with the .1uf to dial in presence. but i think if i did that i'd peobebly be back were i was before. Not quite sure how that would work.

[Chuck H]

Something to keep in mind is that the resistor values used in the NFB loop will also change the knee. Like you, I noticed that most amps built prior to about 2000 seem to use the .1uf presence cap value like it's etched in stone. But that value was made popular by it's use in the early Marshall amps (and the 5f6a Bassman). A typical Marshall has a feedback resistor of 56k to 100k and a shunt resistor of 4.7k. Lets use the 100k value for dicussion. If you change that up and use a feedback resistor of 10k and a shunt resistor of 470 ohms the amp would sound basically the same because the ratio of the voltage divider for the feedback loop hasn't changed. But as soon as you turn up the presence control you would notice that the knee frequency is now much higher than it was with the 100k/4.7k divider. Thats because you changed the impedance of the circuit and a capacitors frequency knee is impedance dependant. I can't remember if there's any square or log formula to the math for figuring the new cap value with an impedance change. It may just be a ratio.

My latest build has a 22k feedback resistor and a 1k shunt. I tried the .1uf presence cap value because that's apparently the the "correct" cap value for a presence control. No dice. Only the very tippy top end would increase with the presence control. I remember asking about it here and figuring that with my feedback loop I would need a .47uf cap to approximate the same frequency response of a typical Marshall type presence control.

Sooo... My NFB divider values were roughly one fifth the size of a typical Marshall. And the cap I needed to make it sound right was roughly five times larger than a typical Marshall. I expect that if I had a voltage divider of half the typical Marshall value that I would have needed a cap of double the typical Marshall value. So thats the ratio... Sans math.

If your 6v6 build uses a different feedback loop (especially if the resistor values are lower) than a Marshall amp that would explain why you needed a large cap to get out of the tinny tipy top area. It's also important to observe the NFB loop resistor value differences between your 6v6 amp and your el34 amp if you plan to try this change on your el34 amp. Otherwise it wouldn't be apples to apples.

Chuck

[Merlinb]

At the risk of taking this forum over completely, (sorry, but i'm rather inquisitive lately ) what are your thoughts on this. i decided to try a larger cap in the NFB circuit. My novice reasoning was that a smaller cap makes the entire amp sound brighter/thinner, so maybe a BIGGER cap would do the opposite. But i never see that in other amps so i figured my reasoning must be flawed and it wouldn't work.

There's nothing wrong with that at all. It simply turns the presence control into a feedback control- less feedback means more gain, tighter bass and a narrower, more 'punchy' bandwidth.

[daz]

I tried it in the el34 amp and got similar results. Not as much but the same sort of effect. But then that amp always has less overall effect with NFB anyways for some reason. Both amps have typical marshall values.

Check, are you saying that bu using a 10k/470R in place of the current 100k/4.7k that i would have the same sound as i do now with the 2.2uf cap but that when i turn the presence pot up it would add very high frequencies to it? If thats what you are saying i'd like to try that because i would then have this thicker tone but with the ability to get back some of that very top when i want that. What i was thinking of doing was putting the 2.2ug, the original .1uf, and a 1uf on a rotary switch. But if what i described IS what your idea would do i'd prefer that because i could balance this thincker sound with that very high frequency bit and get any balance of thick/bright that i want.

[Chuck H]

What I'm saying is that a lower impedance circuit will change the frequency that the presence cap controls. If you lowered the impedance of the circuit you have now it won't add top end, it will move the effected frequencies up. Your presence control would have less effect on the midrange frequencies it's controling now.

You mention that you are using a 3.3uf electrolytic cap now. You could try adding a .1uf in parallel to regain some of the top end. It wouldn't change the overall value much. You would go from 3.3uf to 3.4 uf.

You could also use a film cap instead of an electrolytic. That circuit is not high voltage so a 100 volt cap should work fine. A 2.2uf or 3.3uf 100V film cap should be a managable size.

What you have done is actually pretty clever. As was mentioned in an earlier post your presence control is now behaving more like a variable feedback control. But because of the cap in the circuit you are not changing the amount of bass in the NFB loop. So your bottom end stays tight and you get a boost of raunchy harmonic laden mids.

IIRC your el34 amp has some a somewhat different NFB loop than typical. That could be why you are getting less effect from the circuit change in that particular amp.

Chuck

[daz]

Thanks Chuck ! Thats good stuff, and i see now what you mean. This opens up a whole new world that for some puzzling reason it seems most amp builders don't seem to explore. but i love this because it gives you a totally different way to EQ things that works very differently than EQ'ing the preamp. i will try your suggestions and i can even see having pots in place of the 100k and 4.7k to experiment with the infinate variables here. this is fantastic. It's truly puzzling why this doesn't seem to be a very well explored area, but i sure as haeck will explore it. thanks again. Now that i understand it i know where to go from here.

Edit: well, i tried the .1uf in parallel but couldn't tell a difference, at least nothing i was sure of. I did however replace the 2.2uf electro with 3 1uf films in parallel. Not quite sure yet but that seemed to sound a bit better. Not sure where i'm going next with this, but i think i may try a second pot for presence and adjust it for very high frequencies to add a touch more sparkle to this big meaty tone. Here i was thinking the amp was done for the last couple months or so, but my inquisitive nature had me playing with it again ! But i'm glad i did. No amp is perfect, but i keep getting a bit closer to that impossible goal.

[GibsonLover]

Here's a completely different application of a Presence control, although the discussion so far completely applies for a different take on Presence and EQ. I have a Lexicon Signature 284 amplifier, and noticed the Presence control, along with the rest of the EQ, is very useful. Check out the fourth gain stage in the schematic. This amp is so versitile, it can nail just about any tone well, except maybe a clean Fender sound.

[daz]

That one has me lost. I'll stare at it some more when i get a chance, but it's just not registering right now.

[Chuck H]

It's not as exciting as all that really. Experimenting with the values will have the primary effect of changing the feedback ratio. And IIRC you worked pretty hard on the NFB loop in that amp. If you change the ratio now you will alter the tone for better or worse. Go with your ears.

Changing the values in ratio, as in 10k / 470 ohm instead of 100k / 4.7k, won't alter the tone of anything but the presence control. And then it will only shift the effected frequency. You can do the same thing with the cap value too. I just don't think you'll find anything too exiting there. Once you get the feedback ratio where you think it sounds good that's about all you can do. You can then add stuff like presence and resonance controls.

I do know that Dan Torres (Torres Engineering) in San Mateo California does a dual presence control circuit. I'm not sure how he does it but if you order the "kit" you'll know soon enough. You might change the cap values he has in the kit so one knob effects the mids and the other the top end. His is set up to control the top end and the very top end. But buying the kit will at least show you how he's doing it because he includes a schematic.

Chuck

[daz]

Chuck,

I'm at that point right now where i'm seeing this for what it is. basically just EQ'ing the PA. But it does have a lot of potential the same way there is a difference between EQ'ing a signal pre or post distortion. granted, in my amp both are post, but still the results are different and i can get effects that don't work the same when tweaking the tone stack or preamp stages. I'm trying to figure out what it is that makes my 6V6 amp so much more sensitive to these mods. They DO work with my EL34 amp too, but they aren't as beneficial. If i can figure that out then i will strat experimenting with 2 or more filter sections in the NFB.

[daz]

I think i now know why this hasn't been explored more. It worked great on my 6v6 amp. But it's really not necassarily even good on my EL34 amp. I guess it's the way the different tubes work. It's not the first time i've found things that worked good with my 6V6 and not my el34 amp. But anyways, since most 6V6 amps are fenders which AFAIK usually have no NFB, or and least not with adjustable presence, then it wouldn't be doable there. And it doesn't seem to do much good with el34 amps which are the ones that usually DO have NFB and a presence. So thats likely why this hasn't been explored more. My 6V6 amp is undeniably better with a 3.3uf cap in the NFB, but with the el34 the jury is still out.

[hasserl]

With the larger cap you're just dumping more lower frequencies from the nfb circuit, which means those same frequencies are boosted in the PA. Instead of just boosting the higher frequncies typical of a Presence control, your now boosting a wider bandwidth with more lower frequencies. As Merlinb put it in hius post above, you've changed your Presence control into more of a variable nfb control, kind of like what some guys call a Soul control.

[Steve A.]

... I posted this just to get some thoughts from you guys as to why this worked so well yet i never see anything bigger than a .1uf in any schematic i ever looked at. Both amps are high gain preamps with cathode biased PA's and of course NFB.

FWIW the 70's *umble schematic shows an Accent switch instead of a presence control and it uses a 3.3uF/35vdc electrolytic cap.

Steve Ahola