2x6V6 push-pull power amp - Fender Deluxe reverb copy

[wizard333]

IMO - that seems more opinion that reasoned. A string slap or pick impact has a whole range of freq's in it. One man's "noise" is another man's percussion. I don't think we should call drums "noise" just because they are non-harmonic. I'm OK to say it's a matter of opinion and taste so let's not dwell on the word here.

OK - interesting point. But we're not talking about fundamentals under 80Hz anyway - as pointed out above... The low stuff is more percussive in nature.

All those things you mention have a whole overtone frequency, which is what you hear in a mix, that is well above that subsonic range. Are we talking about things produced by a guitar, or by a bass drum? Even then, same rules apply in mixing.

You're original question was about "Chunka Chunka" guitar and I answered it. But don't take my word for it, stick the biggest coupling caps you can in your amp and listen to the results.

[uneumann]

In the last year I was messing with a JFET preamp that used an opamp servo to set the JFET bias. The problem was a peak in the very low bass, usually between 5Hz to 10Hz of 3 to 6dB. The source of the circuit I was trying to work with came from the Sunn Stinger 60. The schematic is posted on Ampix, but I can't access it just now. Anyway, the circuit just didn't sound right with that low frequency bump. The most practical fix is to roll off the bass at the input to the stage, and I didn't want to do that so I abandoned that circuit.

That sounds tricky. There must be some LF time constant for the servo and if that's too fast or changing in the range where you can hear it, it would cause strange effects.

The key to good sounding bass seems to be a slow dropoff in response staggered between stages, not all the poles right on top of each other causing too fast a rolloff. A shelf type response works in many cases.

Agreed - if you're going to roll off at any point, a series of "close" stages is likely to behave and sound better than exact repeated poles. A bit of speculation here, but could be related to the accumulated phase shifts all changing at the exact same pole(s) that makes it sound weird.

[uneumann]

Yes - you're right --- thanks for bearing with me... I'm playing devils advocate because I have conflicting evidence in my own amps and it's puzzling to me that few discussions seem to drill down into this issue.

I have several amps that I've built or moded for extended low end responses and without any ill side effects... (see my site at the signature URL) What got my attention here, was the initial discussion about the DR power amp (like mine) and then the follow on that went into cutting LF response. I had to respond to that.

My DR clone (described on my site) has very good LF response and I get none of the effects that are commonly associated with larger caps. I'm not trying to say anyone is wrong here. I know that simply extending LF response can and does cause bad things in many amps. What I am saying is that LF response doesn't HAVE to lead to these ill effects. They can be overcome. For example, the biggest LF issue is probably blocking in the PI and power stage, but simply increasing the value of grid stoppers can all but eliminate that blocking. I can't hear (or see on a scope) any significant blocking effects at all. Anyway it's a matter of taste in the end so there's no right or wrong, but I have to say it - LF response does not HAVE to lead to poor tone.

That said, my high gain channel does include some LF cuts. Distortion of LFs is not pleasant sounding due to the sum and difference frequencies produced as IM distortion. But that's a whole different issue. In the clean channel, I didn't cut LF anywhere, although I can't stop the OT and speaker from doing their part to limit whatever LFs get that far.

[Steve A.]

IMO - that seems more opinion that reasoned. A string slap or pick impact has a whole range of freq's in it. One man's "noise" is another man's percussion. I don't think we should call drums "noise" just because they are non-harmonic. I'm OK to say it's a matter of opinion and taste so let's not dwell on the word here.

I don't disagree with you. It seems to me that whenever you play two different notes you also get at least trace amounts of the sum of the two frequencies and the difference of the two frequencies.

However I do think that if a DR or other Fender BF/SF amp is exhibiting farting characteristics the easiest fix is to go with a smaller coupling cap. Besides the frequency response as determined by RC I think that larger caps pass a stronger signal than a smaller cap, and IMO that needs to be taken into consideration as well. When Leo was designing his amps I don't think that he was always calculating RC values in determining the value of coupling caps- with his eyelet boards it is really easy to just swap caps and see how it sounds. I think that theory is great for getting you into the ballpark but once there I would rather trust my ears than my calculator.

Steve Ahola

[Bruce / Mission Amps]

I don't disagree with you. It seems to me that whenever you play two different notes you also get at least trace amounts of the sum of the two frequencies and the difference of the two frequencies.

However I do think that if a DR or other Fender BF/SF amp is exhibiting farting characteristics the easiest fix is to go with a smaller coupling cap. Besides the frequency response as determined by RC I think that larger caps pass a stronger signal than a smaller cap, and IMO that needs to be taken into consideration as well. When Leo was designing his amps I don't think that he was always calculating RC values in determining the value of coupling caps- with his eyelet boards it is really easy to just swap caps and see how it sounds. I think that theory is great for getting you into the ballpark but once there I would rather trust my ears than my calculator.

Steve Ahola

Yeah.. the best is to simply, empirically test it all for your own satisfaction and proof...

One more thing... because I read a number of people loosely quoting this....

"the sum of the two frequencies and the difference of the two frequencies"

Look up what that really means with respect to the sub fundamental frequencies that the original poster is talking about....
I think what he is describing is actually more like harmonic mixing or sub harmonic mixing.... which are a different mathematical "formula" then that of simple frequency mixing, (which is linear mixing).

[uneumann]

One more thing... because I read a number of people loosely quoting this....

"the sum of the two frequencies and the difference of the two frequencies"

Look up what that really means with respect to the sub fundamental frequencies that the original poster is talking about....
I think what he is describing is actually more like harmonic mixing or sub harmonic mixing.... which are a different mathematical "formula" then that of simple frequency mixing, (which is linear mixing).

Bruce - just so I'm not misunderstood, let me clarify...

Distortion comes in many forms. Broadly defined, distortion is any output signal variation from a multiple of the input. A perfectly linear amp adds no distortion. Today's SS amps are good approximations of that. Now a clean channel in a guitar amp produces some harmonics, which we all like and sound good to us. See this definition of distortion in general, but specifically harmonic distortion. These are multiples of the input frequencies, so they're "musical" in a the sense that some of them (second and third harmonics) have musical pitch relationships to the input. This distortion is NOT what I'm talking about in my above post.

When you have a high gain channel, the whole point is to clip and distort the input. The extremely non-linear behavior of that channel gives rise to another form of distortion called Intermodulation or IM distortion. See the above link again for that definition. The IM distortion is by definition - the sum and differences of different input frequencies passing through the distorting circuit.

So, let's take two notes - my favorites for testing this are an open A (110 Hz) and a low G (98 Hz). Play those at the same time into a dirty channel and you'll get lots of stuff coming out - including their sum and difference frequencies. So the sum is 208 Hz, and their difference is 12 Hz. Neither of these is usually musically related to the input notes. Although in this case the 208 Hz happens to be a G#, so its a flat-9 above the input G note. That might sound OK... but the real problem here is the 12 Hz difference frequency. This is the one that I find noticeably irritating. It's too low to be audible, but it produces a motorboating effect that's very grating and has a "ring-modulator" quality to it - but not a pleasing one. Anyway - that's the "sum and difference" that I'm referring to.

That's why I (and most designs) choose to limit the LF response of the high-gain channel. It's intentionally producing a lot of distortions and I don't want these ultra low difference-frequencies to be created in the first place, nor do I want them passed along if they are created. Note that this is very different from what happens in the clean channel - these sum/difference frequencies are not created (or not much of them anyway) so the LF response of that channel doesn't cause any problems. Note also, the PI and power stage has to pass both the clean and dirty signals, so it has to have good LF response and all the LF limits of the dirty channel have to be implemented in the pre-amp circuitry. Once either signal gets to the power stage, it's passed to the speaker.

Hopefully that clarifies what I'm saying in terms of distortion.

[Steve Conner]

There is no such thing as "linear mixing". If the circuit is linear, mixing just won't happen.

All forms of non-linearity produce both harmonic distortion and IMD. It just depends on how you look at it. If you test with a single frequency, you'll see harmonics. Test with multiple frequencies and you'll see IM products too.

The only "gotcha" is that a perfectly symmetrical circuit doesn't produce even-order distortion harmonics, and doesn't produce difference frequencies or other even-order IM products. It doesn't really matter in practice, as even push-pull tube output stages happily unbalance themselves when overdriven.

Removing LF at the beginning of the gain structure and adding it back at the end is a classic high-gain trick.

[Steve A.]

Removing LF at the beginning of the gain structure and adding it back at the end is a classic high-gain trick.

And adding HF at the beginning of the gain structure and removing it back at the end is a classic noise reduction trick.

Steve

P.S. The interaction of the notes on two or three guitar strings is one of the most important things to me in the sound of an amp, an FX pedal and/or a pickup. If it sounds sour then something in there is wrong- at least for me.

[wizard333]

Clean and distorted are entirely different worlds when it comes to low end. You can get gobs of low end out of a clean channel and make it work as long as you don't distort it. But non distorted guitar isn't exactly "chunka chunka" eh?

If you distort your clean channel, you get the same issues. That's the reason why BF DRs sound thin and nasty clean, but good when cranked, and SF DRs need the bass knob at 0 to do the cranked thing without farting. PI input coupling cap, .001uf vs .01uf.

If you add lots of lows in the pre-amp then cut them all out with small couplers in the power section, you've then limited your lows haven't you? Its just a question of where. Big grid stoppers? Now we're getting into other tonal issues.

You *can* do *lots* of things, but in the end, if you don't limit your low end somewhere, you're going to get unpleasant effects and noise. As I said before, if you really want those frequencies, the best solution is a powered sub woofer.

[dobrota]

I tried various preamps - classic Fender preamp - one 12AX7 with 1k5/22u in cathodes and EQ and gain between stages, I have to set bass about half and both middle and treble very low else sound is very thin and trebly. With Marshall JCM800 2204 model preamp design it's worst, I have to set both middle and treble to zero and then sound is still thin again. Thirdly with Soldano / MB recti preamp design sound isn't so bad with normal EQ settings but it's still little strange and like acoustic. Maybe reason is DR power amp design and it's frequency response.

[wizard333]

Its not the power section. I've modded plenty of those and if anything that power section is a tad hot in the high mids, I dial it back some in the pre-amp but it is definitely not a thin sounding design at all.

[g1]

As far as the question about powered subs for guitar, ISP makes them: http://www.isptechnologies.com/portf...tar-subwoofer/
There are a few different models available with various speaker configurations and power levels. Some models even remove some of the low bass from the thru-put that goes back to your combo speakers, which should make them less muddy.

[Steve A.]

There are a few different models available with various speaker configurations and power levels. Some models even remove some of the low bass from the thru-put that goes back to your combo speakers, which should make them less muddy.

Along those lines could we just design a high pass crossover circuit that would allow us to cut the low bass frequencies? Preferably with a control to allow us to dial out as much low bass as we want and perhaps with a switch to select the cut-off frequencies?
This would not be a cure for blocking- you deal with that in the amp circuit- but sometimes there is too much bass but when you dial back the bass and mid controls in the tone stack the sound gets too thin. You might compare it to a Cut control which cuts high frequencies going into the power tubes- only it would cut low frequencies after the OT. It could be installed inside the amp chassis- or in an external box.

Whaddaya think... is something like that feasible?

Steve Ahola

[Steve A.]

I tried various preamps - classic Fender preamp - one 12AX7 with 1k5/22u in cathodes and EQ and gain between stages, I have to set bass about half and both middle and treble very low else sound is very thin and trebly. With Marshall JCM800 2204 model preamp design it's worst, I have to set both middle and treble to zero and then sound is still thin again. Thirdly with Soldano / MB recti preamp design sound isn't so bad with normal EQ settings but it's still little strange and like acoustic. Maybe reason is DR power amp design and it's frequency response.

Cutting and pasting different preamp circuits with a given power amp can lead to mismatches in levels and tonalities, perhaps even impedance. If you posted the circuits you are experimenting with I think that the experts here could give you more specific advice as to solving the problems you have mentioned.

Steve Ahola

[g1]

They do it with a high-pass filter ( http://www.isptechnologies.com/portfolio/vector-112/ ), so it's probably fixed as they don't mention it being adjustable. I guess the feasibility of it would come down to the economics of making an adjustable unit.