KT90 Arcing Internally

[tweedeluxe]

Firestorm, I had considered doing just that, but I think the problem may have been a faulty tube the whole time.

I changed the screen grid stopper to 3k3 and changed the leg off the B+2 node (after the choke) to 5k. This brought the idle screen voltage to around 50-60V below the plate.

However, the tube was still arcing, even at low volumes (had to turn out the lights to see it). I'm not sure, but I think the plate was arcing to one of the filament pins.

I swapped out the tube and everything is fine. No arcs, even pushing the output stage really hard. I don't know if the tube was just bad or I damaged it before, but at least I have a working amp now!

Thanks for all the replies!

Lowell

[dartanion]

Remember that in SE designs, you are conducting max current at idle, so high volume/low volume situations shouldn't matter, unless it's a physical issue with the tube (rattling the innards at high volume).

FWIW - Most new tubes have horrible quality control issues. So, chock this up to a bad tube as it appears the other KT90 is working fine.

[Andy Le Blanc]

Bad tube.....

So far the electro-harmonic products have been good in my experience.
Every now and again Ill get a bogey, but they end up as bench spares, or ...
sacrificial goat....

I really like the EH KT90 so far, good tube, nice tone, and it is a new tube in relative terms.
The amp I built around them is very nice, reliable, great tone , power etc...
I would like to hope that manufacturing is just as reliable, but its a numbers game.

The more we shoot our mouths off, try different projects that demand tubes.

The amp your describing sounds like its not too far off from a HiFi.
What sort of pre are you running? feed back?

[Merlinb]

Remember that in SE designs, you are conducting max current at idle, so high volume/low volume situations shouldn't matter,

That's not true for the screen though.

Once you get an electrode (nearly always the screen) to arc, then the damage has usually been done- it will keep on arcing. Hopefully it won't happen again!

[dartanion]

Remember that in SE designs, you are conducting max current at idle, so high volume/low volume situations shouldn't matter,

That's not true for the screen though.

Once you get an electrode (nearly always the screen) to arc, then the damage has usually been done- it will keep on arcing. Hopefully it won't happen again!

Oops. I meant to say plate current.

[tweedeluxe]

The amp your describing sounds like its not too far off from a HiFi.
What sort of pre are you running? feed back?

It's my first go at a high-gain amp, sort of a blend of a Komet/TW preamp. I kept the 100k/10k stage and used 1/2 a PI, eventually settling on a 100k/5k NFB/presence from the 4ohm tap.

Instead of a split-load Rp on the 2nd stage (a la Komet), I used a pot underneath the preamp board, with a series resistor on "top", and coupling cap coming off the wiper. It's actually quite functional.

I'm still tweaking the amp though. I tried a 6L6GC tube, and it actually sounded much better than the KT90 (to me). Going to try an EL34 next. I have enough room to drill another octal socket so I might try parallel SE too.

The preamp needs some work - I'm thinking of dropping a stage or two, and maybe trying different tubes. It sounds okay, but it doesn't sound quite as rich or full as I'd like, and it lacks some sustain.

I used 3 6AV6 tubes (first two in parallel) and a Russian 6H2pi tube for the 100k/10k stage and the faux-PI circuit. I'm definitely going to take out the Ruskie and replace with a 12AX7.

Thanks again for all the responses. Here's an early gut shot (the power supply/VVR board wasn't mounted yet and I have since made many changes):