Request: Circuit examples with EL34's pin 1 disconnected?

[jmaf]

Do any of you know of any successful circuits which leave the suppressor grid disconnected on EL34 based power amps? (Not connected to the cathode OR any other negative voltage, nothing, pin 1 is just left floating there).

Do you know of such circuits? Can you suggest some?

One last question: has anybody accidentally tied pin 1 to the screen on a EL34, such as when you use it as a contact bridge for 6L6 style tubes that have no pin 1? What happens when that grid is at high voltage? Mushroom cloud? I've never blown a tube up like that and I'm curious.

[Amp Kat]

Certainly a good question and the main function of the suppressor is to catch electrons from the plate that stray away from the plate and screen. The Suppressor should always be negative with respect to the anode. In either case of not connecting anything or sending positive voltage would make the tube very unstable. Not sure it would blow up with high voltage but would basically put it at screen grid potential with no effect of supression. Using a negative volatge is said to make the tube much more efficient or linear. I 'm sure you could do a gm chart and use a graph to show the curves with decreasing negative voltage. I believe Traynor used to send about -15 volts to his Suppressors to make them more efficient and believe a while back someone posted here about the tube performing very well and good tone with a negative voltage to pin 1. Try it and see what happens as you really have nothing to lose.

[jmaf]

Thanks for that, Amp Kat. I'll experiment with old el34's i have lying around. On one of KOC's books I read that Traynor had improved el34 tube life and reliability on the road by using a negative voltage on pin 1 while still keeping very high voltages on the plates. I've never heard about experiences in the opposite though, accidentally making it very positive.

[kg]

disconnecting grids is generally frowned upon.

however, when i have done it it has not appreciably changed the operation of the tube... since absolutely no current may flow in or out of a floating node, it tends to "self regulate" at some arbitrary voltage. the more positive it gets, the more electrons hit it, which makes it more negative, and then less electrons hit it, etc.

in the case of a disconnected g3, you will end up with a tetrode. they work fine, until vg2 exceeds vp, and then ig2 goes through the roof.

the same thing will happen with g3 and g2 tied together--tetrode operation, with it's attendant issues of secondary emission causing a kink in the plate curve.

[jmaf]

So a "real tetrode"(before beamed ones existed) would be more like leaving pin 1 floating, instead of hooking up g2 and g3 together, right? As the latter would put the suppressor at a high potential?

Is there a trace somewhere showing the kink on the plate curve, BTW? I've always been curious to see what it looks like.

[Jazz P Bass]

Tetrode Plate Curve

DF97 Pentode and Tetrode curves

[jmaf]

DF97 Pentode and Tetrode curves

Awesome post, thank you.

[clyde1]

I built a Trainwreck Express once upon a time and failed to wire pin 1 and 8 together. The amp had an annoying parasitic rasp at higher volumes on harder pick attack. Solved by correct wiring.

[Amp Kat]

That is what I would think would happen is you would get all kinds of intermitten and freaky noises going on at different times.

[kg]

That is what I would think would happen is you would get all kinds of intermitten and freaky noises going on at different times.

yes, it's basically impossible to predict exactly how it's going to react when it's floating.

[raiken]

I had one of my amps on the bench that was putting out low power. When I looked at it on the scope, I would see a "jump" or "step" in the the output signal level when I reached a certain point in the volume level. It turned out pin 1 on one of the output tubes was accidentally left floating, and when I soldered it, the problem was fixed. Apparently, some charge builds up on the floating element and causes strange behavior. Don't leave it disconnected.

[J M Fahey]

I think you may make the experiment yourself and find the actual effect on plate_current vs. grid_voltage vs suppressor_voltage
You may set up a PSU with, say, 300 or 400V on the plate and an ammeter in series with it, a variable 0 to -38 or -52 or whatever negative voltage sends that particular pentode into cutoff (so you cover all standard grid voltages) and a switch which either leaves the suppressor grounded, floating, with -15V or +15V.
The grounded switch position will match published tube curves; all 3 others will provide the answer you are searching.
Maybe you needn't do it 100 times, perhaps you are lucky and, say, with 10 or so test points you get a clear trend.
Please post results here, they will be useful for all of us.

[Wilder Amplification]

IF you decide to try B- on the suppressor grids, DO NOT USE THE BIAS SUPPLY AS THE B- SOURCE!!! Derive a 2nd negative voltage source that is dedicated to the suppressor grids.

Reason being...and I've had this happen...suppressor grid short to the heater (after all, the pins are right next to each other and god only knows how close the internal connecting elements of the two are inside the valve). Your heater CT is referenced to ground while your suppressor grids are referenced to the bias supply. If the two happen to short, there goes your bias supply and it's red plate city.

Yes I have had this happen before. Trust me...the "Traynor did it and they didn't have any problems" argument is no excuse. Learn from my mistake...if you're going to do it, build a 2nd negative voltage source that is dedicated to the suppressor grids.

[jmaf]

I just caught up with the discussion, thanks everyone who replied.

clyde1, Amp Kat, kg: Gotya. So a floating pin one then may do random stuff, develop a charge, pick up AM, modulate, whatever... I know that G3 is able to modulate the plate, some folks apply negative feedback to it, so it'd make sense for it to pick up random noise and modulate the output with that noise.

J M Fahey the idea of switching different potentials into it and watching the plate seems to be a good one. I know for a fact that G3 is able to modulate the plate, so switching between -15 0 +15 VDC, etc, will modulate the plate somehow, but it'd be interesting to find out how much exactly. My initial curiosity was whether applying 450 VDC screen voltage to it accidentally will blow up the tube.

raiken, Jon Wilder thanks very much for sharing. I was thinking of using bias for the negative rail, that would've been stupid and I had no clue about that possibility.

Thanks again.

[g1]

IF you decide to try B- on the suppressor grids, DO NOT USE THE BIAS SUPPLY AS THE B- SOURCE!!! Derive a 2nd negative voltage source that is dedicated to the suppressor grids.

Reason being...and I've had this happen...suppressor grid short to the heater (after all, the pins are right next to each other and god only knows how close the internal connecting elements of the two are inside the valve). Your heater CT is referenced to ground while your suppressor grids are referenced to the bias supply. If the two happen to short, there goes your bias supply and it's red plate city.

Yes I have had this happen before. Trust me...the "Traynor did it and they didn't have any problems" argument is no excuse. Learn from my mistake...if you're going to do it, build a 2nd negative voltage source that is dedicated to the suppressor grids.

Jon: Is this common? More common than a grid to heater short? Can you see where I'm going with this? Sorry to be a smart arse but I think you may be going a little overboard here.