18 Jul
2021
18 Jul
'21
10:33 p.m.
I am curious if anyone has attempted to repair (replace) a broken wire in a
PDP-8/e H212 (MM8EJ) core mat (8 k word). The cores are not visible without
a microscope. I cannot imagine how these were even manufactured and wonder
if DEC service centers repaired core mat faults or if faulty boards were
simply discarded.
CDC 6600 cores were huge in comparison and I would not hesitate to replace
a broken core wire on those.
Best regards
Tom Hunter
19 Jul
19 Jul
12:59 a.m.
Tom said
I am curious if anyone has attempted to repair
(replace) a broken wire in a
PDP-8/e H212 (MM8EJ) core mat (8 k word). The cores are not visible without
a microscope. I cannot imagine how these were even manufactured and wonder
if DEC service centers repaired core mat faults or if faulty boards were
simply discarded.
CDC 6600 cores were huge in comparison and I would not hesitate to replace
a broken core wire on those.
I have no idea how DEC made theirs but for IBM's System 360, one of their engineers
came up with the clever idea of stretching the core wire so it necked and broke, leaving
a work-hardened tapered point to thread the cores with.
They patented it https://patents.google.com/patent/US3314131A
(Source: page 187 of 'IBM's 360 and early 370 Systems' by Pugh et al)
On a Youtube film about the 360 they show cores being vibrated into the correct
orientation on a jig board.
1:40 a.m.
------ Original Message ------
From: "Tom Hunter via cctalk" <cctalk at classiccmp.org>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Monday, 19 Jul, 2021 At 06:33
Subject: DEC PDP-8/e H212 core mat repair
I am curious if anyone has attempted to repair (replace) a broken wire
in a
PDP-8/e H212 (MM8EJ) core mat (8 k word). The cores are not visible
without
a microscope. I cannot imagine how these were even manufactured and
wonder
if DEC service centers repaired core mat faults or if faulty boards were
simply discarded.
CDC 6600 cores were huge in comparison and I would not hesitate to
replace
a broken core wire on those.
Best regards
Tom Hunter
I believe much of the core manufacturing for DEC minicomputers was
outsourced, but a lot of it had become much more automated by the late
60's and early 70's. I believe it was done by machine, with a tray to
hold the toroids in place, and a very fine needle-like "bobbin" that
threaded the wires through the toroids. I believe threading the cores by
hand had become largely obsolete by the time the PDP-8 came onto the
market.
Though i can't confirm it, i highly doubt that DEC engineers would
repair core planes. These would more likely be sent back to the
manufacturer for "recycling", with the cores being recovered and reused.
It's worth noting that most computer manufacturers appreciated the
fragility of core memory planes at the time, with most of them being
protected with either PCB's or perspex/plastic shields on top of the
core planes. In this way, it was rather difficult for a clumsy ol'
technician to put his thumb through the planes as he was servicing
machines.
Josh
2:50 p.m.
On 7/19/21 3:40 AM, Joshua Rice via cctalk wrote:
I believe much of the core manufacturing for DEC
minicomputers was
outsourced, but a lot of it had become much more automated by the late 60's
and early 70's.
I've got a trio of planes here, two of which are from a Lockheed MAC-16,
but the other one is made by Keronix out of Santa Monica for an unknown
machine (dated 1973, model number "P4" and p/n 816335 if that means
anything to anyone, approx 16"x16" with two 100-pin, double-sided finger
edge connectors on 0.1" spacing).
Anyhoo, the Keronix one has a sticker on it saying it was repaired by DMA,
inc. in Amery, WI in 1980 - which might suggest that there were third
parties around working on boards, rather than them having to go back to the
manufacturer for repair. (I have no idea what the nature of the repair was,
of course; maybe it was to surrounding logic rather than the mat itself).
It's worth noting that most computer manufacturers
appreciated the
fragility of core memory planes at the time, with most of them being
protected with either PCB's or perspex/plastic shields on top of the core
planes.
Yes, that's how all the ones I've ever seen have been. The Keronix one has
an additional shield over the top of the entire PCB, on top of the one
protecting the cores.
Jules
3:06 p.m.
Yes most core stringing was outsourced.
By hand under magnification was used.
I cant recall any references to automation.
That would br down to the supplier.
The story I heard was at least some were done by embroidery girls in
Hong Kong
Rod Smallwood ? -- Digital Equipment Corporation? 1975 to 1985
On 19/07/2021 22:50, Jules Richardson via cctalk wrote:
On 7/19/21 3:40 AM, Joshua Rice via cctalk wrote:
I believe much of the core manufacturing for DEC
minicomputers was
outsourced, but a lot of it had become much more automated by the
late 60's and early 70's.
I've got a trio of planes here, two of which are from a Lockheed
MAC-16, but the other one is made by Keronix out of Santa Monica for
an unknown machine (dated 1973, model number "P4" and p/n 816335 if
that means anything to anyone, approx 16"x16" with two 100-pin,
double-sided finger edge connectors on 0.1" spacing).
Anyhoo, the Keronix one has a sticker on it saying it was repaired by
DMA, inc. in Amery, WI in 1980 - which might suggest that there were
third parties around working on boards, rather than them having to go
back to the manufacturer for repair. (I have no idea what the nature
of the repair was, of course; maybe it was to surrounding logic rather
than the mat itself).
It's worth noting that most computer
manufacturers appreciated the
fragility of core memory planes at the time, with most of them being
protected with either PCB's or perspex/plastic shields on top of the
core planes.
Yes, that's how all the ones I've ever seen have been. The Keronix one
has an additional shield over the top of the entire PCB, on top of the
one protecting the cores.
Jules 
3:16 p.m.
On 2021-07-19 6:06 p.m., Rod Smallwood via cctalk wrote:
Yes most core stringing was outsourced.
There's more detail on early core production in the book "IBM's Early
Computers", iirc. (And possibly "A Few Good Men From Univac".)
...
Rod Smallwood ? -- Digital Equipment Corporation? 1975 to 1985
On 19/07/2021 22:50, Jules Richardson via cctalk wrote:
> On 7/19/21 3:40 AM, Joshua Rice via cctalk wrote:
>> I believe much of the core manufacturing for DEC minicomputers was
>> outsourced, but a lot of it had become much more automated by the
>> late 60's and early 70's.
>
> I've got a trio of planes here, two of which are from a Lockheed
> MAC-16, but the other one is made by Keronix out of Santa Monica for
> an unknown machine (dated 1973, model number "P4" and p/n 816335 if
> that means anything to anyone, approx 16"x16" with two 100-pin,
> double-sided finger edge connectors on 0.1" spacing).
>
> Anyhoo, the Keronix one has a sticker on it saying it was repaired by
> DMA, inc. in Amery, WI in 1980 - which might suggest that there were
> third parties around working on boards, rather than them having to go
> back to the manufacturer for repair. (I have no idea what the nature
> of the repair was, of course; maybe it was to surrounding logic rather
> than the mat itself).
>
>> It's worth noting that most computer manufacturers appreciated the
>> fragility of core memory planes at the time, with most of them being
>> protected with either PCB's or perspex/plastic shields on top of the
>> core planes.
>
> Yes, that's how all the ones I've ever seen have been. The Keronix one
> has an additional shield over the top of the entire PCB, on top of the
> one protecting the cores.
>
> Jules
1022
days inactive
1022
days old
5 comments
6 participants
participants (6)
-
ccth6600＠gmail.com -
jules.richardson99＠gmail.com -
rice43＠btinternet.com -
rodsmallwood52＠btinternet.com -
steven＠malikoff.com -
toby＠telegraphics.com.au