Datron 1081 wierd display fault

[splin]

I just acquired a Datron 1081 which had some display issues with 'unlit' segments shimmering. It was perfectly usable but replacing the faulty 350V 10uF (measured 490pF!) on the power supply board sorted that.

One fault remains however; when displaying certain numbers an extraneous glow appears on the top right as shown in the photo. It looks a bit like an apostrophe but on closer examination it is actually glow that appears to spread out from an always lit segment that is normally hidden by a small tab/mask over that part of the display that is obviously designed to hide it. I don't know if that tab is part of the Panaplex display or something Datron added. You can see the lit segment if you view it at a sharp angle from above and to the side.

When the unwanted glow appears it seems to be because that hidden segment gets a lot brighter. By entering numbers via the keyboard I can trigger the problem but I haven't figure out the pattern.



4 triggers it if it is the right most digit. The following show the problem:
4
44
14
124
4444
77654
3333333   <= this changes sometimes it doesn't
4444444
5555555
6666666
7777777
888888
8888888

But the following don't:
41
141
144441
1111111
2222222
333333
555555
666666
777777
88888
999999

Of the following, only 877 triggers it.
87
877    <-
8777
87777
877777
8777777

It's likely that temperature changes things because I'm sure some of the combinations have changed from when I was testing yesterday.

So anybody got any ideas? The cathode voltage is a bit high at -185V (supposed to be -175V).

What is the mysterious lit but hidden segment?

[splin]

Almost certainly temperature related. Just switched it on again and entered 4444444 without triggering the fault until it warmed up a couple of minutes later.

[doktor pyta]

These displays in Datron gear used to have two bright dots in the upper part.
Probably some technology related feature.
They used smal pieces of black tape glued to the display's glass to cover these.

[Gyro]

These displays in Datron gear used to have two bright dots in the upper part.
Probably some technology related feature.
They used smal pieces of black tape glued to the display's glass to cover these.

Those are the keep-alive cathodes, they're designed to keep a source of ionisation going so that the digits light promptly - particularly important when the display is multiplexed.

They're there on the original Sperry/Beckman displays used on the previous generation 1051, but at the bottom (Datasheet attached).


P.S. The custom Datron display probably has more or less identical drive characteristics and currents.

[AVGresponding]

These displays in Datron gear used to have two bright dots in the upper part.
Probably some technology related feature.
They used smal pieces of black tape glued to the display's glass to cover these.

Those are the keep-alive cathodes, they're designed to keep a source of ionisation going so that the digits light promptly - particularly important when the display is multiplexed.

They're there on the original Sperry/Beckman displays used on the previous generation 1051, but at the bottom (Datasheet attached).


P.S. The custom Datron display probably has more or less identical drive characteristics and currents.

This is useful information to know, thanks Gyro!

I recently acquired a Solartron 7075, and though it's not been used much on account of waiting for a workspace reorganisation, eventually it'll be switched on most weekends.
You've saved me a potential worry!   

[splin]

Those are the keep-alive cathodes, they're designed to keep a source of ionisation going so that the digits light promptly - particularly important when the display is multiplexed.

Excellent information - thanks.

From the schematics I see there are indeed two keep alive cathodes driven by 330K resistors to -175V. I hadn't spotted the left hand one. Since the LH KA doesn't change brightness with the display pattern that suggests that the -ve rail is probably ok, except that it is a bit high which may be overdriving the KAs?

The right hand KA 330K resistor may be suspect or it could be a fault with the display itself. I'll try adjusting the resistor value to see if that helps.

[Gyro]

You're welcome.

As long as the supply is above the striking voltage (which it clearly is), it is relatively non-critical. Looking at the 1051 schematic, the inverter puts out 200V on the main board schematic - which magically transforms to 180V on exactly the same node when you flip over to the to the display schematic, so 175V sounds fine.

On the 1051 the KA cathodes on the Sperry displays are driven by 220k resistors to a -8.2V rail (it's a convenient a PMOS logic chip supply rail) so around 185V total. They probably upped the resistor values on the 1081 to generate a bit less nuisance glow. As long as they're glowing gently, the KAs are providing the required ionisation source, so the current isn't critical.

If the RH KA resistor is faulty (open) then the cathode may have suffered a bit of poisoning, so may not glow immediately but it should come back (sorry, looking back at your photo, the RH KA looks too bright *, so maybe the resistor value has dropped).

P.S. * That may just be a camera artefact of a constantly lit cathode next to the multiplexed digits.

[Gyro]

Agreed (EDIT: oops maginnovision's reply disappeared), as long as the glow isn't tinged with pink.

The KA cathode is bound to flicker a bit as the display segments will be 'stealing' current and voltage from it via the anode resistors, the anodes are transparent metallic film on the inside surface of the front glass. The supply voltage is unregulated too.

I'd probably just mask the KAs off again and ignore them.