RESOLVED: Peavey Deuce Negative Feedback Loop - Presence Capacitor

[ryanpg]

EDIT: See below for solution to "dull tone."

Obsession? Foolish compulsion? Fun learning experience. Probably a bit of each.

I might take this over to AudioKarma but I thought I'd float it out here first. I've already documented my changes to the preamp section (changed tone stack, and diode clipping added). But I still feel the amp is lacking in high-end presence.

I decided to do some basic testing (which I should have done first) and plugged my guitar directly into the "PWR.AMP IN" jack. Dull tone. I then ran through a MXR active eq and boosted the 3.2k band, and boom! There's the frequency I'm missing.

Low pass or high frequency signal to ground? So I started looking for low pass or band pass filters in the power section. So far my novice eyes have found nothing. BUT, there's a NFB in there. And it looks to me like:

IC25 39pF and IR24 1k (and IR23100k) create a high pass filter? Would lifting the NFB (possibly with a switch) or better yet, sending more high signal to ground help bring in presence? This is the way early Fender amps implement a Presence control. I wonder if it would be safe to just lift one leg of IC24 to see what happens without the feedback loop at all.

Wrong as pointed out by Enzo below -> Maybe a simple mod: omit IC25 39pF, replace IR24 1k with a 75pF capacitor. This would send signal above 21k to ground. 82pF would send everything above 19.5k to ground.

[Enzo]

Do you think 20kHz is what is missing from your sound, that's a long way from 3.2kHz? That looks to me more about stability.

If you replace R24 with a cap, you destroy the NFB. The whole idea here is a voltage divider. For anything below the C25 freq, there is a 100k over 1k voltage divider, so for most audio freqs, we get 1% of the output voltage as feedback.

Can you lift that cap? Of course you can. But I think you are overlooking the largest audio filter in the amp - the speakers. There is no one thing in any amp that affects the sound more than the speakers. Do you have the Black Widows? or do you have stamped baskets? Or other?

Have you unplugged the amp speakers and connected the chassis to other cabs?

On other fronts, you played the guitar into the power amp and found dull, OK, but then you added an EQ and boosted the high end to get what you wanted. Then did it occur to you to tweak the preamp to add that same top end?

[ryanpg]

Enzo, thanks much for your response. Hope you don't feel like you have to solve my tone problem! But you're the man when it comes to these (and many other) devices!

Yes of course, you're right, I was way off on my math re low pass filter on the NFB. As you point out, my overall understanding was really flawed anyway.

"If you replace R24 with a cap, you destroy the NFB. The whole idea here is a voltage divider. For anything below the C25 freq, there is a 100k over 1k voltage divider, so for most audio freqs, we get 1% of the output voltage as feedback."

Thank you for that explanation. I understand the voltage divider now (R23 and R24). Thanks. So anything below cutoff of C25 passes to the NFB. Perhaps changing the value of C25 to lower the cutoff would help?

Excellent general guiding questions you pose.

Re: Speakers. This was a really big part of my consideration. I've got the original BW speakers loaded. With 10" 50-5000hz speakers it sounded a bit brighter, but still muffled to my ears. Through my 4x12 celestion cab, still dull. (My Blackstar and blackface bandmaster are both quite bright through the celestion cab, and through the BW speakers.) I think the BWs can replicate the frequencies I feel I'm missing.

Re: Preamp. Yes, I did EQ mods to boost the high-mids and highs. I did some signal tracing, and at one point entirely bypassed the tone stack. Same dull sound. It wasn't until I plugged directly into the power amp section that I realized the source of the high-end loss was likely found in the power section. (My last experiment is to take the preamp out from the Deuce and into a very neutral power amp that I know sounds good. My guess is that it's going to be very very bright.)

Since this is a passive circuit, I can only cut mid and bass and pass as much treble as possible. (I thought about finding a place to implement a "treble boost" circuit by running a cap in parallel to an op-amp input resistor.)

Here are the values I settled on:

C30 from 270pF to 450pF
R45 from 47k to 90k
C32 omit



I've been looking all over the schematic to find where the ~3khz signal is lost. It sounds like the high-end is being shelved somewhere. I've ruled out several spots. The NFB is/was my last hope to understand and find the "issue."

A big part of this is my aging ears too. Maybe the amp just isn't for me. But I feel like there's an amp I'll absolutely love just hiding in that Deuce!

[Enzo]

So disconnect the NFB and find out if it is involved. Simple enough, and much faster than batting theory around.

meanwhile if the existing NFB is at 20kHz, I have to think raising it to say 30kHz won't help.

I think you should reconsider the preamp. The tone stack is only one small part of it. The whole amp is the system, if the power amp lacks emphasis at 3k, and you want some, then why would it matter where you get it?

In general, on these fine old amps, I expect to have to replace all the small electrolytics, those 2uf/35v ones in particular. They dry out, get leaky, and can affect your tone. I am too lazy to do the math, have you figured what all those 100pf across the feedback resistors are doing on the op amps?

The tone stack is only one small part of this. Look up top, coming off U3a is C9, a whopping 0.1uf across the signal path. What if you reduced that a ton? Same thing at U2a. C5 and C24 are less aggressive, but could be a factor. Lower right C72, R113. Could they be leaning on your signal?

I think you could easily bump your 3k in the preamp.

[ryanpg]

I think you're right about revisiting the preamp. Today I took it from pre out to another power amp. It didn't have the brightness I hoped for.

". . .have you figured what all those 100pf across the feedback resistors are doing on the op amps?"

I've been researching those odd little 100pf/33k and 100pf/22k RCs in the feedback loop. I can't find an exact reference schematic, but I've come across something that looks similar. The circuit resembles an active non-inverting lowpass filter (as illustrated here http://www.electronics-tutorials.ws/fil ... ter_5.html).



They seem to have a cutoff of 48k and 72k respectively. But again, changing values produces no significant change in tone.

Which makes me think about those 2/35v caps you mentioned. I guess if one is out of spec it could drag a lot of signal to ground.

"The tone stack is only one small part of this. Look up top, coming off U3a is C9, a whopping 0.1uf across the signal path. What if you reduced that a ton? Same thing at U2a. C5 and C24 are less aggressive, but could be a factor. Lower right C72, R113. Could they be leaning on your signal?"

Thanks, I've actually explored a both of these areas. Starting with R8/C5 with a cutoff frequency of 16hz. I thought for sure that would be an influencing factor. But no, experiments with changing that cap produced nothing noticeable beyond a tiny bit more hiss. In fact, completely omitting C5 and then C9 only let a little more hiss through.

I looked at the line out section, specifically C72 and R113, but I think that's a high pass that allows everything above 72hz to line out. I don't see any signal going to ground or anything.

I think for now it's just got me beat. Been thinking about your speaker suggestion. I found eminence Red Coat Wizard speakers are really efficient and bright. Might just try that.

It's tough for me to admit defeat though. Especially when I know you're right, a nice 3khz bump is in here somewhere; finding it would save me the huge expense of a new pair of speakers.

Thanks again Enzo.

[ryanpg]

SOLUTION: I found the source of my perception of "dullness." It is definitely the NFB.

Following Enzo's advice I didn't "destroy" the NFB, nor did I eliminate it.

Placing a .22uF (I'm using a 100v) capacitor between the negative leg of IC24 25/25v and ground brought a good portion of the high end sparkle in.

Basically, adding this cap to ground creates a low pass filter in the negative feedback loop. Since high frequencies are eliminated from the signal coming from the 4ohm tap on the output transformer, those frequencies are NO LONGER attenuated by the NFB. I don't have the equipment to do testing, but the effect is pretty profound. There is definitely more noise (hiss) but along with that there's quite a bit of shimmering highs introduced into the tone.

NFB_presence_cap.png (271 KiB) Viewed 2719 times

Hope this is helpful information for other OCD types that feel the Deuce/Mace is a tad too dark. Thank you Enzo for talking through this with me.

[ryanpg]

To illustrate the difference in tone, I did a horrible phone recording. This does NOT represent an overall quality sound! It's a shaw t-top humbucker in a LP style guitar. You'll hear without cap, with cap, without cap, and with cap - one take, just inserting and removing the cap. There's a lot more high getting through with the cap in place. (This is with bass 10, mid 4, and treble 10 and all the tone stack mods too. Again, not meant to sound good, just get as bright as possible.)

https://soundcloud.com/ruinaudio/nfb-presence-cap-demo