• Please review our updated Terms and Rules here

Diablo Series 30 connectors

you also have to be careful about addressing because Diablos treat the id pins as individual selects while the rk05 does a 1 of 8 decode
thinking about it, i that may also be an issue between the RK11-C and D
Yes, that's definitely an issue. The RK05 operates in one of two modes based on the state of the BUS_RK11D_L signal:
  • BUS_RK11D_L not active – RK05 interprets the drive select signals as individual enable signals in sort of a one-hot scheme. This is the same selection mechanism used in the Diablo drives. (BUS_RK11D_L is not driven by RK8-E)
  • BUS_RK11D_L is active – RK05 interprets the drive select signals as a 3-bit binary drive address that it compares with the physical drive address. (BUS_RK11D_L is driven active by RK11-D/E)
The RK8-E controller outputs individual drive select signals and doesn't drive BUS_RK11D_L, so Diablo drives can connect to the RK8-E and work in the manner that the RK8-E expects (except for the AC Low and DC Low signals)

In a PDP-11 system with an RK11-D/E (maybe the same for RK11-C but I haven't checked) a Diablo drive could be respond at drive address 1 or 2, but not 0, in a configuration with up to 3 drives. Think about binary address 3: two select lines would be active so that wouldn't work. It's a bit strange but with RK11-D/E, Diablo needs to be configured as address 0 to respond as drive address 1, or configured as address 1 to respond as drive address 2. When the RK11-D/E selects drive address 0, no select lines are active so the Diablo drives won't respond.

We could make a special Diablo Adapter for PDP-11 systems that would have an address decoder chip and re-drivers for the select signals. This would be only the first adapter in the chain which converts binary addresses to individual select signals, and it would only allow Diablo drives to appear at drive addresses 0, 1, 2 or 3. If someone wants to attach more than 4 Diablo drives to an RK11/D/E then some other digital gymnastics seems possible, but that seems like not a very likely configuration.
 
Last edited:
I can 3D print any MRAC housings you want if someone generates the STL file. I am absolutely terrible with solidworks.

I prefer to print in Prusament PETG or ASA but I can also print in PLA for a greater variety of colors.
 
I'll do further work on the OpenSCAD model taking Al's replacement pin dimension notes into account, and put the script on here. In it, all the pin placement is calculated and parametric. The more ideas the merrier!

Regards running the Diablo off an RK11-D, the PDP-11 field guide indicates there was a variant of the M7255, the M7255-YA for the Diablo. I've never seen this variant documented in the RK11 print sets but if I was to guess, it might include a 2-to-4 decoder for the individual drive select lines?
EDIT
I didn't read gwiley's last paragraph above before I posted
 
In a PDP-11 system with an RK11-D/E (maybe the same for RK11-C but I haven't checked) a Diablo drive could be respond at drive address 1 or 2, but not 0, in a configuration with up to 3 drives. Think about binary address 3: two select lines would be active so that wouldn't work. It's a bit strange but with RK11-D/E, Diablo needs to be configured as address 0 to respond as drive address 1, or configured as address 1 to respond as drive address 2. When the RK11-D/E selects drive address 0, no select lines are active so the Diablo drives won't respond.
I actually did implement this on an RK11-C back in the 80's. It worked well under RT11. One RK05 was the OS drive. I used the Diablo and other RK05 for duplicating files or the entire drive.

Jerry
 
the problem is they are soldered to the pcb, so the housing has to be heat-resistant
No idea what others use but for most of my models I've standardised on PETG as it has a relatively high glass temp compared to say PLA/PLA+ and I think it could survive soldering. Previously I had some thoughts of a MRAC housing that is composed of a few individual but mechanically interlocked layers, will put more thought into that.
 
the problem is they are soldered to the pcb, so the housing has to be heat-resistant
I can print with either ASA (better than ABS) or even Prusement PC Blend (also high temperature).

Prusa prints their hot end parts in ASA. And if you can turn the iron down to as low as possible to still melt the solder that might help.

I've never tried nylon the but Prusa is supposed to be able to print it.
 
JLCPCB is currently building these (almost complete):
1716444114120.png
The top one, Adapter P, can be built with an MRAC42P mounted on the bottom side and 4 headers on the top side.
The second one, Adapter P Keystone, can be built with 42 Keystone 1362-2 Micro Pins mounted on the bottom side and 4 headers on the top side. This will be a fun experiment.
The third one, Adapter S, can be built with an MRAC42S mounted on the bottom side and 4 headers on the top side.

Adapter P Keystone gets all of its structure from the pins soldered to the PCB. A rather thin (0.192") plastic piece with 42 holes is planned to fit over the pins. That plastic piece is fastened to the PCB with two #2 sheet metal screws. The only purpose of this plastic piece is to prevent the pins from seating too deeply into the socket connector on the drive. A thumbscrew like this one, 12mm length, holds the adapter to the drive on one side. Haven't figured out the other side.
1716445106446.png
Micropins, headers, and two Meanwell LRS-150F-15 15 volt 150 watt power supplies (to power the Diablo model 31) are on the way from Digikey.
 
Do you need me to print anything for you or do you have it all covered?
Thanks Mike. I think I'm okay with respect to 3D printing. I will ask you for help with testing the emulator as a tester with the Arraid emulators using the Diablo adapter. Will email with details.
 
However I can help.

Thanks.

BTW, I received the tester board. I'm in the middle of putting my 8/A together and then I will test my Plessey Controller and then convert the drive to a tester. Does the tester link you sent me contain instructions on how to use it as a tester?
 
I just realised my Diablo 30 terminator (p/n 11175-03) is slightly different to Al's version (p/n 11174-03) so here it is. There appear to be more resistors on this one.

Diablo_Systems_Inc_Model_30_Disk_Drive_Terminator_pn1175-03_front.png
Diablo_Systems_Inc_Model_30_Disk_Drive_Terminator_pn1175-03_back.png

Diablo_Systems_Inc_Model_30_Disk_Drive_Terminator_pn1175-03_right_side.png
 
I think the difference is male or female pins on the terminator.

Diablo made 6 different terminators (options 35 - 38, 57 & 58).

Male and Female with and without resistors and Male and Female with resistors without erase gate.
 
I will also need at least 2 Diablo-30 terminators. Pins will be taken from a spare Tek4014 cable that could be used to separate the monitor from the cabinet.
Targets will be a Datapoint 9350 disk, and an ICL1503.
Datapoint connectors depicted below :
20240526_121831.jpg20240526_122914.jpg
 
It's funny that last week there was someone selling ten new Diablo cables on ebay and they all sold.
Wasn't aware there was that much demand! O_o
 
It's funny that last week there was someone selling ten new Diablo cables on ebay and they all sold.
Wasn't aware there was that much demand! o_O

I missed them or they were too expensive, I can't remember . I do get tired of ePay prices some times.

People are solely interested in the maximum dollar rather than fostering the vintage computer community with reasonable prices.

Like the guy that want's $9500 (originally $15,000) for a PC04 with PC8E controller.
 
I missed them or they were too expensive
I bought two cables for $60 ea. That is cheaper than you can buy a connector pair for, if you can find them at all.

"People are solely interested in the maximum dollar"

You need to realize 1) he isn't a hobbist 2) you're trying to buy parts that have either been out of production for decades or if they are new, you'll pay MIL prices

The days of picking up a warehouse full of scrap minicomputers ended in the 20th century.

FWIW, the cables are NOS, they were still in the original plastic bag

"DEC chose a more sensible and cheaper path with the RK05 terminators and cabling."

Made sense for DEC at the time. They're Unibus cables and they were building them and Unibus terminators in volume.
Duplicating that flexprint cable is impractical today and the old ones are delaminating
I can't think of any other minicomputer company that used flexprint cables that way.
 
Last edited:
Well, I guess I will have to rely on George and the rest of you to come up with some kind of adapter or substitute.

Did anyone figure out the differences in the terminators?
 
It's funny that last week there was someone selling ten new Diablo cables on ebay and they all sold.
Wasn't aware there was that much demand! o_O
Probably from folks who bought the Diablo drives from the same seller.

I also noticed that the 15-pin socket connectors are gone, which is what you need to connect the external power supply to the Diablo Series 30.
 
Back
Top