HELP WITH HEATHKIT SB-634 - cathode poisoning reversal

[Alex [Temple-Heath List]]

At 11:57 PM 4/10/2007, you wrote:
>I HAVE SUBJECT ITEM PURCHASED ON EBAY. IT WAS ADVERTIZED AS SEGMENT 
>"F" NOT WORKING ON ALL SIX DISPLAY TUBES. I FOUND A 100K RESISTOR 
>OPEN IN THAT PART OF THE CIRCUIT. I REPLACED THE RESISTOR AND ALL 
>SEGMENTS ARE NOW WORKING EXCEPT ALL SIX NOW ONLY HAVE A SMALL 
>PORTION THAT IS LIGHTING. THE ENDS OF EACH REMAIN DARK AND THE 
>CENTER PART LIGHTS UP TO A LENGTH OF ABOUT 1/8". I HAVE REPALCED THE 
>TRANSISTOR (Q101) WITH SEVERAL OTHER GOOD ONES AND NO JOY. MY 
>QUESTION IS, CAN THIS HAPPEN TO SIX DISPLAY TUBES AT THE SAME TIME? 
>SEEMS DOUBTFUL BUT DON'T KNOW WHAT ELSE IT COULD BE. ANYONE OUT 
>THERE ANY KIND OF EXPERT ON THESE DISPLAY TUBES?
>
>MURRAY

Hi Murray,

It seems to me that the display was used for a long time without the 
F segment working, and most likely what happened is that those unused 
cathodes of the gas discharge display where "poisoned" by material 
sputtered off the nearby segments that continued to see normal 
service. This process essentially builds up an insulating layer of 
material over any unused cathode and is the reason why they now only 
glow on the middle portion, which is further away from any 
neighboring segment, so the area that still has some glow was 
affected in a lesser way. During normal use the poisoning is kept in 
check because every time a segment glows, it cleans itself a bit. But 
if it becomes totally inactive, like in this case, the build-up 
continues and gets worse over time. This same thing also happens to 
unused digits in Nixie tubes, like in clocks where for example the 
minutes' "tens" only uses digits 0-5, so 6-9 slowly get poisoned 
because they are never used.

The good news is, if the poisoned segment or digit still has some 
glow, that it is very likely to be a restorable candidate . I 
recently did this in a SP352 of a ID-1590 where some segments where 
starting to loose glow towards the ends. So there where small 
portions without any glow at all, and numbers where no longer a 
continuos -and nice looking- trace of light. I once also restored 
about half the upper portion of the A segment of the "1" numeral in a 
3-1/2 digit SP354 unit of a ID-1390.

Even though reversing cathode poisoning in Nixie tubes is much more 
straight forward, or less risky I might also say, one has to be very 
careful not overdoing it in Beckman displays, as the inside of the 
glass front cover might be darkened, and so the digit would be doomed 
permanently. Reversing poisoning involves driving the digit/segment 
between 2-10 times the normal current (we're talking in micro-amps 
here), so as to speed up a bit the cleaning effect. Still it might 
take hours, or sometimes days to completely restore a cathode. As 
said before, one has to go much slower with the 7-segment Panaplex 
units. The SP352 is rated at 330uA per segment, and I had good 
results with a current of 700uA (aprox a 2:1 ratio) which took about 
4-5 hours to clear up each segment. But keep in mind that in my case 
the segments only showed a small poisoned area, so your mileage may 
vary. Also make sure that the voltage across the segment does not go 
over 180V, which is the maximum allowable. In my setup it measured 
about 140V pretty much at any current setting. In any case if you're 
not in a hurry you can try a lower current, you just will have to 
wait longer to see any effects. Still you have to make sure you can 
safely increase the current to the point where at least a small 
portion of the poisoned area starts to glow a bit, without the rest 
being abnormally bright. Remember you need it to glow, even if only 
slightly, to get the healing process going.

I use a self-build electrolytic cap restorer to power the process. 
This way I can easily vary the current supplied to the cathode of the 
display by switching voltage ranges and/or forming rate speeds, 
because basically both are resistor divider networks connected in 
series with the output of the cap restorer's HV power supply. My unit 
goes from 100 to 600V in ten steps, and has four forming rate 
positions: slow, normal, fast, and direct. BTW, the building article 
of this restorer appeared in the Winter 1974 Electronic 
Experimenter's Handbook. If anyone wants a copy of the article 
contact me direct.

If you have access to a HV power supply, like the Heathkit IP-17, you 
could hook up a current limiting resistor (300K~1M) between the 
output and the display, and set the voltage for the desired current. 
For example, if you set the power supply to 150V, and use a 390K 
resistor, you'll get a current of about 385uA. Ohm's law (I=V/R) 
should help you figure out other workable combinations.

You can read more about cathode poisoning tutorials here at this site 
(pertains Nixie tubes), but still is very useful info:

http://www.tube-tester.com/sites/nixie/different/cathode%20poisoning/cathode-poisoning.htm

Pinout guides and specs of the Beckman Panaplex displays used in 
Heathkit units are here:

http://www.tube-tester.com/sites/nixie/data/sp-352/sp-352.htm

Good luck with the project. These display units are becoming more and 
more scarce every day, so its a very worthwhile cause to try and restore them.

-Alex  

-----------------------------------------------------------
Products bought, sold or traded here is the responsibility of the
parties involved.  This list and the City of Tempe are not responsible
for losses or misrepresentations of any kind.  Buyer beware!

-----------------------------------------------------------
This list is a public service of the City of Tempe, Arizona
-----------------------------------------------------------

Subscription control - http://www.tempe.gov/lists/control.aspx?list=HEATH
To post - HEATH at LISTSERV.TEMPE.GOV
Archives - http://listserv.tempe.gov/archives/HEATH.html