• Please review our updated Terms and Rules here

DEC units on ebay

Signetics N8881N on ebay:
Ebay

“Item condition: New, old stock.”

This may be of interest for some of us? I bought some and my LEAPER says PASS.
But I have not used them yet, came today.

Volker
 
gnupublic said:
Signetics N8881N on ebay: This may be of interest for some of us? I bought some and my LEAPER says PASS.
But I have not used them yet, came today.

Volker
Click to expand...

Thanks for the tip. I've bought one lot. For that price I can take the risk :) It might be interesting to test a chip under load. I have a few originals in stock, so I can compare them...

Regards, Roland
 
Al Kossow said:
A pdp-11 clone. killed by DEC because they violated the Unibus patent
The CPU is a four board set with over the top connectors
Click to expand...
I assume DEC only played hardball because the company was building clone CPUs? DEC in that era was pretty much "We build CPUs, memory, and a couple types of I/O controller - do your own thing if you want to make your own controllers".

I'd also assume that the only time a company needed to actually approach DEC to license peripheral controller technology is if they wanted the [T]MSCP specs.

Later, there were 3rd-party CPU boards built out of DCJ11 CPUs. For the Unibus, I only know of one being sold by itself as an upgrade, and that was the Nissho upgrade for the 11/24. I was approached by RSTS Engineering and asked "Can you handle making the RSTS/E patches for them for each release?" because at least that part of DEC didn't want to be seen as directly supporting 3rd-party CPUS. ATEX also made a Unibus DCJ11-based CPU, but it was limited to their typesetting systems.

There were lots of Q-bus DCJ11-based CPU boards from other companies - Mentec (who ended up getting the rights to a lot of PDP-11 software from DEC), ROI (who later went on to design what became the "DEC" 11/93 CPU), etc.

The J-11 product manager (a very nice woman named Cathy Berida at the time) had the thankless task of calling anyone who ordered a DCJ11 CPU to ask what the customer was going to do with it - even for Qty 1 sales. That's because DEC couldn't decide if the DCJ11 was a merchant semiconductor product or intellectual property to only be used in DEC systems. This is far from the only time DEC grappled with that dilemma.

DEC didn't start playing hardball until the BI bus came along, and insisted everybody license the technology, purchase a BIIC for each board sold, and implement the "BI corner" of the board exactly to DEC specs. And that certainly killed off a lot of the aftermarket controller business for those systems. Though DEC was making a much wider range of peripherals and controllers by that late in the game.
 
I don't think i've seen the expansion board for the E&M 8080 trainer before
I'm also curious what the IBM modules look like
Wonder where this stuff came from. Shame there isn't more of the 12 there
Those M-series connector blocks are a good find
 
That is a lot of stuff. The 8/e in the right cabinet doesn't have any memory or processor cards, that much is obvious. I can't identify any of those boards from the photo but no top block connectors means none of those are multi board controllers like an RK8E. It is really tough to say if this is a good value. My gut feeling is that it is probably worth the asking price but it would be nice to see detail of the card loadout in both 8/e chassis.
 
It's the DECtape transport and RK05 drives which would tempt me if I had enough cash to burn on it, even if the 8E's best bits have been plundered.
 
rk-2.png

what's going on with the RK05 drive numbering? The two RK05J have RK05F-style drive numbering and the RK05F has four drive numbers? It gives the impression we're looking at removeable double-density RK05J's and a 'quad density' RK05F?
The RK8-E controller could only address unit 0 - 3. The RKS8-E (or RK8-L) controllers could address 0 - 7.
 
Last edited:
rk-3.PNG

Does anyone recognize that half-size controller in the bottom of the full rack? FTI logo? Maybe Forward Technology Inc.? maybe Multibus boards?
Could it be a 'special' controller that somehow doubles the storage density of the RK05's? Or maybe it really is just a collection 'for parts'.
 
ftcnet said:
View attachment 1297971

Does anyone recognize that half-size controller in the bottom of the full rack? FTI logo? Maybe Forward Technology Inc.? maybe Multibus boards?
Could it be a 'special' controller that somehow doubles the storage density of the RK05's? Or maybe it really is just a collection 'for parts'.
Click to expand...

Interesting. Maybe the drives are split in smaller parts for an os which didn't support the full size? Or split so every user had its own part?
 
ftcnet said:
Does anyone recognize that half-size controller in the bottom of the full rack? FTI logo?
Click to expand...
It's the Fabri-Tek logo. It's gotta be a core memory expansion system. Bitsavers shows the Model 8 for the 8/I and 12 which looks virtually the same, but perhaps they made an Omnibus one, too.
 
ftcnet said:
what's going on with the RK05 drive numbering? The two RK05J have RK05F-style drive numbering and the RK05F has four drive numbers? It gives the impression we're looking at removeable double-density RK05J's and a 'quad density' RK05F?
The RK8-E controller could only address unit 0 - 3. The RKS8-E (or RK8-L) controllers could address 0 - 7.
Click to expand...

Roland Huisman said:
Interesting. Maybe the drives are split in smaller parts for an os which didn't support the full size? Or split so every user had its own part?
Click to expand...
OS/8 can only handle a device with up to 4095 blocks (of 256 12 bit words). A removable RK05 exceeds this (6496 blocks) so it was split into two smaller devices of size 3248.

My understanding is that he RK05F, because it had a fixed platter was able to operate at twice the track density so it would need to be at least four devices (a little over 3 full size devices actually) from an OS/8 standpoint.

I don't have an OS/8 handler source (or binary) for the double track F version in my collection.
 
DougIngraham said:
My understanding is that he RK05F, because it had a fixed platter was able to operate at twice the track density so it would need to be at least four devices (a little over 3 full size devices actually) from an OS/8 standpoint.
Click to expand...

That is true, but the RK05J is split into A and B parts under OS/8 with just one drive select number. The RK05F responds to two drive select numbers, with each split into the A and B parts.

Maybe they used a special OS/8 device driver that didn't use the A and B parts of a single drive select number, and used two OS/8 drive numbers for each drive select number?
 
m_thompson said:
That is true, but the RK05J is split into A and B parts under OS/8 with just one drive select number. The RK05F responds to two drive select numbers, with each split into the A and B parts.

Maybe they used a special OS/8 device driver that didn't use the A and B parts of a single drive select number, and used two OS/8 drive numbers for each drive select number?
Click to expand...

That A/B construction is exactly the reason why I thought about something else than OS8...
 
Roland Huisman said:
That A/B construction is exactly the reason why I thought about something else than OS8...
Click to expand...
rk-2.PNG

I agree, there may be other than OS/8 software here. Can you read the label on the middle pack? I think it looks like MULTOS SYSTEM

Thanks to antiquekid3 for identifying the FTI logo. It seems bizarre to adapt a PDP-10 8K - 36-bit core plane to provide 24K of 12-bit memory for an Omnibus system.
 
You must log in or register to reply here.
Back
Top