Direct Coupling with Zener Diodes

[J M Fahey]

Assuming we can't all look at a scope trace of a dummy loaded amplifier driven by a signal generator and know what it means adjectively to the tone.

You would be surprised
To be more precise, *changes* in scope traces can directly be related to *changes* in sound.
After seeing+hearing lots of different waveforms along time, you get to the point that hearing something evokes the waveform it *should* be producing and vice-versa.
Then, when you turn the scope on, you often find something very similar to what you expected.
As far as gain+cf stages, Soldanos have one *before* the effects loop and another one *after*.
There must be a good reason for that.
To boot, the attenuator between them is calculated so they clip at about the same time or, to see it from another point of view, to guarantee that no matter what you plug in the loop, either one or the other will clip in the particular CF way when expected to do so.

[woodyc]

Guys, we are all looking at the same data. I just have a different opinion on the relative merits of using a stage as a cathode follower when you don't need the low output impedance.

I’m totally up for a debate. But first I’d like to wire up a simple objective test kg’s assertions about direct coupling. I won’t have time until this weekend. Can it wait til next Monday before I pit my manly 35dB against your puny CF?

In the meantime, a few things to ponder… After Leo Fender pulled the CF out, he produced the blackface amps, which are not tonally inferior to anything. When Marshall pulled the CF out, they produced the later JCM800s (Michael Doyle cites the 2210 as being among the best distortion amps of all time.) And speaking of Marshall, believe it or not, a popular mod in pro amps since the early 1970s has been to convert the CF into a gain stage. These days, people are even selling new amps based around that mod, i.e. Splawn. Also, Trainwrecks, Dumbles, and a host of other great sounding distortion amps have no CF. Et cetera.

[kg]

Guys, we are all looking at the same data. I just have a different opinion on the relative merits of using a stage as a cathode follower when you don't need the low output impedance.

that's fine. i have a different opinion than most people on the relative merits of having a 400hp subaru, when i don't "need" the power.

I’m totally up for a debate.

see, this is where i'm not following you. what is there to debate?

in fact, this ENTIRE THREAD has been very argumentative, with many experienced people (not speaking of myself, mind you) all agreeing, and a bunch of other people disagreeing.

i don't really give two flying shits if you believe me (or them) or you don't... it's a free country.

But first I’d like to wire up a simple objective test kg’s assertions about direct coupling.

so, objective tests... like steve already did? like rudd did?

what else are you going to objectively test?

I won’t have time until this weekend. Can it wait til next Monday before I pit my manly 35dB against your puny CF?

is this a contest? do you have something to prove?

if so, do i get to pit my 100# 600w power amp against your unquestionably smaller one?

what if we just take turns dropping our power amps on each other's feet? would that be a good debate?

In the meantime, a few things to ponder… After Leo Fender pulled the CF out, he produced the blackface amps, which are not tonally inferior to anything. When Marshall pulled the CF out, they produced the later JCM800s (Michael Doyle cites the 2210 as being among the best distortion amps of all time.) And speaking of Marshall, believe it or not, a popular mod in pro amps since the early 1970s has been to convert the CF into a gain stage. These days, people are even selling new amps based around that mod, i.e. Splawn. Also, Trainwrecks, Dumbles, and a host of other great sounding distortion amps have no CF. Et cetera.

and a lot of "great sounding distortion amps" don't even use tubes.

thank god it doesn't bother me.

[jmaf]

After seeing+hearing lots of different waveforms along time, you get to the point that hearing something evokes the waveform it *should* be producing and vice-versa.

Kinda like reading music in fact....it's just a different language. I'm by no means a master of the scope, but I've seen folks who stare at a trace and go directly to the faulty capacitor.

[jmaf]

About this thread, I had this saying in my head all the time, terribly translated : I don't believe in ghosts, but that they exist I'm sure.
(Yo no creo en espantos, pero que los hay, los hay.)

I was unable prove shit I said here but if you find yourself in front of a Bassman that farts, match the resistors, keep the 2nd grid negative, profit. (Ducks and runs.....)

[jmaf]

i don't really give two flying shits if you believe me (or them) or you don't...

You should be working in diplomacy, no kidding.

[Steve Conner]

Here are some thoughts:

Maybe the diode bridge gizmo in later Marshalls is meant to mimic the clipping characteristic of what we just called a GS/CF. I have tested these things, and the clipping of the high voltage diodes they use is very soft.

If you have a Bassman that farts, it could actually be the sound of Ken's Subaru from 5 blocks away. Every time I hear one of those things go past, I want one, but then I think of the gas mileage and the insurance, and the feeling goes away.

Or, if those two 100k resistors were badly mismatched, like one was 100 and the other was 235, I can see how it would sound pretty bad with high output pickups. And resistors can drift that far with age. But I don't buy the theory that they need matched better than the usual 5 or 10% of resistors straight out of the box, hence I object to the use of the word "matched".

Of course jmaf measured the ones he removed and can tell us what the value was.

[jmaf]

Steve, this is such a case where I wish folks were less spread around the world and we could meet up and the whole thing could be shown in minutes. That CF sounds beautiful on even overdriven state from the previous stages(on normal conditions, in a working amp) when the 2nd grid is biased at or more than 2V DC negative.

Now, with the unmatched resistors, I found from 80k to 110k in my builds and they fart, probably due to someting that's already been discussed here.

"Match"(whatever you wanna call it) them to a few percentage and they work a dandy, like I said my theory lacks me, but I think it's related to what Merlin said in his book about that stage self-balancing, the two triodes work together, the 2nd stage rarely drives the grid positive under "normal" conditions. You folks can probably theorize why, I tried.

Just build a(nother) 5f6-a and see for yourself, unmatch those resistors wildly, etc, test and see(better, listen!). You can probably help me/us understand.

[kg]

You should be working in diplomacy, no kidding.

you know, a decade ago, before two kids, before a mortgage, before the list of stuff that has to get done, every day, got so long, i would have spent more time, and perhaps had a more patient and forgiving attitude. in fact, i did just that. hence the "lifetime member" under the handle.

i do not think that gives me carte blanche to come on here and disrespect the board or its members.

but for god's sake, why is there so much competition? it didn't use to be this way. there was a lot more agreement.

the tubes are the same they were back then, and i daresay there isn't a damned thing new circuit wise.

so for the life of me i don't understand the attitude that people are here to feed you (not you, jmaf--speaking generally) bullshit and stroke their own egos. there are the same old names that have been here for a very long time, and they STILL provide little nuggets of wisdom, here and there--but only if you're a willing recipient. if you're not, YOU'RE the one who loses the opportunity--not them.

having a huge discussion as to whether or not you can achieve a pleasing tone without a DCCF is like asking an italian chef if he could cook without oregano. sure, of course he can. but then the question is, why NOT use oregano? because oregano costs money? because it's a waste of an ingredient, and you can only have 10 in a recipe? because you tried oregano 30 years ago when you were a kid and you hated it so you're never going to touch or taste it again? because the dutch don't use oregano, and their food is edible?

it's an utterly pointless discussion. if you don't like oregano, don't use it.. but don't then claim that oregano is worthless just because you don't have a use for it. leave it to those who like it, and appreciate that they may actually like the taste.

[Chuck H]

I hate the taste of cathode followers. But to be fair I hate the taste of all electronic circuitry equally so I need to remain diplomatic.

[jmaf]

kg, i picked on that line tongue in cheek really. i get where you're coming from.

i liked the analogy you used too, and if you think about it learning cooking and any other art for that matter is tough and it's knowledge that can't be passed forward so simply.

so, bringing to my personal experience, the only reason i "competed" with you earlier instead of just asking where you thought i was wrong was because of the way you approached me in the first place. i thought, this guy has four stars on his shoulder but he ain't gonna talk to me that way.... perhaps you get similar reaction in other cases, and i don't know if i'm totally off here, but that may be the case. having said that, i've noted how you contribute lots of knowledge when folks ask you, and i've come to admire the stuff you contribute back. but let's just say it took me a bit of a cultural adaptation.

as bill cosby would probably put it, i just just said all that just to really say that i didn't read your entire chat with woodyc and i've got nothing better to contribute to this other than my freudian post-beer analysis.

[Alan0354]

3) B.K.Butler Tube Driver:
As before, the first triode is zero biased, same considerations.
The second triode is positive biased through a 470K resistor to +12V

How does the 12AX7 work in this schematic? the first half of the tube has no cathode resistor, it is biased in zero bias!!! The grid of the second half is pulled to the voltage same as the plate. How is that going to work. can you explain this?

This is my guess, for the first half, even though there is no cathode resistor, but the internal resistance rk=1/gm can by like 1K. That is the one they use for dropping the voltge to set up bias. Is this true?

Thanks

[Merlinb]

How does the 12AX7 work in this schematic? the first half of the tube has no cathode resistor, it is biased in zero bias!!! The grid of the second half is pulled to the voltage same as the plate. How is that going to work. can you explain this?

This circuit runs at low voltage, so everything is not as it first appears.
The first triode is grid-leak biased. A negative voltage develops at the grid due to grid current, and I would estimate a bias of about -50mV in this case. This might not sound like much, but with only 9V supply the valve will only have about 600mV total input headroom!

The second triode's grid is also negative because of grid current, and is probably around -0.1V in this case. These figure will vary a great deal with the actual valve plugged in, however, due to manufacturing variance.
http://valvewizard.co.uk/Triodes_at_...s_Blencowe.pdf

[J M Fahey]

How does the 12AX7 work in this schematic? the first half of the tube has no cathode resistor, it is biased in zero bias!!!

So what?
"Zero bias" is maybe a way to name this, but it's probably confusing you because it sounds like "no bias", which is not the case.
This triode *is* biased ... meaning its grid sits at a definite voltage relative to cathode .
And such voltage is not random, but was carefully chosen by the designer ..... it just happens such voltage is 0 volts.

Quoting Enzo:
0 volts does not mean "no" volts, it's "0", a very definite value.

But maybe it has some negative voltage there by some other means? As in grid leak bias or something?
Well, not in this case.
To develop a negative grid leak bias you must resort to capturing some electrons which wander away from the cloud that surround the cathode, and happen to be physically captured by the grid.
I remark physically because there is no voltage that makes them go there.
This electron capture makes the grid become progressively more negative and if unchecked, would soon put the tube into cutoff, and in a random and uncontrolled way.
So, you need to provide a path to discharge that current to ground (where the cathode is also referred to), by means of a grid resistor.
How big?
Typically, 1 Megohm or less. Meaning: "with a 1 Megohm or less grid resistor you minimize grid leak influence on bias as to put it out of the picture".
Higher Grid Resistor values will add enough "extra" bias as to shift the operating point.
Or, as RDH4 puts it:

And ... what about *LOWER* than 1M grid resistor values?
Well, it's clear the grid leak bias generated will be so small as to be irrelevant.
And what about a 3.3 kilo ohm grid resistor?
You must be kidding !!!
Nay / nada /zero / zilch grid bias with those values. No way.
Ok, finished my cup of coffee. 08:15 AM here.
Will go finish an amp I must deliver today, will answer your next question during my next coffee break.

The grid of the second half is pulled to the voltage same as the plate. How is that going to work. can you explain this?

This is my guess, for the first half, even though there is no cathode resistor, but the internal resistance rk=1/gm can by like 1K. That is the one they use for dropping the voltge to set up bias. Is this true?

[imaradiostar]

I think it largely doesn't matter because that circuit sounds about a hundred times better when you add cathode resistors and bypass caps.

That's the "bias mod" now used by Eric Johnson and David Gilmour. It was included from the factory on the rackmount versions of the tube driver. I've owned both- I only own the rackmount version now as it sounds better and can still do the incorrectly biased sound of the original.

Jamie