Captain Obvious here. Those edge card power terminals are wasteful, you
need to snap them over both sides, and those sides MUST be for the same
signal / voltage...

Pin / sockets are denser.

On Sat, 27 Nov 2021 15:40:58 -0600, Louis Ohland <ohland@charter.net>
wrote:

>Captain Obvious here. Those edge card power terminals are wasteful, you
>need to snap them over both sides, and those sides MUST be for the same
>signal / voltage...
>
>Pin / sockets are denser.

I never figured out why they used that method to connect the 50/70 PS.
I know those machines were developed in a different plant and I assume
they didn't want to duplicate anyone else's design. It does make a
more compact design and I think that was a big part of it. They wanted
to get away from that "tower" thing.
I always liked the 70. The 50 and 60 never made much sense to me tho.
It was basically just a microchannel PC/AT with all the limitations of
both.

My SWAG is the automated assembly line they used needed the reduced
degrees of freedom of the planar edge connector?

On 11/27/2021 19:03, gfretwell@aol.com wrote:
> On Sat, 27 Nov 2021 15:40:58 -0600, Louis Ohland <ohland@charter.net>
> wrote:
>
>> Captain Obvious here. Those edge card power terminals are wasteful, you
>> need to snap them over both sides, and those sides MUST be for the same
>> signal / voltage...
>>
>> Pin / sockets are denser.
>
> I never figured out why they used that method to connect the 50/70 PS.
> I know those machines were developed in a different plant and I assume
> they didn't want to duplicate anyone else's design. It does make a
> more compact design and I think that was a big part of it. They wanted
> to get away from that "tower" thing.
> I always liked the 70. The 50 and 60 never made much sense to me tho.
> It was basically just a microchannel PC/AT with all the limitations of
> both.
>

Cost was probably the #1 reason behind that design decision.

The way they implemented it in the 50/70 is ok-ish. But the same can't
be said about the Model 90 / 5494. Thankfully they abandoned it for the
35/40, 56/57 & 76/77.

On 28.11.2021 4:28, Louis Ohland wrote:
> My SWAG is the automated assembly line they used needed the reduced
> degrees of freedom of the planar edge connector?
>
> On 11/27/2021 19:03, gfretwell@aol.com wrote:
>> On Sat, 27 Nov 2021 15:40:58 -0600, Louis Ohland <ohland@charter.net>
>> wrote:
>>
>>> Captain Obvious here. Those edge card power terminals are wasteful, you
>>> need to snap them over both sides, and those sides MUST be for the same
>>> signal / voltage...
>>>
>>> Pin / sockets are denser.
>>
>> I never figured out why they used that method to connect the 50/70 PS.
>> I know those machines were developed in a different plant and I assume
>> they didn't want to duplicate anyone else's design. It does make a
>> more compact design and I think that was a big part of it. They wanted
>> to get away from that "tower" thing.
>> I always liked the 70. The 50 and 60 never made much sense to me tho.
>> It was basically just a microchannel PC/AT with all the limitations of
>> both.
>>
>

On Sat, 27 Nov 2021 21:28:44 -0600, Louis Ohland <ohland@charter.net>
wrote:

>My SWAG is the automated assembly line they used needed the reduced
>degrees of freedom of the planar edge connector?

The pictures I have seen of the PS/2 assembly areas look more like all
of this stuff was hand assembled on a work bench type operation where
one person built the whole machine from FRUs. I will pop this question
up on the retiree BB and see what the guys who made these things say.

IBM also had the 5394 (with the same case and PSU used for the smallest "desktop" 3174 submodels) where the '02A' submodel is a reduced planar with PCB fingers into the PSU housing for power instead of the earlier harness. All of the stamped metal shielding around the planar is removed (in addition to the fan). The power connection is more elegant in a 50/70 style (and not using a power harness with the connectors like the Model 90 and 5494), but still a kludge inside the PSU - It would be interesting to see if the slot was already present with the earlier wire-harness versions, where it may have not been used for not enough power delivery.

Sounds more like special bid than the normal variants offered in
announcement letters. Still, anything is possumble.

On 11/28/2021 11:38, gfretwell@aol.com wrote:
> On Sat, 27 Nov 2021 21:28:44 -0600, Louis Ohland <ohland@charter.net>
> wrote:
>
>> My SWAG is the automated assembly line they used needed the reduced
>> degrees of freedom of the planar edge connector?
>
> The pictures I have seen of the PS/2 assembly areas look more like all
> of this stuff was hand assembled on a work bench type operation where
> one person built the whole machine from FRUs. I will pop this question
> up on the retiree BB and see what the guys who made these things say.
>

On Sun, 28 Nov 2021 22:02:15 -0600, Louis Ohland <ohland@charter.net>
wrote:

>Sounds more like special bid than the normal variants offered in
>announcement letters. Still, anything is possumble.
>
I am watching the thread on the retiree page and I haven't seen any of
the usual PS/2 team chime in. I am pretty sure we have a few people
who were in Austin or Boca building this thing. I don't remember which
was which but I am pretty sure the 70 and 80 came from different
plants. It's sort of like the K/Pok and Endicott thing in mainframes.

Nope, Not Invented Here

On Sat, 27 Nov 2021 21:28:44 -0600, Louis Ohland <ohland@charter.net>
wrote:

>My SWAG is the automated assembly line they used needed the reduced
>degrees of freedom of the planar edge connector?
>

I saw 1 answer that might have been on point by 2 guys

G McG
Back in the old days Reliability, Availability and Serviceability
were the key tenants of the field service divisions. We had a career
path for Service Planning Reps who fought for these attributes daily
with manufacturing, engineering, marketing and others.
Using an edge connector instead of discrete wiring would have been the
type of thing that might have arisen from whoever did his sort of
thing initially.

·
P F
agree serviceability was key and a single FRU you could plug in
without making a mistake. IBM dealers had to ramp up staff quickly and
Inexperienced Field techs would go out of there way to plug a cable in
backwards even when they are keyed or forget to connect a cable. PS/2
were idiot proof, disassemble quickly (they were tool less using thumb
screws and clips as long as you could find the keys) and replace the
appropriate module. It was hard to make a mistake or forget to plug
something in.

Most of them were just the old N.I.H thing at Boca. They started that
design on a clean sheet of paper and didn't use much of anything the
other machines used.

It was sort of like why the 55s and similar used the Dallas clock.

I must agree, it would be VERY hard to install a 50 / 70 PSU wrong.

On 11/30/2021 15:28, gfretwell@aol.com wrote:
> On Sat, 27 Nov 2021 21:28:44 -0600, Louis Ohland <ohland@charter.net>
> wrote:
>
>> My SWAG is the automated assembly line they used needed the reduced
>> degrees of freedom of the planar edge connector?
>>
>
> I saw 1 answer that might have been on point by 2 guys
>
> G McG
> Back in the old days Reliability, Availability and Serviceability
> were the key tenants of the field service divisions. We had a career
> path for Service Planning Reps who fought for these attributes daily
> with manufacturing, engineering, marketing and others.
> Using an edge connector instead of discrete wiring would have been the
> type of thing that might have arisen from whoever did his sort of
> thing initially.
>
> ·
> P F
> agree serviceability was key and a single FRU you could plug in
> without making a mistake. IBM dealers had to ramp up staff quickly and
> Inexperienced Field techs would go out of there way to plug a cable in
> backwards even when they are keyed or forget to connect a cable. PS/2
> were idiot proof, disassemble quickly (they were tool less using thumb
> screws and clips as long as you could find the keys) and replace the
> appropriate module. It was hard to make a mistake or forget to plug
> something in.
>
> Most of them were just the old N.I.H thing at Boca. They started that
> design on a clean sheet of paper and didn't use much of anything the
> other machines used.
>
> It was sort of like why the 55s and similar used the Dallas clock.
>

On Tue, 30 Nov 2021 17:33:47 -0600, Louis Ohland <ohland@charter.net>
wrote:

>I must agree, it would be VERY hard to install a 50 / 70 PSU wrong.
>

I think the biggest thing was this was a group of engineers working on
a blank page. They didn't care how anyone else did it.