DCC-116 E / DATA GENERAL NOVA 2/10 / Nixdorf 620 - Restoring and restarting

[Christian Kennedy]

On 4/1/17 12:33 PM, Dominique Carlier via cctalk wrote:

[snip]

> Anyway, the Nixdorf 620 is actually built by "Digital Computer Controls"
> and after some researches it seems that it is a "DCC-116 E", the 17 slot
> version of the "DCC-116" which Is a clone of the "Data General Nova 2/10".

Er.  My DCC-116 is more a clone of the mapped Nova 840 than the Nova 2,
with the DCC/Keronix "64Kword" hack that gives up more than one level of
indirect memory reference in exchange for having 16, rather than 15,
bits of logical memory reference.

[snip]

> Here is the list of the boards I have with the references written on
> them and their position in the machine:
> 
> *17    HEX 0 – 15                 ENTREX INC 62 00 01842 002 REV A*
> 
> *16    HEX 16 30 31             ENTREX INC 62 00 01842 002 REV A**
> 15    OPTION 2                  Empty**
> 14    OPTION 1                  Empty**
> 13    PRINTER                   ENTREX INC SN598**
> 12    TAPE 556/800 BPI     2433 LFI 213**
> 11    TAPE 1600 BPI          Empty**
> 10    DISK                           2456 00 MP-Kontroller D44an620
> 6393500215 2456 7 0 1577**
> 09    COMMO                     Empty**
> 08    SCANNER                  Scanner BD 2431 NCAG 54147.1.15 2431 02394**
> 07    MEM                           1609 0 7 02616**
> 06    MEM                           Empty**
> 05    MEM                           1609 0 8 02367**
> 04    MEM                           1615 01 9 5596**
> 03    TTY                             Empty**
> 02    DO NOT USE             Empty**
> 01    C.P.                              1509 05 4 04436*
> 
> Concerning this I also ask a few questions:
> 
>  *
> 
>    What are the boards in slots 16 and 17 for?

Packaging aside, this looks a lot like an Entrex 480 system, which was a
key-to-disk/tape system:

https://www.cs.auckland.ac.nz/historydisplays/FourthFloor/ReportsAndViewer/Reports/Entrex.pdf

My guess is you're looking at bespoke 16-port multiplexers for talking
to their terminals.

>    I do not have a COM card, does that mean that I could not attempt a
>    serial transmission (type rs232) with this actual setup ?

IIRC the functionality of the basic I/O board normally found in slot
three is subsumed into the CPU card of the DCC116, so you should be fine
as long as you can find the signals on the back of the machine.

> 
>  *
> 
>    I have no idea how many kilobytes are present, apparently there are
>    two core memory cards (8kb each? 16Kb each?), And non-core RAM board.

It's a word oriented machine, so memory is sized in KW, not KB.  You can
probably tell how big the boards are by looking at the physical address
jumpers; core boards were typically 8KW, MOS boards were typically 16 or
32KW, depending on vintage. Hint:  The more jumpers, the smaller the board.

> Unfortunately, after a couple of hours, the "POWER FAIL" light has
> started to light up. The problem here is that I have no schematics, it
> is a modular power supply consisting of two elements:
> 
> http://www.actingmachines.com/classiccmp_forum/psus.jpg
> <http://www.zeltrax.com/classiccmp_forum/psus.jpg>

I have no idea how this machine packages things, but in jumbo-chassis
Novas the lower PS generates five and 15 volt supplies, the top PS
generates only five.  The reason for this is that the 15 volt supply is
only used by core memory boards and the memory bus is only available in
the lower chassis.

> One seems to be dedicated to the 15 VDC regulated, the second module is
> in charge of the regulated 5VDC. Both elements have a "Power fail
> module". When the machine is completely populated I now observe that the
> + 5VDC collapses at +1,9VDC.

Sounds like one of the +5V switchers is having issues.

[snip]

> 
> Another thing I noted during my measurements, where I should (I think)
> measure 15V, I have 18V, looking at the documentation on the NOVA 2 (4
> slots), I observed according to the diagrams below That there is 18V but
> before a zener diode and not at the level of connectors where there must
> be 15VDC, should I warn me ? Would the diode be dead ?

+/-5V and +15V but not -15V should be regulated.  In the original DG
power supply both five and 15 volt supplies were derived from a common
+30V unregulated supply, but it's not clear that DCC followed that
model.  It may be helpful to know that the +15V supply is not regulated
to 15V but temperature compensated downward with increasing temperature,
such that at 55C it's closer to 14.4V; this is done to maintain margins
on the core planes. The consequence is that you're probably looking at
something having drifted in the voltage divider that feeds the voltage
comparator or something having drifted in the temperature compensation
stuff (if it's actually there) rather than a simple zener gone bad.

> 
> I tried to join temporarily a secondary regulated power supply to
> reinforce the + 5VDC so as to be sure that the problem did not come from
> the logical part of the computer which checks these voltages and
> triggers the Power Fail status. By coupling this external regulated
> power supply, the computer restarts again, thus it is definitively the
> power supply (or the comparator component(s) of the power fail module)

/POWER FAIL is asserted by the power supply itself, as is MEM OK.

> 
> If I removes all the boards (printer, core memory, scanner, disk
> controller, etc.), the Power Fail light eventually goes out, I get again
> the 5VDC, so the power has become "too weak" to power the computer when
> it is fully populated.

It's a switcher; look at the caps in the LC filter (downstream of the
series pass transistor) that, together with the inductor, form the
energy storage mechanism of the power supply; check the source supply as
well.  The fact that it eventually comes back suggests that the
reference, comparator and pass device are probably functioning.

[snip]

> Notes that the Operating system I used to make these tests is a very
> limited OS named DIDOS which was distributed by Nixdorf for his 620
> Serie. As the machine is a clone of DG NOVA 2/10, I wonder if the
> computer is compatible with programs for the NOVA series of Data
> General? So I wonder if the DCC-116 was compatible with the software for
> NOVAs?

Should be, as long as the machine isn't running in extended logical
address mode (my DCC has a front panel switch to enable/disable this
feature).

> The other questions I ask myself at this point are the ways to transfer
> softwares or bootable OS on this machine? I have no other disk pack on
> another system but I have a nine track tape TU81 + connected to a VAX
> 4000-605A under OpenVMS 6.2, and I can put this computer on my network.
> Do you think there's a way to make a bootable tape for my NOVA 2 clone
> with the Pertec 8840A from a VAX 4000 with a TU81+ ?

Probably.  There's nothing magic about DG tape images.

Nice job on getting the machine to boot!

Cheers,
Chris

-- 
Christian Kennedy, Ph.D.
chris at mainecoon.com	AF6AP | DB00000692 | PG00029419
http://www.mainecoon.com	PGP KeyID 108DAB97
PGP fingerprint: 4E99 10B6 7253 B048 6685 6CBC 55E1 20A3 108D AB97
"Mr. McKittrick, after careful consideration…"