EL84 Amp Design...Why Such High B+ Voltages?

[archaos]

Maybe one of the speakers died and the load impedance is doubled.

Actually no, both speakers work correctly.

[stormbringer]

The Mullard data sheet (1961) specifies the following limiting values:

Va(b) max = 550V
Va max = 300V

Vg2(b) max = 550V
Vg2 max = 300V.

I don't know the meaning of the parenthetical (b), but it didn't appear to be a footnote. Can anyone enlighten me?

Here is the URL (sorry I couldn't add an image for whatever reason):
http://tubedata.milbert.com/129/e/EL84.pdf

[Enzo]

Rather than put footnotes on every page, they probably have a chart of them at the front or back of the book.

Just a guess, but I;d have to say the 300v is the max design voltage they recommend in a circuit. The (b) voltage is perhaps the max voltage they want on it period. In other words design it to run on 300v, but it is OK for it to sit there with as much as 550v on its plate.

[ThermionicScott]

I believe the (b) means at zero current, or as close to it as possible.

[Gingertube]

A few "shotgun at the barn door" comments (in other words "All over the place like a mad womans breakfast"):

Traynor tied G3 to the raw negative bias supply rather than "elevating" it as suggested above. That approximately halves the fault current if the bias applied to G1 disappears. The HiFi designers are recommending this arrangement for true pentodes.

Tube ratings which appear on the old datasheets were "design centre" and you can often use them at higher voltages than are listed on the datasheets.

Depends how you use them.

6V6 for example were used as television Vertical Deflection Amplifiers at 1200V on the anode (yes twelve hundred volts) BUT 15% duty cycle maximum.

I run JJ EL84s in Ultralinear on one of my HiFi Amp designs at 360V, never had a problem with that (270 Ohms series screen resistors) but when using the same tube in Guitar Amps I use 1K 5W screen resistors minimum, because when you start overdriving (with fixed G2 volts rather than ultralinear), then screen current goes up quite dramatically as the anode voltage drops below the screen voltage and the screen wires overheat and can "sag" so that whether they were originally aligned to the control grid wires or not becomes a mute point. In fact you can actually melt the screen wires if you abuse them too much.

Having said all that my design philosophy for guitar amps these days is to run lower voltages and higher idle currents to maximize gm and minimize rp. I've found that I just like the way they sound that way and am not fussed about sacrificing 20% of output power to achieve the sound I want.

Example:
I run EL34 at 380V which gives me 35 Watts out, compared to the more usual Marshall etc. voltages of 100V more which gets another 15 to 20 Watts out of them. I just don't like the sound of those extra 15 or 20 Watts.
I run 6V6 at 285V (data sheet value) for about 12 Watts output. Yes you can screw another 6 to 8 Watts out of them by running high voltages (but claims of 25W from a pair were just marketing B.S.). I'm a bit conservative with this as I mainly use vintage ST shaped 6V6G although I have some late 1970 Russian 6V6GT which sound almost as good (the box they came in still had the KGB sticker on the side).
I like to keep EL84 at 350V max.

Remember that to get twice the sound level, you need TEN times the amplifier power OR (more sensible option) twice as many speakers with the same power.

Cheers,
Ian

[Steve Conner]

Traynor tied G3 to the raw negative bias supply rather than "elevating" it as suggested above. That approximately halves the fault current if the bias applied to G1 disappears. The HiFi designers are recommending this arrangement for true pentodes.

Can you explain how this is possible without also halving the power output?

[Steve Conner]

Traynor tied G3 to the raw negative bias supply rather than "elevating" it as suggested above. That approximately halves the fault current if the bias applied to G1 disappears. The HiFi designers are recommending this arrangement for true pentodes.

Can you explain how this is possible without also halving the power output? I can't: loss-of-bias is the same condition as peak signal current.

[Gingertube]

Steve,

Good Question - sorry I don't have a good answer. Better treat that claim with "a grain of salt" or just assume I was wrong. It certainly would'nt be the first time and there is a high chance it won't be the last.

Cheers,
Ian

[pdf64]

Maybe the loss of bias, ie bias reference, could result in the control grid being pulled significantly positive wrt the cathode (might secondary emission contribute to that)?
That may cause higher current than a peak signal condition, at least 'non grid current' drive arrangements.
Pete

[MZOONY]

This is a good thread

I just built a 4 x EL84 cathode bias amp, and have the same issue where the tranny I am using (as that is what I had available is 350-0-350 where the design calls for 290-0-290
I tried the zener diode in series cathode to ground via the CT but it made no change i teh voltage
Anybody tried that and had that issue?

[Chuck H]

This is a good thread

I just built a 4 x EL84 cathode bias amp, and have the same issue where the tranny I am using (as that is what I had available is 350-0-350 where the design calls for 290-0-290
I tried the zener diode in series cathode to ground via the CT but it made no change i teh voltage
Anybody tried that and had that issue?

Flip the diode.

[MZOONY]

Thanks Chuck

I tried that thinking I was missing something
I used two different types
CZ5352B
and
1N5352BRLG
Zeners all 5W 15 v

Tried going one at a time to see a difference to the 350-0-350 but I never see a change
I am baffled

[MZOONY]

I did further tests to see what the issue might be and this is what I found
As these diodes are 15V I noticed that when I measure (I have the diodes on a terminal board) from the first diode I notice a drop in voltage, and subsequently keeps dropping voltage as I go to the next diode and so on
The problem is when I get to to the last one which is connected to ground the voltage than measures back to the original 350-0-350 of the PT
So what I gather because of some other part of this amp that shares the ground it does not let me get the drop in voltage

[g1]

Tried going one at a time to see a difference to the 350-0-350 but I never see a change
I am baffled

I'm not sure exactly where you are measuring, try measuring the DC voltage on your first filter cap.

[MZOONY]

Thanks G-one for the reply
Am I missing the the whole point what I am measuring?
The way I understood it is by putting the zeners in line with the CT and if I have on the secondary originally (e.g.) 350-0-350
I should be getting less (e.g.) 300-0-300
Am I missing something?
Just tried measuring from the first filter cap and without zeners I get 348 and with zeners I get 335
does that sound right?