PDP8/e programmers consoles

[allison]

On 03/16/2018 08:28 PM, Charles Dickman via cctalk wrote:
> So the PDP8/e used the KC8-EA programmers console which has
> incandescent indicator lamps and the PDP8/f used the KC8-FL
> programmers console which has LED indicator lamps.
Wrong.  The 8f uses the same as the 8e.  It was the 8m that had the
LED console.  If memory serves it was date of design that influenced it.

I can be certain as I have an 8f built in 1973.  I've since moded it to
take bright yellow LEDs as lamps are not easy to find and annoying
to replace.  If I care to I can put the lamps back.  Since I put
operational
over absolute faithful its better as is, less digging in the box.

> Why?
>
> The maintenance manual says that LEDs are more reliable. The operation
> is as far as I know the same. The LED version doesn't need the +8VDC
> supply, but that seems an advantage for it. I remember reading that
> the EAE had some revision dependencies with the console, so were there
> bus timing problems with the LED console that limited it to short
> buses?
Cost to redesign and the cost of LEDs, and their lack of brightness then
didn't justify it.  The 8m was aimed at the industrial user that would use
it differently, usually wrapped in a box to protect it from the shop floor
dust and grime and generally less maintenance.

Many sites have the M and the F swapped on the LEDs.  However the
LED equipped machines are rare and all the 8f machines I know of
have the 8v power for lamps.  The LED version can use 5V or 8V.

However its possible to swap the panels, I tried it once and it pass diags
and worked fine.  Didn't like the red.

The biggest difference is the lamp panels have resistors to keep
the filament warm but not glowing for improved life and those
are not there for LEDs (it would keep LEDs lit).    The LED did
not need that resistor but needed a current limited resistor
for the LED.  Otherwise they were the same and there was
no timing impact.

Keep in mind they all (e,f,M) could use the the industrial minimal
panel that allowed for reset and run using boot roms.

> Maybe this changed over time, but it overlapped long enough for them
> to have both described in the maintenance manual and specific to the
> application.
The Incandescent console was higher volume and lingered for that reason.
The volumes for the 8m were low as that market was also seeing inroads
from the PDP-11 and not much later the LSI-11.

There was overlap in production the 8e was very popular until the 8A
which was cheaper and more compact.

> Another console question...
>
> Has anyone ever seen a PDP8/i style console for the PDP8/e? Looking at
> the way the PDP8/e console functions, it should be possible to time
> multiplex all the register to the data bus so they are all visible at
> once when the machine is halted.
No and can't!   The 8i (nor the 8L) was not omnibus.  The boards to make
up the console were not compatible with the 8e.    Yes, you can display
the same stuff if you were to make your own, all the signals are on the
omnibus and it would take some buffering maybe some latches.  The
8e form used fewer lamps and fewer lamp driver parts so it was less
expensive to make and maintain.  Yes, one could kluge it, taking into
account you need a hand wired backplane for the M-series flipchips
that support the 8i panel and getting the signal polarity right.

Its forgotten that the 8 series runs from transistors, early DTL, TTL,
and later CMOS and had seen at least 5 iterations from the PDP-5
though the late-8 CMOS chip based series (decmate I/II/III).

The 8e represented the first of the family of pdp-8 that had a unified bus
both in signals, timing and physical structure.  It was the a major
repackage
of the PDP-8i to both a smaller, less expensive, easier to manufacture,
repair, and test.   Whats missed was the 8e/f/M were the easiest of the
8 series to interface to anything and they did indeed get a lot of hand
made one off interfaces for specific missions.  As a result the 8e was
nearly the ubiquitous small computer.


Allison


> -chuck