G2 mod gigs well

[Structo]

Hmmm, I wonder if this is the source of the common complaint against the Dumble amps, that fizz we try hard to eliminate.

[Andy Le Blanc]

your not missing anything, I got lucky... anything less than 220k cooks
up the wattage...

It could be the dumble issue, the fender design is the common starting point
since the mid 60's if it has a fender power side it could very well be

I got on to this when I heard a funny little distortion in an otherwise textbook
fender based circuit, I've had experiences dialing in fenders with their owners
to avoid the fizz you had to bias colder than you would expect
biasing hot would drop the plat voltage relative to the grid...

I got poking around in my amp, and even though everything was 5x5
the amp had a little fizz at softclip

I got poking around on the inside and found that I could grab a v difference
between the screens and plate and that the relationship changed with the bias

I've been looking to fill in the gap ever since,... what, why, how...
The design recommendations for the "new" beam power tubes
only pop up in a couple places, the g2 are slightly positive to keep the
virtual suppression grid in the technically ideal place, you see those
old schematic with the g2 tied to the ct of an opt with no resistors
this is why, but old gear was never ment to be pushed into clip
distortion is a relatively modern taste. you see it addressed by some
makers but its usually a series resistor, you couple this with DIY
misinterpretations of "bias" and your cooking tubes thinking you got good tone.

this nothing new, I bet theres some HI FI guys laughing in their sleeves

but you have to build a good argument, get an easy DIY measurement
try it out, give it a listen...

[David Root]

Thanx Andy! I think I'm going to put this into my 183 based build I just started. And I think I'll go back & put it in my #124 too becuz I had the fizzies for a while with that one. As well as the EH150, which definitely needs it.

[ampdoc1]

I put a screen drop switch on one of my amps about 6-7 years ago. I did it a little simpler by switching the screens between my B2 (310VDC) and B3 (260VDC) nodes. Works perfectly for dropping the output noticeably, without any degradation in the tone. Course, I was probably just lucky that the drop worked out well at that point.

I don't know the math, but the amp gets regular use and is still going fine.

a'doc1

[markr14850]

Thanx Andy! I think I'm going to put this into my 183 based build I just started. And I think I'll go back & put it in my #124 too becuz I had the fizzies for a while with that one. As well as the EH150, which definitely needs it.

It seems like the balancing resistors on the series capacitors in the Dumble precision power supply could be adjusted to achieve this effect without needing to add any new components.

[David Root]

Please expound on that. You are talking about tightening/loosening the regulation by changing the balancing resistors, right?

So if one were to lower the balancing resistors on the screen node caps to increase current to ground and reduce screen node voltage that would accomplish the same thing?

That would be loosening the regulation I think. Tone/feel-wise is it the same?

[jelle]

Thanx Andy! I think I'm going to put this into my 183 based build I just started. And I think I'll go back & put it in my #124 too becuz I had the fizzies for a while with that one. As well as the EH150, which definitely needs it.

It seems like the balancing resistors on the series capacitors in the Dumble precision power supply could be adjusted to achieve this effect without needing to add any new components.

not really/.. from the first post:
'bleeders between screen grids and
ground at the tube socket, after the typical 470r screen grid resisters'

Jelle

[Tone Lover]

Guys please bear with me I know newbies are a pain Im one, so were talking referencing say a 220k resister to ground after the 470ohm resister off of pin 4 of a 6l6. to change g2 below posative voltage by a volt or so . Its late Im tired hope I made sence.
Thanks Bill

[Andy Le Blanc]

yes...

and no newbie worries, I've been at it since well before this interweb thing
it used to be much harder to find reference or people who had even half a clue.

you need to look at a couple measurements

1)... bias...The method I've used is to disregard the tube and use the bias
to set the current at idle, at around 50% of the power transformers current rating.
I set this and maintain the current in the amp at this known setting for all
other measurements, this removes a variable, it gives you a base or foundation and a stability
for all other current and voltage readings.

set and maintain a bias point for all your readings
OR..
take all your readings at the same bias point

2)... P and G2... measure the voltage DIFFERENCE between the plate and
screen grid, with a 6L6 its pins 3 and 4 at the tube socket, plate is pin 3
and g2 is pin 4.

The voltage at G2 controls a virtual hump in the electrical gradient thru the valve.
The "Ideal" is to have the G2 slightly positive (50- 500ma/mv +) relative to the plate...

so if you have the positive lead of your meter on pin 3 and the negative
on pin 4 youll see a negative reading...

this gets the most efficiency out of the device but has consequences when the amp is pushed,
plenty of volume but a hard tone and fizz

3)... series resistor issue... when you push a amp the G2 will pass current
screen grid resistors are there to buffer the screens from excessive current
the problem is that with little changes in current you get big changes in voltage
so when you push the amp you get voltage swings that will cook the screen
And you can't really known what you effecting when you swap around
screen grid resistor values, you can't measure across the resistor without interference
and must have a known bias point to work from or the readings are bunk.

the best resolution is to be able to set the voltage difference between G2 and the plate.
the better way to do this is with a voltage divider

The simplest way so far appears to be installing a bleeder after the common 470r in older fenders and fender based circuits

You can get even better control with a divider before the screen grid resistors
Or a G2 voltage source, this is done in HI FI with regulated screen supplies
but you can also do it very simply in a guitar amp, the point is to change
the relation ship between the plate and G2 keeping the plate slightly positive
to avoid screen grid issues and better the tone of the amp.

you CAN measure the difference between the screen and plate
and you can CONTROL that difference to "effect" with a simple bleeder
or use a simple voltage divider to source the screen grids

I've done this with some variation just noodling on the bench...
a simple bleeder at the screen, you can also bypass the bleeder
with a filter cap or even a large value bypass etc...

you can provide an extra PS node to source the screens,
with a filter cap and bleeder after an extra resistor from the power rail that can be used to
"set" the g2 voltage difference any where you want, just like a bias, you can measure and you control.

[Tone Lover]

Ok sounds good what are my parameters in other words where is a good place to start with say resisters and caps. use 220k and 22uf cap Im just guessing you never discussed caps in all the stuff I have read. Sorry for asking so many questions.
Thanks Bill
PS thanks for explaining it in a way thats easy to understand

[ChrisM]

You should be able to just put a 220K 2W resistor from pin 4 to ground and be good to go.

[pdf64]

I remain to be convinced that this notion has any validity.
What's this 'fender fizz' anyway?
My view is that Fender AB763 etc design is fine, though as with anything, the devil is in the detail; most pertinently, as many SF Fenders demonstrate, the lead dress.
So an SF with bad lead dress may be on the verge of stability, such that hot bias brings on parasitics.
Same but more so with inexperienced dumble implementations.
But a well implemented AB763 is about as good as it gets and I just don't recognise the fizz issue.
I fail to perceive the relevance of quiescent screen voltage to 'large signal' conditions; guitar amp power supplies are the antithesis of regulated / stabilised supplies - apply a significant signal and everything changes. Apply a large enough signal and even holy cows like bias become irrelevant.
So, to sum up;
a/ there is no problem
b/ the rationale provided to justify the idea, re conduction angle etc, is nonsense.
Re screen grids, I have concerns regarding the screen grid dissipation that classic designs put on modern production tubes under overdrive conditions; but the additional 220k resistor suggested has an insignificant impact on this.
Pete.

[echuta13]

So you have experimented with this and didn't find that it had any relevance Pete?

[pdf64]

Yes, for the sake of experimentation I gave it a go. The amp sounded fine either way, I couldn't tell any difference. At quiescent it causes the plates to run a fraction cooler. With a signal I couldn't measure any significant difference, certainly nothing that wasn't within the realms of normal line fluctuation.
To my mind, that backs up the analysis that such a minor amendment to the quiescent conditions has negligible impact on large signal operation.
Also what's with the 'biasing to 50% of PT rating'! As if there aren't enough 'inventive' biasing methodologies already.
Actually I prefer the tone when biased to 60% of PT rating, but maybe that's just my ears adjusting for the increase in line voltage over the years. Pete.

[Gaz]

Pete,

I did quite a bit of experimentation with Andy's set up as well, and noticed no difference in tone personally. I was interested in the setup myself because I've had far too many screen failures in non-master amps with the typical PS arrangement, even with large screen stoppers (well, up to 1k), and the voltage divider seemed like an economical idea. For me, I couldn't get enough voltage drop to make it worth while without having to use very high wattage resistors to dissipate all the heat. It then became just too impractical, and active screen regulation began looking more and more economical! Ironically, I think the most economical solution is just letting the tube die - they are pretty cheap afterall.