valve testing el84 only

[bob orbell]

Just to let Phil and all know you are not alone with this valve problem, our member KC Jones brought his KAT 6550 amp to me last night with the very same problem, one valve red hot and the cathode resistor burnt, a quick fix though. BOB

[Richard]

How old were the valves in that amp Bob?

[Richard]

Originally Posted by bikerhifinut

Out of curiosity I put the stereo20 on the oscilloscope this afternoon.
On one channel the waveform was a complete mess.
After a process of elimination it came down to a bad valve on a matched pair of the edicron EL84's (6P14PEV). But before we jump to conclusions about valve quality, they are over 10 years old probably 12 to 13 and have had a fairly hard life.
But the voltages were checking out ok, no cooked components underneath.

food for thought?

And the odd thing was I never noticed a problem and the amp was distorting badly from around a watt output and that affected channel couldn't even deliver 5 watts flat out. I think a phase was down. a straight swap top the other channel confirmed the diagnosis.

A.

Had similar with a KEL84 Andy! It sounded fine at lowish levels but distorted on one channel when it was wound up. Scoped it and found it worked fine up to 4W on the offending channel. Removed each valve in turn and found one wasn't connected - it was running as SE - but the waveform looked fine up to 4W! Isolated the problem to a small area of the pcb by flexing it and watching it cut in and out. Never found the exact joint but soldered a bunch of joints in the area and got it

[bob orbell]

Originally Posted by Richard

How old were the valves in that amp Bob?

I will have to check with Kevin Richard, but I would say at least 5 plus years. Been doing some voltage tests on the amp this morning, and all 4 coupling capacitors are leaking, C11-14, they are Jupiter Beeswax, I am replacing with Panasonic polyester. BOB

[Richard]

Originally Posted by bob orbell

I will have to check with Kevin Richard, but I would say at least 5 plus years. Been doing some voltage tests on the amp this morning, and all 4 coupling capacitors are leaking, C11-14, they are Jupiter Beeswax, I am replacing with Panasonic polyester. BOB

Ah, good to find the reason Bob, I stopped using PIO for coupling/blocking after reading firstly Noel's comments in the WAD 300Bpp article and then talking with a well known parts and equipment supplier who said that, even with good ones, a leakage up to several hundred mV when used in such a position was to be expected

[bob orbell]

I was reading almost 10 volts on one of them, I have never used PIO in my own kit, just good polyester, or polypropylene.

[Greg.]

Hi Phil, I'm not sure if this is the thread of choice because you have been posting prolifically recently, but I'm sure you will read this.

If you need new EL84 valves, have a look at this sales thread.

http://www.audio-talk.co.uk/phpBB3/v...hp?f=16&t=6294

[Phil Y]

Hello Greg,

Thanks for spotting that but my plan is to stick with JJ for now. I used JJs for about 6 years with no issues then decided to try some other brands to see if I could find something better sounding. Basically, with all the other brands I have tried, I have had at least one of the four valves fail after between six and twelve months. I have JJs back in now so I will see if I get a period of reliability.
If not, then I shall have a closer look at the amp.

Regards, Phil.

[bikerhifinut]

You've either been unlucky with Valves or there is something in the amp.
I think it's the former if all voltages check out and there's no leakage past the coupling caps. Did you check the coupling caps leakage with the output valves unplugged? That should be the definitive way. Apologies for teaching granny to suck eggs.
As I may have said before, I'd check what's happening at the Screen grid (g2) of the output valves as you have done a pure pentode mod. Now a bit of digging around has got me wondering if at times the valves are drawing up to 57mA at full chat into ClassB? This would be quite normal in certain configurations. But I think the cathode bias prevents that situation happening?
I would still check that particular bit of the amp.
Can anyone with the knowledge here tell me if the bypass capacitor to earth on the screen grid feed to the output valves is necessary here? It's just that you seldom see it on commercial designs, they usually just feed the grid via a resistor after the PSU filter (choke). What springs to mind here is perhaps a Zener or something to keep the voltage differential from the anode constant?
Just as a postscript, I "Checked" all of my EL84's using the scope and one channel of a SE amp and found a couple of duffers. One ancient "pull" mullard and one of my quad of Russian 6P14PEV. But they're well used and it also may have suffered a bit of abuse during a project build/test.

A.

[Phil Y]

Andy, as I have already said, 3 of the 4 times I have had valve failure was before I put the amp in pentode mode so I see no reason to suspect this. Also, the reason I did it this way was basically because all the designs I looked at online have the screen feed not taken from the same point as the main feed to the o/p Tx. Normal practice seems to be to take it from a more smoothed, slightly lower voltage point, further down the R-C filter chain. I was just trying to copy what seems to be standard practice. I did consider "splitting" the 2.2k dropper in the HT to create a feed but with a PCB design I decided this would be a bit of a messy solution.

http://www.r-type.org/articles/art-003e.htm Scroll down for circuit.

And this one: http://www.drtube.com/schematics/vox/ac156202.gif

I am not a designer, just a repair technician, so I am unable to argue the finer points with you. My plan for now is to leave the amp in use for a couple of weeks and then lift the bonnet again and re-check all voltages, re-test the valves, and see if anything has changed.

Regards, Phil.