Pushing 6p14p too hard

[torrence]

Hi,

I have a pretty standard EL84 push-pull ultralinear amp almost finished, but I am having trouble with the power tubes. I bought a bunch of 6p14p-ev tubes as I wanted to push these a bit beyond the standard EL84 specs. I am running them at 290 V on the plate in ultralinear with a cathode current of 50 mA. After a few minutes they develop a sporadic crackling noise which I presume is some kind of arcing in the tubes. It happens in both channels and for any of the tubes I try, so I presume I am pushing them too hard.

So before I start desoldering things at random, which parameter is most likely causing trouble? Should I try reducing the current first, or would it be more helpful to drop some voltage on the second screen (say with a zener diode)?

The total plate power is just over 14W, which is right at the limit even for the EV variant of this tube.

[kevinkr]

How about a schematic and some build details? In general pushing any modern EL84 variant including the 6P14P_EV to or beyond the plate ratings is asking for trouble.

Some details on how the output stage is wired and whether you are using fixed or cathode bias would be helpful. Are you using grid stoppers on the screens and control grids?

What is a second screen? (suppressor? Which incidentally should be tied to the cathode)

[vinylkid58]

Originally Posted by kevinkr

What is a second screen? (suppressor? Which incidentally should be tied to the cathode)

FWIW, the 6p14p data sheet shows G3 is tied to the cathode internally.

jeff

[kevinkr]

Originally Posted by vinylkid58

FWIW, the 6p14p data sheet shows G3 is tied to the cathode internally.

jeff

I didn't bother to check the data sheet obviously.. Still wondering what he meant by second screen, (control versus screen grid?) presumably he'll post soon with some further details and hopefully we'll get to the bottom of his issue.

[torrence]

Sorry, just meant the screen (second grid I guess it is sometimes called).

OK, I found my schematic. The weirdest thing about the output stage is the 5W zener diodes under the cathode resistor to try to eat up another 60V of B+. I know this isn't ideal, but this was the power supply I had, and the direct coupled input stage works really well at this voltage.

I think I will first try increasing Rk to lower the plate power dissipation. It sounds pretty good when I run this current-limited on a 100 watt lightbulb.

[kevinkr]

I think with some adjustment of the cathode and screen resistors you could run directly on your 340V supply without those 60V zeners.. I have an old Radio Shack amp here running 6P14P at 380V and while I am not totally thrilled by this so far over a number of years it has worked OK..

[torrence]

I guess I should just crack out a HV supply and see what these tubes can take. When I tried this initially on a breadboard, these tubes really didn't seem to like running from 340 volts, but I also didn't try dropping the screen voltage.

[Rod Coleman]

The Peavey Classic 50 guitar amp runs 4x EL84 (usually 6P14P ordinary type) at 10W per anode of dissipation. But the anode supply is well over 400V and the screen is at about 400V.

This is how it gets 50W (claimed) out of 4x EL84.

Would not be a design choice I would always make myself, but I can say that there are a large number of classic 50s around, as far back as the 1980s, and their reputation for tube-eating is not worse than other amps.

Either way, if you have crackling, then I suspect a faulty tube, or some other fault. Va= 360V, Vg2 = 340V presents no problems for the 6P14P, and 14W is within rating, too (just about).

"Second grid" is the term used for the Screen in Russian data sheets, because it is a literal translation of the Russian term for it: сетки второй.

[Depanatoru]

You are pushing too hard those 6P14P . From my experience the crackling noise is generated by metal expansion inside the tube ( G2 or maybe anode ? ) when power dissipation is too high

[torrence]

Thanks to everyone for the comments. Based on the replies, I increased Rk to 680 ohms and have bypassed the Zener diodes. The tubes seem pretty happy running at Vpk = 360 volts with Ik = 22 mA. Along the way I also found a bad solder joint, so that probably wasn't helping. I have attached an 'action shot' to show the whole ugly mess in progress.