• Please review our updated Terms and Rules here

Rolm 1602A mil spec computer

The right-handside DB37 plugs, with the blue wiring, are for an external core-memory box. They match the box connections on your pictures. I have neither the extension box nor the corememory sets, but the internal memory slot, now filled with a PROM board, has ( nearly..) all the signals needed : I am designing a plug-in SRAM board.

That front-panel I found is the right one for a 1602A, but unfortunatly I do need an 1642 front panel interface ( anyone has a spare ?...), and I will also need to wire up a connector. Which is just as well as the 1602A and 1602B use different connectors, and the cable that came with the box is for an 1602B...

Ah well, keeps you busy....
 
Memory: OK, so your 1602A ONLY has got the D-Sub plugs for supplying an external memory brick and there is no internal free slot for e.g. a CMOS memory module. I just pulled one of the 8k CMOS modules from my 1624 IO processor, which also uses a 5605 CPU and this obviously is not going to fit your machine (no free slot):

8kCMOS.JPG - Click image for larger version  Name:	8kCMOS.JPG Views:	0 Size:	237.6 KB ID:	1235283

CPI-Board: Why do you think, that you CPI board is not suitable for driving a panel? The fully populated CPI has got more chips on it to handle TTY and papaer tape than your's - so maybe your's is only a panel interface?

In all Rolms I saw, the panel was a single Deutsch plug with 55 pins - the TTY interface was wired to a different plug on the front panel if used (only a few wires to that one and in 1602B onwards it was a 55pin AMP plug instead of a Deutsch as seen in the 1602(A)). The paper tape interface - if used - was again wired to a different plug, so no Y-cables in that case. The wiring in all my Rolms is done directly from the front plug to the plug of the board within the chassis and you mentioned, that there are no wires connected to your CPI board -> Obviously your Rolm is not prepared for connecting a panel, i.e. it starts right after power up to execute code from the ROM.

My proposal would be, looking at my post from 8/24 which contains the wiring diagram of a 1602B with CPI board - verify that the proposed outputs of that 1642 CPI board are outputs/inputs in your CPI board. If they are open, you probably need a different CPI board, but if they are connected to some chips, you are with high probability just missing the connection of the appropriate pins to a Deutsch 55 pin connector (pinout also given in my post from last summer).

I have got a CPI board pulled from my IO processor, if you did these tests and still are sure that your CPI is not prepared for a panel, than er can probably make some arrangement for a temporary loan of that one or preferred some meeting as I am little reluctand shipping it...

Mem-CPI-Tool.JPG - Click image for larger version  Name:	Mem-CPI-Tool.JPG Views:	0 Size:	175.9 KB ID:	1235284
 
Last edited:
I was unclear : my 1602A has 2 memory slots, one of which is occupied by the ROM board, so one slot remains free. Anyhow with more modern RAMs there is enough space left on the board I am designing to add ROM and USB interface. You 8K board is made up with 32 2114 SRAM's 1Kx4, I will used 62256 and 2764. Finding those green connectors will be the main problem.

The need for a 1642 CPI : that is what the 1602B manual says. But yes, seeing that I have both a TTY and an RTC board in my 1602A means that it never had an 1642 CPI in the first place. It remains strange that I should have a S375 panel interface board, but no internal wiring for it. Then again I know nothing about the history of my particular machine. As you say : it must have been set up to run from core, or boot from tape, seeing as this is the only mass storage controller inside.

My 1635 frontpanel has a 55 pin Bendix socket with matching cable, not compatible with the Deutsch 55 pin socket. So changing some sockets on my machine is a must.

So tasks in order are :
1) finish the RAM board, check "bite" function.
2) as you suggested : draw up schematic of my S375 CPI board and check if it can drive a 1635 panel.
3) find out existing wiring between sockets & IO boards. Find out which socket is going to be redirected to front-panel duty, wire up the front panel using a Bendix socket. ( or create a cable with a Bendix plug on one end, and a Deutsch on the other.
4) find either a 1642 ( and rewire again.. ) or find a 3320 papertape reader if board, as I would like to have papertape functionality for this machine.

Journey is only starting !
 
Thanks for the explanations to Jos. So:

Ad 1: I will put my 8k core boards back into the 1626 IO processor.
Ad 2: Probably no full schematic is needed; I'd just look at the pins relevant to controling a panel (see wirings scheme posted 8/24)
Ad3: What I can offer would be a new breakout cable just found in my spares box at the very bottom - you could wire that to the mainboard as it is done with the Deutsch plugs and the IO connections right now. You could unscrew one of the original Deutsch receptables and push it back into the chassis and the new Bendix receptable would replace it:

Amp55pin-BreakOut-20211209.jpgAmp55pin-20211209.jpg
That will work if one of your cables fits the panel and the other and matches the receptable shown above. Let me know and I am sure we can make some deal and I will send out the insertion/removal tool together with the breakout cable.

Ad4: Maybe at some time in the future a paper tape board will show up somewhere. Often one gets lucky unexpected ;-)
 
That is most certainly the correct plug for the panel / panel cable i have ! Would be a good way to proceed...
 
(1) The cable shown in my post from 12/9/2021 is on its way to you as well as an insertion/removal tool.
Hope both are of help in your efforts.

(2) I often did not find the tools very useful, because they do not work very well. Inserting is much easier
by using the cable to slide them in and eventually add pushing with a tiny screwdriver. For removing use
the tool with has got the non facceted end and slide this in aloing the cable until the very end (there it
unlocks the spring) and now you may pull at the cable risking that the cable breaks at the crimp; so better
push to the pin/socket from the front. If it is not working the first time, pull the tool and re-insert in a slightly
different orientation (sometimes the holding spring does not get unlocked in the first attempt). Good luck ;-)
 
Finished V1 of my schematic for a RAM board for this ROLM.
Features :
- 64K of SRAM with battery backup.
- read/write SRAM contents via USB to a PC.
- ROM space, insertable in 1K increments.
- Reset / Boot commands via USB ( needs additional wiring in the machine, does not replace the frontpanel )

[File too big to attach, find it here] ftp://ftp.dreesen/Rolm/ROLM_ramcard.pdf

...comments welcomed....
 
Interesting diagram showing the development of the Rolm 16XX product line up until 1977; also showing "where" to find Jos' 1602A in the family-tree:
RolmUntil1977.gif
 
Ready to be tested in th next couple of days : the semiconductor memory for my 1602A....


rolm_ram_t.jpgrolm_ram_m.jpg
 

Attachments

  • rolm_ram_t.jpg
    rolm_ram_t.jpg
    228.3 KB · Views: 12
I am afraid it doesn't... and other equally interesting projects came along ( Datapoint 2200, 5500....)
I noticed one of the AMD2901's in the cpu card had been replaced, don't know if that has anything to do with it...
Anyhow the board needs more time. Quite willing to send you a free bare PCB though !
 
Thanks for the offer, but I am having running core ;-) If you are unsure on the processor board I might be able to borrow you one for testing if you promise sending it back within 2 months. But of course just an offer and if there are other more interesting projects...
 
So, in the unlikely case anyone here has a need for it : this is the reconstructed schematic of the Rolm s375 control panel interface .
This is strictly the interface itself, no added functionality as the 1642 board has . It also is not 100% compatible with the 1642, but should be close enough.
I will now have to add the necessary internal wiring, also replacing one of the Deutsch receptacles with the Bendix one. Afterwards I should be able to use the panel.
 

Attachments

  • Rolm_375.pdf
    362 KB · Views: 3
Hi Jos, thanks for the efforts in creating the schematics of your S375. Quite surprising that they designed a special board with just a CPI interface and no reader, RTC and PTP/R to save cost. Great job in finding out all the connections.
Regarding the two signals you are missing (PSTOP and PRST), that is a simple puzzle: They are directly routed to the processor from the external plug. You can see that in the tables I shared Aug 24, 2021 in this thread: PRST heads to J2-30 and PSTOP to J2-33 of the CPU's J2. Good luch in toasting - . . . - ahhm activating your panel ;-)
 
Back
Top