Thoughts on a frequency specific AVC circuit

[1blueheron]

Perhaps this is a newbie question that has been addressed in another thread or come up before.  I did a brief search and didn't find anything that seemed to describe what I am trying to acheive.

My goal is to develop an active gain control that will limit lower frequencies to a specific level so that they do not push an amplifier into distortion prematurely and result in a "farty" sound but at the same time not roll off all the bass response so it sounds thin or not compress the entire signal so it sounds lifeless without dynamics.  I want a tight, well defined bass response that stays sharp on attack but also plenty of mid and high frequency distortion with a nice crunch and sizzle.  Too many amps turn to mush in the lows when you crank up the gain to the point that the distortion is just getting nice harmonics.

In studying the 6BA6 tube, it seems that its design purpose was automatic volume control, primarily for AM radio which tended to drift in and out.  A form of variable compression.   I am using this as somewhat of a jumping off concept.

Is there a practical way to design a compression circuit that has a frequency cutoff point in a tube amplifier which adjusts automatically with the overall volume of the signal.  Any practical exeperience that could be shared would be appreciated.

[shooter]

try using AGC in your surfing, maybe tube based AGC circuit.
as to frequency specific, maybe band trap, band pass filtering

[1blueheron]

Maybe it would be more beneficial to focus on a "Resonance" control in the negative feedback loop?  Just read Aiken and from how he describes it, it should do largely what I am looking for.  I can theoretically cut the low frequency response in the preamp stage which would prevent premature distortion and then boost low frequency by means of frequency selective NFB.  Apparently this is not simple and if not done correctly can cause oscillation and such?

[shooter]

In general, any feedback loop "system" is complicated, more complex than i typically mess with in amps.


If farty is your goal, messing with the bypass caps in the 1st 2 gain stages has "solved" that for me.  If that didn't quite do it, messing with the TS comes next.  Plus I always start with a speaker that is typically 2X my amps power and spl's >90


If none of that gets you there, then maybe something complex

[HotBluePlates]

Is there a practical way to design a compression circuit that has a frequency cutoff point in a tube amplifier which adjusts automatically with the overall volume of the signal. ...

Let's think about this proposed circuit for a moment:

If it is compression that triggers only for low frequencies, then it will compress (or otherwise reduce volume) for all frequencies.  Unless we split the audio path into high frequencies & low frequencies, trigger & apply the compression only on the low frequency path, and then mix the to bands back together.

There's gonna be some phase-weirdness when the bands get re-mixed, but nevermind that for a bit.

... My goal is to develop an active gain control that will limit lower frequencies to a specific level so that they do not push an amplifier into distortion prematurely and result in a "farty" sound ...

Read what Aiken has to say about "blocking distortion."

99% chance what you're describing is bias-shift as the output tubes distort, and blocking distortion caused by slow recovery from this bias-shift because the coupling caps (and grid-reference/bias-feed resistors) are too large a value.  Trying to build a multi-band or frequency-dependent compression circuit is "swatting a fly with a bazooka" when you could just fix the problem at the source.

An additional/alternate solution for the bias-shift/blocking-distortion problem is lowering the output impedance of the phase inverter, and increasing its ability to deliver current to the output tube grids.  Taken far enough, you're converting your amp into a Class A2 or Class AB2 amp (that stays cleaner to a higher grid-drive voltage and output power).

When do you experience "farty bass" or blocking distortion now?

[1blueheron]

In general, any feedback loop "system" is complicated, more complex than i typically mess with in amps.


If farty is your goal, messing with the bypass caps in the 1st 2 gain stages has "solved" that for me.  If that didn't quite do it, messing with the TS comes next.  Plus I always start with a speaker that is typically 2X my amps power and spl's >90


If none of that gets you there, then maybe something complex

Eliminating farts is my goal.  I see where my phrasology was confusing.

[1blueheron]

Is there a practical way to design a compression circuit that has a frequency cutoff point in a tube amplifier which adjusts automatically with the overall volume of the signal. ...

Let's think about this proposed circuit for a moment:

If it is compression that triggers only for low frequencies, then it will compress (or otherwise reduce volume) for all frequencies.  Unless we split the audio path into high frequencies & low frequencies, trigger & apply the compression only on the low frequency path, and then mix the to bands back together.

There's gonna be some phase-weirdness when the bands get re-mixed, but nevermind that for a bit.

... My goal is to develop an active gain control that will limit lower frequencies to a specific level so that they do not push an amplifier into distortion prematurely and result in a "farty" sound ...

Read what Aiken has to say about "blocking distortion."

99% chance what you're describing is bias-shift as the output tubes distort, and blocking distortion caused by slow recovery from this bias-shift because the coupling caps (and grid-reference/bias-feed resistors) are too large a value.  Trying to build a multi-band or frequency-dependent compression circuit is "swatting a fly with a bazooka" when you could just fix the problem at the source.

An additional/alternate solution for the bias-shift/blocking-distortion problem is lowering the output impedance of the phase inverter, and increasing its ability to deliver current to the output tube grids.  Taken far enough, you're converting your amp into a Class A2 or Class AB2 amp (that stays cleaner to a higher grid-drive voltage and output power).

When do you experience "farty bass" or blocking distortion now?

The bass get loose or farty when the gain is turned up to where I am getting some nice crunchy mids and sizzle in the highs. And also when playing in drop D tunings.  It is primarily just on the initial string attack and then evens out. It is just hitting the pre-amp too hard in low frequency output.  I want to avoid changing the tone of the mids and the highs but tighten up the low frequency response without loosing low frequency information altogether, which I fear would happen if I just change bypass cap values.  If I go the other route and increase current flow, I fear the mids and highs will get cleaned up as well.  I guess I almost need to go a bi-amp route to get the sound I am searching for.  Compressing only the signals from say 50Hz to 150Hz.

[Willabe]

Can you post a schematic of the problem amp?

[shooter]

when playing in drop D tunings. 

    I only sound somewhat reasonable using  dropD


no matter what amp i use, that fat string all floppy is a pita.  Not being guitar savvy doesn't help.


Like Willabe said


If you have anything bigger than 10uF in the 1st gain stage, you want to knock it back 1st, then get complicated as necessary 

[HotBluePlates]

Earlier you brought up the Resonance control:

Feedback does a lot of things (not just "reduce distortion" as simple explanations usually say).  Among these, it tightens the amp's control over speaker motion, especially at the speaker's bass resonant frequency, where it tends to flap in the breeze.

Resonance & Presence controls seem to boost response in certain frequencies by reducing the negative feedback.  If you feel like the bass is "loose" now, the Resonance control will reduce feedback & the speaker's control in the bass range & make the amp even more loose.

The bass get loose or farty when the gain is turned up to where I am getting some nice crunchy mids and sizzle in the highs. ...

What is the amp?  Schematic?

Your problem is an issue of amp-design: In Guitar Amplifier Overdrive it is shown that if too much distortion is created at any single stage (or with too much bass), sounds like you describe happen.

So most high-gain designs today, distortion-per-stage is reduced by using "interstage attenuators" (a resistive voltage-divider), and by cutting bass in many places in the amp.  "High gain amps" actually add "a little distortion" over many successive gain stages until the overall high-level of distortion is achieved.

The bottom-line you'll eventually encounter is there is no way to solve the problem you're experiencing while:

• Keeping the amp sounding exactly the same as it does now (though you're trying to fix the sound)
• Keeping the distortion exactly as it is now (you could potentially have more, but distribute it differently)
• Keeping the circuit the same as it is now

A separate issue is when guys practice alone, they often set their amp for unrealistic sounds.  Usually there's way too much bass or low-mids making the amp sound "full" in isolation, but which sounds too muddy to work in a band context.

Sometimes the impulse to sound "full & rich" in isolation artificially creates these unrealistic goals, that also happen to result in side effects like farty bass.  There's a reason every distortion-focused amp since the 1970's Marshall amps strips out huge amounts of bass in their circuitry, and why there are Treble Booster pedals (which don't sound "trebly" when pushing an amp starting to breakup).

[1blueheron]

I really appreciate all the thoughtful comments and feedback. The points are well taken and received.  I don't remember where exactly I left off in tube rolling and cap swaps on this amp and will need to pull it apart and check against my original schematic before posting but honestly I have not heard an amp yet that fits the bill, commercially produced or home built.

I personally am not a player.  I am a listener and twiddler of electronics.  My son is the player.   I have certain things that I like and don't like.

When I was his age, music was changing and sounds came and went.  I am grateful that he has an appreciation and respect for some of the great icons and bands of my era.  His playing is greatly influenced by EVH, David Gilmour, George Lynch, etc.  That said, a current trend for today's young players is towards more than six strings and darker heavier, multi-string power chords in lower tunings overlayed with all the ripping solo techinique in the upper ranges that the guitar hero's of the past debuted.   Most of the classic amps just don't seem to be voiced for this mix despite being outstanding for the classic rock sounds in band settings.  It almost seems like today's aspiring guitarist needs to split his signal into two paths.  One into a solid state bass amp/speaker with the high end rolled off to get the modern "chunk" sound right, and the other into a Marshall or Soldano for the chainsaw ripping lead that can cut through the mix.

Perhaps it is a unicorn I am hunting.  I feel like it is technically possible, if I move outside the box, maybe we just spend so much time chasing after known tigers and bears like Marshalls and Fenders and Vox's we don't pay much attention to what would fit today's music scene best.

Closest thing I have heard so far is the PRS MT15. 

Looking at the MT15 architecture it has 5 gain stages and oversized output tubes for its rating.  This lines up with HBP's assessment of building the gain in more stages and more current to control speaker motion more tightly.  I guess I need more bottles.

[sluckey]

Kemper may have what you seek.

[shooter]

Not a player either, just a geek with a solder pencil, and an audience member of music.


I took the simplistic approach, design goal;
2 knobs and 20W.
The better the guitarist, the better my amp sounded, the better the guitarist, the bigger "range" my amps covered.
add pedals, post processing computers, and you have music for the masses, the fringe's, maybe not 

[jjasilli]

General rule for Overdrive:  reduce lo's on the way in; reduce hi's on the way out.  This is because lo's get muddy, as 1blueheron states; and OD produces tons of harmonics (extra hi's).  This is a known issue.  The simple solution is to further reduce lo's > your OD circuit.  Probable as simple as changing the value of a cap and/or resistor > the OD circuit.


A schematic would be helpful.

[HotBluePlates]

... Perhaps it is a unicorn I am hunting.  I feel like it is technically possible, if I move outside the box, maybe we just spend so much time chasing after known tigers and bears like Marshalls and Fenders and Vox's we don't pay much attention to what would fit today's music scene best. ...

At the hobbyist level, it's fair to guess there may be a lot of uninspired copying.

At a mid-level of experience, the hobbyist often has a lot of inspired experimentation.

I suspect at a high level of experience (where many amp manufacturers are), circuits trend back towards "tried & true" because it was found inspired-imagination didn't align with practical-reality.  The trick is figuring out, "What did the amp-makers know?"

... Closest thing I have heard so far is the PRS MT15. 

Looking at the MT15 architecture it has 5 gain stages and oversized output tubes for its rating.  This lines up with HBP's assessment of building the gain in more stages and more current to control speaker motion more tightly.  I guess I need more bottles. ...

Got a schematic?

You hadn't said what the starting-point amp is yet, so we don't know if the end-goal is unrealistic or not.  But most master volume amps assume the player will almost never distort the output tubes (since that's what the master volume & the preamp distortion architecture are there for).

We might also guess a master volume amp is made not only to get distortion at lower volume, but because blocking distortion at the output tube(s) are such a difficult problem, which is more easily & cheaply conquered in the preamp.  (Said as a guy that prefers simply playing non-master volume amps clean-ish, and using pedals for thick distortion/fuzz.)

[1blueheron]

I appreciate the EL34world brain trust helping me think through this.  The little AO-44 I just put together was a bit of an eye opener for me on what can be done using a good clean amp with a pedal and cements some of the advice given here in this thread. 

I don't think I have the budget for a Kemper, and it fundamentally has too many buttons and knobs and not enough bottles for a baby like me  but the point is well taken.

All that to say I don't think I am going to mod an existing working amp that I have.  I think I am going to attempt to build something out of the AO - 29 that I just got and of course I need to finish that 6973 Plexi project and see how it sounds. 

The AO-29 has plenty of holes to fill to do multiple gain stages.  From what I have read and seen, the MT 15 that comes close, derives some of it's virtues from building its gain in multiple small bottles.  It also uses a pair of under driven 6L6's to ensure clean headroom for the harmonics developed in the previous stages rather than trying to push the power tubes into distortion which is what I was previously trying to do and running out of steam before I got there.  Perhaps some version of this can be put together using the PT and OT from the AO-29.

I will abandon the idea of splitting the signal and automatic gain control for now, although I do think it might be an interesting project and has some theoretical merit, albeit plenty of technical hurdles such as phase shift and delay that would need to be addressed.

[pdf64]

Your query brought to mind a Frank Marino interview I read maybe 30 years ago, where he described how he’d developed a rig that used tube amps to get the right mids and highs, but solid state amps for the bottom end.
Can’t remember any detail sorry, eg if actual crossovers were used, but it may be somewhere on the internet now.

[shooter]

I do something similar with tubes;
I use a self biased PP 6V6 for my "cleans n tone" through a 2X12
guitar "y'd" into an SE KT88 for the "red meat" through a 1X15


neither amp hits 20W music, but together they rattle windows just fine 

[1blueheron]

I do something similar with tubes;
I use a self biased PP 6V6 for my "cleans n tone" through a 2X12
guitar "y'd" into an SE KT88 for the "red meat" through a 1X15


neither amp hits 20W music, but together they rattle windows just fine 

Do you have any active electronics or filters dividing the frequnecies or just a Y plug splitting the raw signal and then relying on the physics of natural frequency rolloff based on speaker cone size?

[shooter]

nope, just the "tone" designed into the 2 amps,  3 total TS's and 3 Vol controls + V & T on the guitar + 3 speakers and a non-player almost sounds good, and a great player is happy as can be with all the knobs