PDA

View Full Version : Modcomp Bailey 1055 on eBay



NF6X
September 9th, 2014, 10:52 AM
Much like the newly-announced iWatch, I don't know what it is, I don't know what it does, but I could tell at first glance that I wanted it:

MODCOMP BAILEY 1055 VINTAGE COMPUTER (item 331314313755) (http://www.ebay.com/itm/331314313755)

I'm neither close enough to flush enough with cash to buy this one myself right now, but it sure looks pretty to me! I hope that one of "us" will get it and then share lots of pictures and discussion here.

I'll ask the seller if it's OK for me to download the photos and post them here for posterity and discussion. I doubt they'll mind, but I won't do it without permission.

Chuck(G)
September 9th, 2014, 10:56 AM
"GOLD RECOVERY"

"I AM STARTING THE BIDDING AT THE SCRAP VALUE OF THE BOARDS."

In what universe are these worth $10K after processing?

Frank S
September 9th, 2014, 11:03 AM
Thanks for sharing this auction.
Never seen that dense spacing TTL boards.
Frank

NF6X
September 9th, 2014, 11:15 AM
I left out the "gold recovery" part to prevent a retroriot. :)

JDallas
September 9th, 2014, 11:30 AM
...I don't know what it is, I don't know what it does, but I could tell at first glance that I wanted it...
I've wondered why I haven't seen any ModComp computers before.

In college I worked at Johnson Controls SECD-Southwest in the engineering co-op program and worked with these things all the time. JCI did control centers for waste-water treatment plants and used ModComp very heavily. I had to build these up, test them and interconnect them in a mock-up area so the coders could start writing the system right away. The last job I did before graduating was "The Columbus Project" (i.e. Columbus Ohio) that used a ModComp distributive computing system instead. I think I had four or five of those things interconnected as they would be later on-site.

The worst part of that job though was running RS232 cables from the mock-up area to the programmers office up through the ceiling panels. Fiberglas itch was unavoidable.

I just figured ModComp vintage computers weren't around because they were put to controls systems use and industry carries those as far as humanly possible to avoid the cost of a new control system design.

Glad to see that Modcomp... they always did look great. IMSAI's big brother. :)


...Thanks for sharing this auction. Never seen that dense spacing TTL boards....
As unbelievable as this sounds, ModComp used wire-wrapped boards (automated obviously) on a lot of the system boards. It was a quality job... first rate sockets solder tapped to the board.

I remember having to rework some of the wirewrap on one set of boards to support Grinnell color display monitors.

The media of choice then was 14-inch platters usually installed in the base of the rack (less vibration, sway and motion being close to the floor pads.)

MikeModifed
September 9th, 2014, 12:22 PM
Located in Kennewick, WA; used by Department of Energy: came from Hanford?

Mike

KC9UDX
September 9th, 2014, 12:33 PM
But it's assumed working. Surely that raises the price... :P

Al Kossow
September 9th, 2014, 05:38 PM
MODCOMP BAILEY 1055 VINTAGE COMPUTER

It is a Modcomp Classic

https://www.flickr.com/photos/magnus_osterlund/238688862/

Bailey Controls is a process control company. I'm sure they OEMed it.

olePigeon
September 9th, 2014, 05:53 PM
There was a guy on Craigslist in Los Angeles not too long ago that was attempting to sell a brand new, unopened original 128k Macintosh for gold scrap (advertised as 68000 GOLD or something.) I emailed him to not scrap it, as it was worth more as a computer than the gold he was trying to get from it. He then went off the handle about nerds, virgins, driving Alfa Romeros, etc., then sent me pictures of him destroying the computer. Might still have the email somewhere. It was the weirdest thing I've ever read.

Pretty much sums up my impression of the gold scrapping industry.

NF6X
September 9th, 2014, 06:22 PM
The seller gave permission to share the pictures here, as well as a preview of coming attractions:


Yes, that would be fine with me, I have about 14 cabinets I will be putting up for auction, there are about 4 of these units, and the rest are disk drives, tape drives, etc. I also have a couple of the large keyboard units.

Here are five pictures that I picked out from the listing:

20400 20401 20402 20403 20404

NeXT
September 9th, 2014, 11:02 PM
I was literally just in Kennewick. :I

That's a very nice looking machine but 7K? Keep dreaming. Won't lie about the condition. It's almost pristine. I bet you it will start on the first try.

MattisLind
September 9th, 2014, 11:06 PM
It is a Modcomp Classic

https://www.flickr.com/photos/magnus_osterlund/238688862/

Bailey Controls is a process control company. I'm sure they OEMed it.

In Sweden Modcomp machines were used by ASEA (ABB) at nuclear power stations. The ones depicted in Al's link above is from a test system for these control systems. The picture are a few years old. A couple of years ago the guy who took the photos asked if we were interested in a couple of incomplete Modcomp system that had been part of this test system. Incomplete since they had to save some spare parts for their other systems. We didn't take them.

There are very little information regarding Modcomp on bitsavers. I could ask Magnus if they saved the documentation if there are any interest.

Qbus
September 12th, 2014, 05:36 AM
Have to wonder why they were produced? They look like real big 16 bit systems that were something along the line of a PDP-11 but appear to have more internal room and maybe a later construction date although there is no sign of any LSI being used so almost looks like an early to mid-seventies design. Were they built as an alternative to the DEC and Data General Family? Looking at the construction would assume that they had to cost more than the stuff that was on the commercial market at the time and looking at the data center pictures also see that they had their own terminals and printers.
Would be great to own something like that but just imagine the problems with trying to find documentation and any useful software like an operating system or interpreter, maybe it’s just because I am stupid or whatever but the acid test for me in buying antique systems is will run Basic, being that’s the only language I can work in. Just try to imagine finding a Basic compiler for something like that.

KC9UDX
September 12th, 2014, 06:11 AM
Just try to imagine finding a Basic compiler for something like that.

Buy it and hire me to write the compiler. That would be fun!

JDallas
September 12th, 2014, 06:46 AM
...wonder why they were produced?...big 16 bit systems...
Probably produced because of their great selection of I/O; they seemed to be the choice for control systems.


...no sign of any LSI being used...
Actually one of the photos does show LSI chips. However most of the photos looked like memory boards with what were probably DRams of the day.


...the acid test for me in buying antique systems is will run Basic...Just try to imagine finding a Basic compiler for something like that.
At Johnson Controls (SECD-SW) we used ModComp computers for process control systems and the programmers wrote in assembly language.

MattisLind
September 12th, 2014, 06:57 AM
Actually one of the photos does show LSI chips.
I would guess that those are Am2901 four bit bitslice ALU ICs. Together with 74181 and Intel 3002 the most common design choice for the ALU at the time when building minis.

JDallas
September 12th, 2014, 01:08 PM
...those are Am2901 four bit bitslice ALU ICs...
Sounds right. I remember reading up on those before I graduated. One of the Johnson Control engineers that worked with ModComps talked endlessly about doing his own bit-slice computer design. I kidded him that if a bit-slice design was in his future, he would have bought an S-100 computer for his home instead of an Apple II. :)

Another lesson learned on bit-slicing:
After graduation, I took a job with a company whose competitor went the Am2901 bit-slice route on a new product design. We looked into bit-slicing but my mentor there said that bit-slicing was a bad choice because it created too many choices that would only delay the design. I designed a state-machine solution (Signetics programmable part) to control all the bus DMA and winchester disk data flow and used a 8085 to do all the setup and I/O parameter block command support.

My mentor appeared to be correct because I had our new board done a year before the competitor got to market. A year later the competitor product appeared and it was a two-board Muliti-Bus set that offered no additional features, just bulk.

So not only did it take them twice as long to design their product, it was too big to sell.

Al Kossow
September 12th, 2014, 01:46 PM
I designed a state-machine solution (Signetics programmable part) to control all the bus DMA and winchester disk data flow and used a 8085 to do all the setup and I/O parameter block command support.

Was this board OEMed? What sorts of disks? Sort of sounds like several different designs I can think of from the early 80's.

MattisLind
September 13th, 2014, 12:28 AM
Another lesson learned on bit-slicing:
After graduation, I took a job with a company whose competitor went the Am2901 bit-slice route on a new product design. We looked into bit-slicing but my mentor there said that bit-slicing was a bad choice because it created too many choices that would only delay the design. I designed a state-machine solution (Signetics programmable part) to control all the bus DMA and winchester disk data flow and used a 8085 to do all the setup and I/O parameter block command support.


Sometimes one has other things to take into account: Companies like Emulex (who used Am2901 heavily) had a whole product line with different Unibus or Qbus controllers. Instead of tailor each controller to each interface using a state machine you simply modify the microcode and adapt the some interfacing logic. Then you can easily handle SMD, ESDi, MFM, PERTEC etc mostly using the same board. A sort of IP that you just put in.

But it will be a more complex and more expensive product. On the other hand a one off product certainly would have been more cost effective using the Signetics PLS chips.

Signetics, by the way had the 8X305 bipolar series which had some use in controllers. I have bunches of those if someone wants to waste time and play with ancient bipolar controllers...

JDallas
September 13th, 2014, 06:17 AM
...Signetics, by the way had the 8X305 bipolar series...
Yes, we were a 8x300 design shop mostly; the company had used it in video drivers, navy map protocol converters and winchester drives in the late 70s. The company had one of the few 8x300 emulators from company that originally made the processor, from whom later Signetics bought the rights.

The design I mentioned was for a faster version of winchesters for which an 8x300 in the data stream was no longer feasible so I had to design a new architecture. My mentor there, one of the owners, made the mistake of specifying that the board would be a two-sided board with no power and ground plane to improve margins as they had done that with all previous boards. However he had to make a second layout version of it with power and ground planes. At this time printed circuit board layouts were still done by tape and mylar. :)

The problem with the 8x300 for new designs at that time was that there was no support platforms from Signetics yet. Getting an emulator at that time was extremely difficult. To compile 8x300 code, we used an Intel ISIS development system's 8080 assembler, using macros for the 8x300 opcodes. The listings were a mess but its code was true.

I hopped to S.D.System to design their Versa-Floppy-Winchester III for the IEEE-696. I recommended using an 8x300 because would have been a better design and it was certainly fast enough for those drive interfaces. But S.D.Systems was locked into the WD1010 chip set and I think they had already invested a lot of time waiting for WD to stop changing their design of the 1010/1014/1015 chip set so the board could be started. The board was so tightly packed with chips that we had to use CAD layout from the same firm in Atlanta that did D.C.Hayes boards to get it to fit on an S-100 board.

When Western Digital later started using 8x300 on the WD100x series, I assumed that was due to one of the following: (1) the same company designed them for WD, (2) an employee from the company hopped to WD and designed it for them or (3) that Signetics had fielded a new emulator and assembler for WD to use.

MattisLind
September 13th, 2014, 06:38 AM
...At this time printed circuit board layouts were still done by tape and mylar. :) ...

I have only done some hand layout in school in the early nineties. For very small boards. It just have to take massive amount of time, right? Rip up and retry? I have this big board in an old Incoterm terminal. Appears to be done by hand. Impressive. http://i.imgur.com/pMOt09E.jpg

JDallas
September 13th, 2014, 07:28 AM
...At this time printed circuit board layouts were still done by tape and mylar...

...It just have to take massive amount of time, right? Rip up and retry?...
Back then we had contractors that would do all our circuit board layouts. They'd use a large drafting table sometimes a light table with sheets of mylar onto which they'd add black tape and stickers special made for circuit board layout. The original were something like 8x the size of the board, so that it was easier to work with, except for flipping to layers of mylar. The final gerber prints were made photographically from the mylar and tape and reduced to actual size.

I only taped two small boards in those years and both at 1x instead of 8x size. One was a special keyboard layout for a control panel, with a 8048 I programed, for the Columbus Project at Johnson Controls where I worked as a co-op student during college (alternating semesters between full-time college and full-time work with night classes) and another for small board on a contract job-on-the-side.

Problems I've Had With PCB Layout Contractors:
I've done a lot of my own board design layouts in CAD since 1998. As I studied drafting in the mid 70s and did drafting work at Texas Instruments before going to college, its an easy thing for me and I can do it a lot faster and with less compromise than the PCB layout contractors I tried to use. I find that pcb layout contractors say something is "impossible" when they mean that wouldn't be easy and as profitable for them to do. That I did it anyway proves it wasn't impossible. :)

PCB Layout contractors also try to make each board of an assembly a separate contract - no way! Instead I bought PADs and did it myself with all the boards on the same panel with a snap-line to separate them AFTER SMT pick-and-place automated assembly, through-hole item insertion and wave-solder. The contractor's method would have also added extra labor hand-soldering the board connecting flex cables whereas my solution had it done in a single pass, tested, and *then* snapped apart and assembled onto the mechanical mounts.

In other words, all the flex cables were done by the single wave solder pass and the boards were all electrically connected allowing them to be tested before snapping them apart. My first board in PADs was a 6 layer panel with five snap-off boards, two or which were small external connector adapters. The prototype run included a test fixture card snap-off used to program the micro's firmware. I learned how to squeeze cost out of a design at Tandy. :)

Mylar & Tape Layout Methodology:
From what I saw of our mylar and tape pcb layout contractor back in 1982 doing board layouts on mylar and tape, it involved a lot of heuristics on what you do first and where you put certain traces on the board so that you minimized ripping up tape when you'd otherwise hit a dead-end. They'd sketch difficult traces through the layers before going back and taping it. They'd work over a grid behind the mylar and use registration marks so the mylar could be carefully aligned for vias (he may have used a thin punch or X-Acto blade point to tack a position through the mylar for aligning vias). Our boards didn't use wavy traces, they used 45 degree radials. This type of printed circuit board layout was fascinating to see, but it was a very slow process.

In contrast, the board I did at S.D.Systems was packed with chips... hardly any space between the sockets. Initially we used a contractor to do the layout by hand but after about 3 months (at least it seemed that long if it wasn't) he was stuck. To get a board we allowed him to put a patch of one of the planes on the back of the board and he delivered the entire power plane on the solderside - which under no circumstance would be acceptable. We punted that and had a 6 layer board (including the power and ground planes) done by CAD layout in Atlanta. They did a wonderful job.

Trinkets From The Age Of Hand-Layout PCBs:
I think I've got the prototype we built of the hand layout of the S.D.Systems board; someone did a full chip build of it anyway. I may have some of the final layout's 6x4 foot gerber prints of each layer too. I used to hang the set on the wall behind my desk as technology art.

Qbus
September 16th, 2014, 05:18 AM
Looks like nobody stepped up and bought it, hate to think because of its size and location this thing is going to be cut up for scrape value. What do you suppose the scrap or gold recovery value of something like that is?
At some point it’s important to know that sort of information because that gives an idea of what someone can offer the sellers of this sort of hardware where the seller will still make a profit beyond scraping the system.
Back in the eighties during the first gold boom remember visiting a place up in Philadelphia where they were recycling equipment for gold and it was no small process. Everything had to be disassembled, broken apart and then rolled in drums for some amount of time so there was a lot of labor and floor space involved. At that time they were getting new IBM system 360 that were all pelleted and ready for shipping and chopping them up, along with tons of other military and government stuff.

JDallas
September 16th, 2014, 06:01 AM
...Looks like nobody...bought it, hate to think...this thing is going to be cut up for scrap value.
That isn't a gold scrap treasure, its a vintage chips treasure. Each ModComp Classic board is literally packed to integrated circuits. That use for it is at another price-break, certainly not at $7K. Either they'll donate it, scrap it, or re-list it at a more reasonable price. Anyone interested at $2K?

NeXT
September 16th, 2014, 07:29 AM
With a lot of effort I could dismantle it and cram it in my car but it was simply far too expensive for me to justify.

Marty
September 23rd, 2014, 02:32 PM
Hi All;
I agree with JDallas, "" That isn't a gold scrap treasure, its a vintage chips treasure. ""
And IF I had the money I would save it for the IC's, and of course trying to get it working again..
But, it would be near impossible telling that to the guy who is selling it, He is only looking to make the Big Bucks off of it..
Thanks, JDallas..

THANK YOU Marty

Qbus
September 24th, 2014, 08:44 AM
Its back, and this time with no reserve and a bid.

http://www.ebay.com/itm/MODCOMP-BAILEY-1055-VINTAGE-COMPUTER-SERVER-MAINFRAME-DEC-DIGITAL-GOLD-RECOVERY-/261600376739?pt=US_Vintage_Computers_Mainframes&hash=item3ce898f7a3

There is also an 11/04 for $500 but they never get the love that the binky light systems get. I have an 11/34 have to say that the octal keypad is surprisingly functional, way more fun than the two switches and three lights on the 11/23. Someday hope to be one of the “in” crowd and have something like an 11/05 with switches and lights. Perhaps that’s the appeal of the Bailey?

NeXT
September 24th, 2014, 09:38 AM
Whoever buys it I hope they know they're getting the raw CPU and not a lot else, unless someone fell for the GOLD RECOVERY title and is about to be screwed into buying up to $150 in scrap gold and $850 in scrap steel, in which case I can call them a tool.

JDallas
September 24th, 2014, 01:18 PM
Instead on pitching gold recovery, they should have pitched it was a 16 bit-slice evaluation board.
Should include a note: "Buyer must be able to read schematics to write his own custom assembler."

Marty
September 25th, 2014, 01:47 PM
Hi All;
"" they should have pitched it was a 16 bit-slice evaluation board. ""
Seeing a Bit-Slice machine, I decided to get out my 2900 Evaluation and Learning Board.. And after Replacing a Switch, and 7 Leds, As far as I can tell it still works..
I have Not Located the Manual for it, So I don't know how to Set it up and Run it..
Does anyone know how to set one of these up and see if it still working..

THANK YOU Marty

dave_m
September 25th, 2014, 03:35 PM
Seeing a Bit-Slice machine, I decided to get out my 2900 Evaluation and Learning Board.. And after Replacing a Switch, and 7 Leds, As far as I can tell it still works..
I have Not Located the Manual for it, So I don't know how to Set it up and Run it..
Does anyone know how to set one of these up and see if it still working..



There is good documentation including the manual:
http://www.donnamaie.com/AMD_Vintage/AMD_2900_ED2900A.html

Here is a schematic:

http://www.giga.nl/walter/computers/Am2900.pdf

Marty
September 25th, 2014, 04:14 PM
Hi All;
Thank You Dave M, I had already been to that site, But for some reason , I either didn't see the rest of the listing, or I din't page down enough on that page, or for some reason I got a limited version of that page..
I did have the schematic copied, which helped me find out what a missing Ic I had was..
I got it, now to look at it.. THANK YOU Dave !!!
It doesn't seem to work, will check all the standard stuff, corrosion, and bent pins and bad Ic's..
No corrosion, and no bent pins.. I will follow the check out procedure in the assembly instructions..
It looks like I got it working, alot of little things, nothing major..

THANK YOU Marty

toober
September 27th, 2014, 05:35 PM
Whoever buys it I hope they know they're getting the raw CPU and not a lot else, unless someone fell for the GOLD RECOVERY title and is about to be screwed into buying up to $150 in scrap gold and $850 in scrap steel, in which case I can call them a tool.

From what I see of the pictures, $1000 in gold scrap is quite reasonable.

The stuff isn't $250/oz anymore!

--
Will