testing 6DJ8s

[hailteflon]

When testing some used 6DJ8s I would occasionally run across one that tested good for transconductance but then the needle would gradually fall into the bad zone.

Emissions check strong on this tube.

What is wrong with this tube? Mark

[SY]

Try putting a scope on the tube while you're testing. It might be oscillating as 6DJ8s tend to do without the proper care and feeding.

[Geek]

Them little suckers also get their grids coated with junk that messup the reading. I've had 6DJ8's that tested like that, so I ran them on the bench overnight conducting 15mA at 90V and they tested and sounded like new the next day.

And if they are on their way out, that test will help them along to the land of "forever glowing filaments"

Cheers!

[hailteflon]

At first thought I would say that an oscillation would come on suddenly. It takes about 20 seconds for the needle to bottom out.

But, this tube (Amperex PQ) was unusally hot so this fits with oscillation. Possibly the oscillation gets worse as the tube heats up.

Releasing the button and starting the check again repeats the original results. It starts out with a good readings and begins to fall.

The question is, why is it oscillating? This is a B&K707 with a real mercury rectifier and good 6BN8 (may be wrong number).

The grid may have some junk on it. I did the conventional overvoltage to the filament for 15 seconds which sometimes helps, but I'll have to repeat this tube....can't remember. I had more than one that did this. Also, no grid emission detected.

If I can figure out how to get a scope lead on it I will post the results. Mark

[hailteflon]

I found a Pamona 1449 9-pin test socket. Just treated it with TarnX. It will take some time to dry it out under a lamp so no fireworks occur.

How does passing 15mA through a tube for several hours clean the grid. Is it literally a scrubbing action by the electrons?

[Tom Bavis]

I have made an inline adapter with 1K resistors in the grid leads... and 6DJ8s test MUCH better.

[hailteflon]

Thanks for all the help folks

Tom, do you mean series a 1K with the test lead to the scope? Or between the tester and the grid connector on the adapter?

Also, Geek said he cleaned the grids with 15mA. Was this d.c.? Seems easy enough to put a load resistor from the plate to B+.

This test socket should be done in a couple more hours and then I will see if I get any oscillations with the scope leads connected to the plate and cathode.



Mark

[hailteflon]

Test results: I found no evidence of oscillation. +/-50V on plate-cathode and also +/-50V on grid-cathode with test button pushed. The scope trace was a very sloppy sin-wave that had very little reduction in amplitude while meter reading was falling. The frequency was approximately 120Hz. This is a transconductance test.

The readings did not fall as low each time with continual retesting. I held the button down for about a minute and the reading was up to mid-scale on the meter. The needle was actually rising back up. I didn’t know if continual biasing will damage the 707 so I stopped. The original low point was 26 on the meter. It was up to 50 when I let go of the button.

Looks like Geek is right. The grid needs a little sandblasting with electrons.

[Tom Bavis]

I had trouble with 6DJ8s oscillating in a Triplett transconductance tester - made a 9 pin adapter (socket to 9-pin plug) wired straight through except for grid pins, which have series 1K resistors. Oscillation was obvious - reading would change as you move your hand around the tube.

[hailteflon]

Could you see the oscillations on a scope?