Don't know what to say on the transistor. It may have been actually made
that way. They did lots of things back then. Yes familiar with the bcd
decoder its used in the 5245 class counters also. I think someone was doing
some funny stuff at lunch time to come up with that. It was the 60's after
all.
How about some pix of a 9815 calc. That would be pretty neat.
On really old stuff if you can't find the part needed its pretty easy to
replace the whole function with modern answers. I know it sort of breaks
the original mode but for me at least its a case of getting it going.
Regards
Paul
WB8TSL

On Sun, Nov 6, 2016 at 1:59 PM, Adrian Godwin artgodwin@gmail.com wrote:

You could try heating the transistor and bushing then give the to5 a shot
of cold spray or supper duster hoping it will shrink enough to slide out.

On Sunday, 6 November 2016, Adrian Godwin artgodwin@gmail.com wrote:

Just homemade opto-isolators. Used in choppers, too. The transistor was
indeed bonded to the heatsink.  Just replace the whole thing with a
3-legged regulator? or simply a modern PNP t0-220 with a little heat
sink on it.
73, Don

On 2016-11-06 15:15, paul swed wrote:

--
Dr. Don Latham
PO Box 404, Frenchtown, MT, 59834
VOX: 406-626-4304

Adrian,

Here's a picture of the package you have, with the TI logo on the "heatsink". I pulled it just now from an old car stereo. The PNP Germanium GC588 crosses - not correctly! - to a differently packaged NTE102A which is rated for 1A and 32V with some decent gain. I suspect the GC588 is closer to your 2N1038-2, which is rated 3A. If one or two of these (curve tracer tested, even) would help you just send me an address off list.

Bob LaJeunesse
Ann Arbor, MI

Yes, further searching reveals that although the 2n1038 is T05, the -2
version is X26 - mounted in a stud by the factory. So it could well be
glued. 2N1038s aren't unobtainable though, so I'll keep it original if I
can. Even though a 7915 would be fine !

I was a bit puzzled by the insulators. Very metallic-looking. Then I
realised they're BeO. Pretty unusual these days (no, I won't be machining
them) so another reason to keep a T05 in there.

On Sun, Nov 6, 2016 at 10:46 PM, djl djl@montana.com wrote:

A 2N1038 is in a TO-5 or TO-11 package.
A 2N1038-2 is a modified or selected part.
Is this transistor pressed into a hexagonal bushing that is about 1/2 "
long with a 10-32 threaded stud for mounting?
If so the bushing/heatsink is part of the transistor.
A google search for 2N1038 will show this this transistor and a similar
one with the integral heatsink.

Glenn

On 11/6/2016 5:46 PM, djl wrote:

--

Glenn Little                ARRL Technical Specialist  QCWA  LM 28417
Amateur Callsign:  WB4UIV            wb4uiv@arrl.net    AMSAT LM 2178
QTH:  Goose Creek, SC USA (EM92xx)  USSVI LM  NRA LM  SBE ARRL TAPR
"It is not the class of license that the Amateur holds but the class
of the Amateur that holds the license"

Glenn,

I think you're describing what Robert has photographed.

There are some of those packages, but they're the 2N1373 driver transistors
and have a smaller mounting hole. I dismounted Q2 and found it's really a
2N2556 (but still T05) and similarly mounted, but in a bigger stud. The
threaded part is wide enough to hold the package itself.

Since the insulating washer can now be kept safe, I can probably just punch
or heat the T05 out (it has a hole right through the stud).

On Sun, Nov 6, 2016 at 11:16 PM, Glenn Little WB4UIV <
glennmaillist@bellsouth.net> wrote:

The real hack in the 5245 Nixie tube decoder is that the neon lamps were
the memory element.

hld

On 11/6/2016 2:15 PM, paul swed wrote:

--
Howard L. Davidson
hld42@att.net

I'm pleased to report this TIC is now running again (albeit with some
temporary hacks, such as a 7915 regulator).

There were a couple of dried-out electrolytics, the shorted series-pass
regulator referred to above, and an open-circuit 2N708 controlling the
timebase gating. Another 2N708 died when I failed to push a PCB fully into
its socket. They're rather delicate flowers - Silicon, but only 15V Vceo :
a bit tight when used as switches with a 15V or 13V rail. Replaced with
BC237s, though that's perhaps a bit slow.

This is the HP5275A from france that  Poul-Henning Kamp mentioned in march
2015. It's been sitting here a while waiting for some attention
​
VID_20161109_003111.mp4
https://drive.google.com/file/d/0B8q6sqvuGdFfMDI2aXA0NnFWRTg/view?usp=drive_web

Paul Swed : I'll add some HP9815 porn when I get it reading the 5275A
outputs.

On Mon, Nov 7, 2016 at 2:46 AM, Howard Davidson hld42@att.net wrote: