PDP-10 KI10 console: which key is red?

[Jörg Hoppe]

I'm in progress of restauring a KI10 console panel.
After giving power to all the push buttons, I found that ONE keycap is
RED instead of white (placed on "NXM" on the photo).

ftp://j-hoppe.dyndns.org/fotos/ki10/ki10-milestone2012-1.jpg

Can someone with a running KI10 please do a lamptest and post, which
keycap has to be the red one?

thanks,
Joerg Hoppe

[Daiyu Hurst]

So beautiful!

[Patrick Scheible]

Unfortunately, I can't help with your question. But that is a great
front panel.

-- Patrick

[John Francis]

In article <kburvl$ac2$1...@news.albasani.net>,

I didn't use the KI much (mostly KA or KLs), but I'd guess it was the
"Power" button/light that was red.

That would free up a pale blue button for sense switch 1, which in turn
would give you a dark blue one for use on NXM.

[Jörg Hoppe]

Am 02.01.2013 09:58, schrieb John Francis:> In article
<kburvl$ac2$1...@news.albasani.net>,

Thanks for seeing that miscolored SENSE1 switch.

I have some more facts now, but still no answer:

- I found an old picture of a running PDP-10 KI10 (shot by myself in
Darmstadt, Germany, 1980. I was age of 17, and it was a big day for
me!). Only few keys are illuminated, but POWER is a white light there.
See ftp://j-hoppe.dyndns.org/fotos/ki10/ki10-thd1980.jpg

- I also shot another a picture from the PDP-10 in the Kiel Computer
museum, Germany, as it was before restauration in 2010.
There POWER and LAMPTEST were painted red by hand.
See ftp://j-hoppe.dyndns.org/fotos/ki10/ki10-kiel2010.jpg
But moving my red keycap to LAMPTEST makes no sense, because LAMPTEST is
the only key without a bulb (the bulb terminals in the switch are not
even connected).

regards,
Joerg Hoppe

[jmfbahciv]

I don't remember seeing red. another brain cell gone?

/BAH

[Rich Alderson]

This had to wait until I was back in the office after the winter holiday.

There are *NO* red keycaps on our KI-10. None at all.

Given that it was at one time part of the dual-processor 1077 at the University
of Kiel, I'd say that the painting of the buttons on the CPU remaining in
Germany was done at a later date, as was the substitution of a red bulb in your
own console.

--
Rich Alderson ne...@alderson.users.panix.com
the russet leaves of an autumn oak/inspire once again the failed poet/
to take up his pen/and essay to place his meagre words upon the page...

[Rich Alderson]

Rich Alderson <ne...@alderson.users.panix.com> writes:

> This had to wait until I was back in the office after the winter holiday.

> There are *NO* red keycaps on our KI-10. None at all.

And a couple of days later I walked in to find that the POWER keycap was
glowing red. Turned out that the bulb had burnt out, and the restoration lead
decided to make the power button easier to see.

So that's the way that sort of thin happens.

[jmfbahciv]

Rich Alderson wrote:
> Rich Alderson <ne...@alderson.users.panix.com> writes:
>
>> This had to wait until I was back in the office after the winter holiday.
>
>> There are *NO* red keycaps on our KI-10. None at all.
>
> And a couple of days later I walked in to find that the POWER keycap was
> glowing red. Turned out that the bulb had burnt out, and the restoration
lead
> decided to make the power button easier to see.

Boy! that's enough to make you paranoid for the rest of your life :-))))

>
> So that's the way that sort of thin happens.
>

yup.

/BAH

[Duty Calls]

On 02 Jan 2013 14:47:39 -0500, Rich Alderson

<ne...@alderson.users.panix.com> wrote:

>Given that it was at one time part of the dual-processor 1077 at the University
>of Kiel

Ah, the only thing more fun -- twice as fun, of course -- than
playing with a 1070 is playing with 1077!

I had the good fortune to encounter one during my career.
Even then it seemed like a bit of an artifact, as all the other
machines I was working with were Model B KLs running
TOPS-20. Given my relative youth, I probably didn't appreciate
it as much as I should have.

I recall one day I was sitting in the machine room and it started
issuing many strange non-halting STOPCDs. I had never seen
anything like it and I was truly puzzled for a minute. What the
hell was going on? Then I glanced over at a thermometer we
had mounted on a nearby column. It was already 75 degrees (F)
and getting warmer. That was all I needed to see. We shut it
down and called facilities maintanence.

When the decision was made to decommission it, I wrote a LIGHTS
program and added it to the system startup, which amused some
of the operators.

[Jörg Hoppe]

> Then I glanced over at a thermometer we
> had mounted on a nearby column. It was already 75 degrees (F)
> and getting warmer. That was all I needed to see. We shut it
> down and called facilities maintanence.

Well, on my KI10 panel, the big red OVERTEMP ALARM bulb was burnt out.
Seems this PDP-10 was used to run in hot environments ...

>
> When the decision was made to decommission it, I wrote a LIGHTS
> program and added it to the system startup, which amused some
> of the operators.

Would you hand it over? The SimH behind my panel could use some light
effect programs in the near future.

[fishtoprecords]

On Sunday, January 20, 2013 9:28:17 AM UTC-5, Jörg Hoppe wrote:
> Well, on my KI10 panel, the big red OVERTEMP ALARM bulb was burnt out.
> Seems this PDP-10 was used to run in hot environments ...

In some? I think they were born warm.

Did the KI have something similar to the terrible linear power supply that the KL was cursed with?

[Johnny Billquist]

Probably something worse, as the KI is older.
However, it's "only" the memory boxes of the KL that used that horrible,
huge power supply. The rest was more switched, unless I remember wrong.
But that PS for the memory was enormous (especially if you wanted to
take it out, and was alone...)

Johnny

--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: b...@softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol

[Rich Alderson]

Johnny Billquist <b...@softjar.se> writes:

> On 2013-01-20 22:43, fishtoprecords wrote:
>> On Sunday, January 20, 2013 9:28:17 AM UTC-5, Jörg Hoppe wrote:

>>> Well, on my KI10 panel, the big red OVERTEMP ALARM bulb was burnt out.
>>> Seems this PDP-10 was used to run in hot environments ...

>> In some? I think they were born warm.

>> Did the KI have something similar to the terrible linear power supply that
>> the KL was cursed with?

> Probably something worse, as the KI is older.
> However, it's "only" the memory boxes of the KL that used that horrible,
> huge power supply. The rest was more switched, unless I remember wrong.
> But that PS for the memory was enormous (especially if you wanted to
> take it out, and was alone...)

You misremember, Johnny. The input from the wall to an unmodified KL-10 is
60A 220V 3phase (13.2KVA), fed into a huge transformer. The memory power
supplies are more modern.

[Johnny Billquist]

On 2013-01-22 02:55, Rich Alderson wrote:
> Johnny Billquist <b...@softjar.se> writes:
>
>> On 2013-01-20 22:43, fishtoprecords wrote:

>>> On Sunday, January 20, 2013 9:28:17 AM UTC-5, J�rg Hoppe wrote:
>
>>>> Well, on my KI10 panel, the big red OVERTEMP ALARM bulb was burnt out.
>>>> Seems this PDP-10 was used to run in hot environments ...
>
>>> In some? I think they were born warm.
>
>>> Did the KI have something similar to the terrible linear power supply that
>>> the KL was cursed with?
>
>> Probably something worse, as the KI is older.
>> However, it's "only" the memory boxes of the KL that used that horrible,
>> huge power supply. The rest was more switched, unless I remember wrong.
>> But that PS for the memory was enormous (especially if you wanted to
>> take it out, and was alone...)
>
> You misremember, Johnny. The input from the wall to an unmodified KL-10 is
> 60A 220V 3phase (13.2KVA), fed into a huge transformer. The memory power
> supplies are more modern.

Darn! I bet I'm confused because KL10 might mean so different things...
I've only played with the DEC-20 variant. Looking at the bitsavers
documents, it should be 21.6 KVA (60A @ 120 VDC or 30 A @ 240 VAC).

And I've swapped the PS of the memory system of our -2060, which is why
I remember the misery. :-)

Johnny

[Johnny Billquist]

Trying to read and remember more, I might be wrong anyway. As far as I
can put together things, the thing I remember is the H760, which in the
docs say "RAW POWER SUPPLY". So I guess it should be for all of the machine.

Johnny

[Johnny Billquist]

On 2013-01-22 15:12, Johnny Billquist wrote:
> On 2013-01-22 14:45, Johnny Billquist wrote:
>> On 2013-01-22 02:55, Rich Alderson wrote:
>>> Johnny Billquist <b...@softjar.se> writes:
>>>
>>>> On 2013-01-20 22:43, fishtoprecords wrote:

Sorry for the spamming here. Reading on manuals sometimes takes some
time and additional readings...

The PS for the MA20 is a H7420. This is a linear PS.
There are actually two per MA20.

The actual 3-phase is fed (as with all DEC stuff) into a power control.
Bye the name 863 in the case of the DEC-20.
From there it is fed in parallel to all power supplies in the DEC-20.

And two of these sits on the back door of the CPU cabinet, which are the
ones I remember swapping. Heavy suckers! Two more sits inside the CPU
cabinet, if you have a second MA20.
And then you have three more in the I/O cabinet.
So I was wrong in that only the memory subsystem used that stuff. There
are some more. Looking at the drawings, the H7420 is used in:
DTE-20
DIA/DMA-20
RH-20

That said, the H760 feed the CPU, and that too looks pretty linear.

Johnny

[John Francis]

In article <kdm9t9$i5g$1...@Iltempo.Update.UU.SE>,

Johnny Billquist <b...@softjar.se> wrote:
>On 2013-01-22 15:12, Johnny Billquist wrote:
>> On 2013-01-22 14:45, Johnny Billquist wrote:
>>> On 2013-01-22 02:55, Rich Alderson wrote:
>>>>
>>>> You misremember, Johnny. The input from the wall to an unmodified
>>>> KL-10 is
>>>> 60A 220V 3phase (13.2KVA), fed into a huge transformer. The memory
>>>> power
>>>> supplies are more modern.
>>>
>>> Darn! I bet I'm confused because KL10 might mean so different things...
>>> I've only played with the DEC-20 variant. Looking at the bitsavers
>>> documents, it should be 21.6 KVA (60A @ 120 VDC or 30 A @ 240 VAC).

By the time we got our 2050 in Reading, the power supply was a little
more reasonable. I don't know how much of that was because of redesign,
and how much was because by that time solid-state memory was standard.

But I had also used the original KL (system 1026?) in Marlboro, and
that had a pre-production power supply. The main power input (referred
to as the Elephant's Trunk) was about a 2" diameter conduit, and I think
it had it's own special dedicated circuit - this in a machine room that
had a good number of RP-series disk drives, which themselves used a fair
amount of power (especially when spinning up a newly-mounted disk pack).

[Johnny Billquist]

On 2013-01-22 18:27, John Francis wrote:
> In article <kdm9t9$i5g$1...@Iltempo.Update.UU.SE>,
> Johnny Billquist <b...@softjar.se> wrote:
>> On 2013-01-22 15:12, Johnny Billquist wrote:
>>> On 2013-01-22 14:45, Johnny Billquist wrote:
>>>> On 2013-01-22 02:55, Rich Alderson wrote:
>>>>>
>>>>> You misremember, Johnny. The input from the wall to an unmodified
>>>>> KL-10 is
>>>>> 60A 220V 3phase (13.2KVA), fed into a huge transformer. The memory
>>>>> power
>>>>> supplies are more modern.
>>>>
>>>> Darn! I bet I'm confused because KL10 might mean so different things...
>>>> I've only played with the DEC-20 variant. Looking at the bitsavers
>>>> documents, it should be 21.6 KVA (60A @ 120 VDC or 30 A @ 240 VAC).
>
> By the time we got our 2050 in Reading, the power supply was a little
> more reasonable. I don't know how much of that was because of redesign,
> and how much was because by that time solid-state memory was standard.

The numbers I cited *is* for machines like the 2050... :-)
The same design was used in the 2040, 2050, 2060 and 2065, as far as I know.

Johnny

[Duty Calls]

On Sun, 20 Jan 2013 15:28:17 +0100, =?ISO-8859-15?Q?J=F6rg_Hoppe?=
<j_h...@t-online.de> wrote:
>>When the decision was made to decommission it, I wrote a LIGHTS
>>program and added it to the system startup, which amused some
>>of the operators.
>Would you hand it over? The SimH behind my panel could use some light
>effect programs in the near future.

I don't know if I still have it. If I do, it's on media that I
can't access easily.

The LIGHTS UUO wasn't supported for the KL, which had no front
panel. On the KI, the LIGHTS UUO only controlled the lights of
the data switchs -- that's one row of 36 lights.

The program worked by using the LIGHTS UUO in a loop that set a
pattern in the lights, sleeping for a fraction of a second (say
about 200 milliseconds), then changing the pattern slightly. This
can create the illusion of moving lights. The ROT and LSH
instructions get used a lot.

One simple pattern is a single light that moves down row and then
"jumps" back to the beginning when it reaches the end. It looks
like single light that goes round and round. A slightly different
pattern is to have the light "bounce" back and move the other way
when it reaches the end of the row. A variation on either of
these is to increase the number of lit bulbs with each loop. So
first one light seems to travel the length of row. At the end it
travels the other direction, but now it's two lights. It reverses
again and becomes three lights. Eventually more bulbs are lit
than not, and it appears to be a dark spot moving back and forth.
This can be reversed to start with one unlit bulb and reducing the
number of lit bulbs with each iteration. All pretty simple stuff.

My program had several different patterns. Every two or three
minutes it would change and pick another one at random. From time
to time I would add a new pattern.