mike wrote in :

Adam Goldman wrote:
In article , mike wrote:

Virtually every TM500 mainframe I've bought used had at least one bad
pass transistor.
Checking them with an ohm meter is definitely better than nothing.
But...they're often overheated for long periods of time.
On a curve tracer at high current, they can look more like resistors
than transistors.
Breakdown voltage can also come way down. Beta can go to hell, then
the driver in the plugin overheats and shorts out blowing up the
plugin when the voltage goes to the rail.


Thanks for all the info!

The transistor junctions all measure about .54V drop one way and open
the other, on diode test. The high-power NPN is a little lower than
the others. I guess I'll still look into rigging up a curve tracer.
How much current/voltage do you usually have to put into them to make
the problem show?

There's not a cheap sub for these transistors if I just wanted to
replace them all preemptively, is there?


I wouldn't sweat the ESR of the caps.


I think you're right because on the two I checked it was too low to
measure. (No, the caps aren't shorted.) I still haven't gotten at the
board to check the others and reflow the solder. It looks like I'll
have to remove half the fasteners and panels in the unit?!


Just put on a load and look at the
ripple. IIRC, full load minimum line should have a trough voltage of
at least 22V on the 25V supplies. Don't remember the number for the
11.5V, but think it was 7V trough. But it's been three decades since
I worried about the numbers.





I didn't see those numbers in the TM504 manual. Are they given in the
custom plugin kit manual or something?

Nope. You won't find 'em published, cause if you believe them, you'd
also have to believe that some of the early plugins won't always work.
Those numbers, assuming I remembered them correctly after 30 years,
are VERY conservative and very unlikely to occur in practice with a
random selection of plugins, but they guarantee everything will work
and were the guidelines for my design teams...IIRC.

mike


Take a good look at the connectors. Replace the plastic key. You
wouldn't think it important, but it helps guide the plugin so you
don't blow it up when you plug it in hot...at least not as often.
Same thing can happen if the top and bottom of the connector wears on
the fiberglass and lets the plugin slop up and down.


The barriers appear to be in decent shape. Are replacements readily
available, anyway?


And use your expendable plugin to test the hole for the first time.


Uh-oh. I guess the DC504 will have to be the guinea pig.

BTW, my unit is too old to have the two 600V/1A anti polarity
reversal diodes shown in the manual addendum, so I'm going to put in
1N4007s where they're shown in the schematic.

Thanks again for sharing your experience with me.

-- Adam




TEK actually made an 067-xxxx-99 fixture for testing TM500 mainframes,it
checked each slot individually,measured load and no-load voltages and
ripple using a DMM,scope,and external load resistors.It came with a brief
pamphlet with procedure and voltage/ripple charts for each TM500
mainframe(1-wide,3-wide,4 wide,5-wide,6-wide),measured the transistors
beta.It was never sold outside of TEK.(It also was not ISO-9000 compliant)

I suspect the remaining TEK service centers(Factory and DC) have several of
them,and the data sheet,and still use them.I know Orlando's was shipped to
DC.

--
Jim Yanik
jyanik
at
kua.net