Inter-Cathode Leakage

[Trevor Haddox]

I have a B5750 that has weird leakage between the 0 and 2 where one will dimly light the other. It's still readable, but annoying. There's no continuity between them (while in circuit too). It's not the driver either, as that's just transistors, and I've already tried swapping the logic chips driving them. I've also tried flicking it with my finger at various angles. Is there anything I can do or is it dead?

[gregebert]

Can you swap out the tube ? That way you can isolate between a tube problem versus something on the board.

Sounds like some bench-testing of the tube might be needed

[Trevor Haddox]

On Monday, February 5, 2018 at 7:55:46 PM UTC-5, gregebert wrote:

Can you swap out the tube ? That way you can isolate between a tube problem versus something on the board.

Sounds like some bench-testing of the tube might be needed

I'm pretty sure it's not the board. I've swapped around all the logic chips in that circuit and all that's left is the cathode drive transistors. I'll desolder the tube when I get a chance and apply voltage to it. Someone else suggested running the HV between the defective cathodes, but could that make the problem worse? 

[Terry S]

Have you inspected the pin area for debris -- something that might become conductive at high voltage?
Also the PCB in that same area.

Terry

[MichaelS]

It's possible that you have a whisker between the digits.  You can blast it by applying high current between the digits, but it may do more harm than good.

On Monday, February 5, 2018 at 4:46:19 PM UTC-6, Trevor Haddox wrote:

[gregebert]

I've had that happen many times with IN-1's, and once with a Burroughs 6091.

However, in the case of shorted cathodes, both will be equally lit. One is not brighter than the other.

[Tomasz Kowalczyk]

If the bridge created by sputtering still has significant resistance (sputtered layer has sub micron thickness), then it doesn't behave like a short. Imagine this as some other digit being connected to the same driver transistor, but through a resistor of 10-100k value. Then the other digit will be fainter.