Experiences with 3DG4s?

[Rob Mercure]

Folks,

I was rummaging around the "whatever" tubes in the attic and again noticed the 3DG4s - really hadn't paid much attention to these bottles since I worked on TVs in the early 70s. But checking out the specs I noticed that they have greater current capacity than 5U4GBs with lower voltage drop. But they have that 3.3 volt filament at 3.8 amps. Hmmm, but that's only 0.8 A more than a 5U4 and I suspect that some trannies are robust enough to pull that extra current so you could wire a socket with pins 4&5 tied, pins 6&7 tied, pins 1&8 tied with a 0.45/10W resistor between pins 2&3 creating a socket that would, with sufficient filament current, accomodate the 3DG4 or a 5 volt rectifier.
Aha, of course there is a "rub" - my manuals state that pin 2 is "internally connected" - but to what? The four valves I've got have a thick internal wire going from pin 2 to the bottom mica and ending there unconnected. Does anyone have any idea why this is or why it would prevent a dual tube/voltage socket?

Rob

[Enzo]

Did you do resistance checks to the other pins?

One concern I mave right away with those socket pins wired together is the reason the pins are left off of the tubes in the first place. They want that space around the HV AC leads to reduce the likelihood of arcing. You double up those pins and now that HV is right next to the other one. 700-800 volts less than a quarter inch apart.

And that internal pin, even if it is not connected directly to something, it might be close enough to something else that if it got near ground or something potential for arcing is high.

[Enzo]

You seem to be rummaging around quite a bit today...

[Rob Mercure]

That's what I was concerned about

Hey Enzo,

You've exactly echoed my concerns. As to the close pins a dollop of silicone glue or even hot melt would probably prevent arcing between the pin solder connectors and perhaps a small slit between the pin holes on the top. While I realize the 700VAC is high some UL designs have 300VDC between adjacent pins (screen and plate) not including whatever AC swing is present.

But the kicker is the "internally connected" pin. Yeah, if you looked at a bottle the internal wire is about 1/4" from each plate tab as it penetrates the lower mica spacer. I've got atmospheric charts somewhere that correlate air temperature, pressure, humidity with the distance that various frequencies can strike an arc. But I don't have anything for a "vacuum" - whatever that means in this case cuz I'm not sure how drawn down the rectifier vacuum is (pretty "hard" I would suspect though) so I don't know if 1/4" is "close" or not (no nice sharp points to arc between though).

Just trying to determine if these bottles are worth storing or trading.

Rob

[Rob Mercure]

More info - as long as I'm "talking to myself"

Having looked at other 5 volt rectifiers I've noticed that all of the ones close at hand - 5U4s, 5AW4s, 5AS4s, etc - have an additional support wire connected to an external pin (pin 1 on 5U4) that either barely enters the lower mica or actually protrudes 1/4" or so. In fact the only real difference I can tell between the 3DG4s and these tubes is that the wire - for support I suppose - in slightly closer to the plates than of the 5V tubes and I assume that this account for the "internally connected" designation in the tube manuals.
While I'm still not sure of the practical consequences I'll make a stab: I suspect that my substituting scheme might work for lower plate voltages - say 350 VAC per plate - and lower frequencies (but when is the last time you wired up a rectifier for a 400 Hz AC supply?) and one might even get away with it at higher voltages. Unless the vacuum was "polluted" by gas emissions from the tube elements or the tube was leaky the spacing between the wire in question and the plates is still farther than even a 1KV spark should be able to jump in a vacuum! But, further caveat, I would only use the tube in an upright position (but, again, when is the last time you wired up a rectifier horizontally?).
Guess I'll just have to try it someday and see.

Rob