PDA

View Full Version : Got me a Columbia Data Products MPC 1600-1!



SomeGuy
December 21st, 2013, 11:17 AM
Well, I'm getting really excited about this.

In a moment of weakness I decided to buy a Columbia Data Product 1600-1 of off eBay. No idea where I am going to put it :)

Surprisingly there isn't much information on the net about these, so I will post some here. Reportedly it was one of - if not THE - first IBM PC hardware compatible clones.

I know you people here like to tout your Genuine IBM products, but back then it must have taken some real balls to try and rip off Big Blue. In today's litigious society they would have been shut down faster than you can say "infringement"!

A quick look inside shows date codes in 1983 on most of the parts. In theory, an earlier 1982 model might be more revealing.

There was no software with it, except a non-bootable backup disk labeled "Fast Graphs" left in the drive.

Stupidly, I missed an opportunity to get a set of Columbia Data Products OEM MS-DOS 2.11 disks off of eBay earlier -
http://www.ebay.com/itm/7-Vintage-1984-Software-Columbia-Data-PC-Computer-5-25-Floppy-Format-Disks-/121228138598


Hmm, were they also the first Microsoft DOS OEM?

----

164271642916426

So anyway I opened it up and it looks surprisingly clean on the inside. The power supply covers part of the motherboard. I'll try to pull it out in a bit and get some better pictures.

This machine has a CGA compatible video card, and a memory expansion card. It has two full height double sided 5.25" drives, and the disk controller is built in to the motherboard. No cassette port! :(

I don't have an RGB monitor, but it has a composite out, so for the moment I will just hook it to my TV. I also have a composite to VGA converter that seems to work OK.

It powers up ok, but the video card seems to have a bad RAM chip.
With no disk in the drive it displays the message "@LEACE INCEBD A BOODABLE DICK" :rofl:

Even though I don't have an OEM boot disk, it seems to boot plain old MS-DOS 3.3 just fine. (Except I can't read any of the text)

16428

Here it is booting Microsoft Flight Simulator II - I expect the graphics are a bit off due to the faulty video chip. Also, my TV doesn't seem to like the color video, it has a bunch of wavy lines although it is stable in monochrome mode. I'll just use my Composite to VGA converter instead.

16430

And Pinball construction set

Ironically the thing came shipped in a big box for some huge brand new "HP TouchSmart" touch screen system. What kind of retard would actually use a touch screen on a monitor that big? Yea, I know.

Still a few things to figure out. The memory card is an "Apparat 512k ram card", but Google turns up little. Does anyone have the manual for this? Also, there is a port on the back that looks like a serial port, labeled "console". What is that for? Also, since most reports don't mention the MPC as a "100% compatible", what exactly is incompatible about it?

Any other CDP collectors out there?

Stone
December 21st, 2013, 12:05 PM
I know you people here like to tout your Genuine IBM products, but back then it must have taken some real balls to try and rip off Big Blue. In today's litigious society they would have been shut down faster than you can say "infringement"!What did they rip off?

pearce_jj
December 21st, 2013, 12:15 PM
AFAIK most of the PC 5150 design was open source - after all, the schematics and BIOS source were in the tech ref. But certain parts were copyright, the keyboard interface for example.

MikeModifed
December 21st, 2013, 12:20 PM
Apparat (the only one that I know of) started out with NewDOS/80 for the TRS-80 Model I (an update of the original, miserable RadioShack operating system). As the IBM phased in and the TRS's out, they attempted to branch out into hardware. I have two Apparat Spool/64 printer spoolers sitting on a shelf.

The same company??

Mike

SomeGuy
December 21st, 2013, 12:48 PM
What did they rip off?
That is actually a very good question. Possibly nothing at all.

Most of the old articles I can find about CDP tend badmouth it and suggest that the BIOS was not clean-room reverse engineered like Compaq's, or other such FUD.

Just a glance at the BIOS, it is not a copy of the IBM BIOS (unlike Apple II clones that usually had exact copies of the Apple II ROMs). It looks minimal, and like later clones lacks ROM Basic. It also appears to be a later release marked 1983 (The earliest would have been 1982).

Chuck(G)
December 21st, 2013, 03:06 PM
Not the most uncommon CDP system by a long shot.

The Commander 964 preceded the 1600 and used twin Z80As. If you read the company history, it's as if anything before the 1600 was a black hole. Sad, really.

SomeGuy
December 21st, 2013, 03:08 PM
Now, this is interesting - looking at the disk that was in the drive, even though it does not have system files, it has a copy of command.com that claims to be "v1.17". As the hand written label suggest, the disk also contains the 1982 "Fast Graphs" software.

16432

In the boot sector there is a date of "August-3-1982", an OEM signature of "[CDP inc.]", the files it looks for are IBMIO.COM and IBMDOS.COM, and instead of Robert O'rear, the name "Michael D. Luke" (Lukeu?) appears. Who is that?

16434 16433
I located and replaced the bad video ram chip. So it seems to work right now. No more of this gibberish.

16435
NYAN!

Guess it is set to expect a monochrome monitor. Will have to play with the DIP switches.

mbbrutman
December 21st, 2013, 03:11 PM
AFAIK most of the PC 5150 design was open source - after all, the schematics and BIOS source were in the tech ref. But certain parts were copyright, the keyboard interface for example.

What? References please?

None of it was open source. Published technical specifications != open source.

The BIOS was definitely copyrighted. Compaq was the first company to successfully reverse engineer the BIOS and sell a reasonably compatible clone. The BIOS ROMs also included the BASIC interpreter from Microsoft, which was protected by their copyright.

orion24
December 21st, 2013, 07:38 PM
I know you people here like to tout your Genuine IBM products, but back then it must have taken some real balls to try and rip off Big Blue. In today's litigious society they would have been shut down faster than you can say "infringement"!


I'm sure IBM regreted many of its past decissions, but history cannot be changed. IMO IBM got out of business because it failed to convince that it could build a better PC than the clones. I enjoyed watching these happening when I was a kid.

SomeGuy
December 22nd, 2013, 03:59 PM
16447164481644916446
A few more pictures, showing the full motherboard, memory card, video card, and power supply.

I managed to bump the RAM up to a full 640k by filling up the memory card. Each switch in the 8-switch set just indicates if a bank in installed or not (0-7 from left to right). The 4-switch set must be the start address, which is already set appropriate for the 128k on the motherboard.

The thing about these early clones is that they were charting unexplored waters here. IBM's specs may have been "open", but it wasn't fully hashed out what was/wasn't permissible to copy or re-implement, or how it had to be done. And even if something was technically permitted, who knew if legs would mysteriously start breaking?

barythrin
December 23rd, 2013, 08:41 AM
The reference CDP "ripped off" was the IBM BIOS by making one of the first but unauthorized IBM compatible clones. That got them sued pretty much to oblivion by IBM. Compaq did the appropriate thing and kept their developers out of the IBM systems and just told them what the interrupts did then had them code the equivalent from scratch to prevent any copyright infringement and then became their own history with a huge amount of sales from their cheaper but fully compatible IBM clones.

Yes the IBM PC was made from off the shelf parts, making it a great target to be able to be cloned but the firmware/bios was what IBM wanted to keep private.

SomeGuy
December 25th, 2013, 07:09 AM
16490
Now I have it hooked up with the composite to VGA converter, and it runs the Seirra Xmas card demo from floppy very nicely. Lots of colors for CGA.

A few other things I discovered. This machine doesn't use a twisted cable, so it runs the motor on both floppies when either is active. I tried feeding it an 8-bit RLL hard drive controller and it recognized that fine, but when I tried an 8-bit compatible VGA card didn't seem to like that.

Also, while playing with the motherboard DIP switches, one of them activated a "remote bootstrap" feature in the BIOS. When powered on it would just display a message that is was remote bootstapping and just sit there, probably expecting data from a serial port or something. No idea how to use that, but interesting!

mbbrutman
December 25th, 2013, 08:07 AM
The first version of the PC BIOS did not look for BIOS ROM extensions. So if you added a hard drive controller or a card with it's own BIOS it would not be found at boot time, basically rendering the card as not usable unless you had a different way (outside of BIOS) to access it.

I am kind of surprised the RLL controller was recognized because that would imply a BIOS that looks for BIOS ROM extensions. Maybe the Columbia is late enough where it is using a more modern BIOS that knows to scan for those. If that is the case then the VGA card should have been detected, but there might be other reasons for it not working.

vwestlife
December 25th, 2013, 08:56 AM
If that is the case then the VGA card should have been detected, but there might be other reasons for it not working.

Probably he would need to change the motherboard DIP switches to indicate "no video" in order for the VGA card to be recognized, just like on a real IBM PC or XT.

SomeGuy
December 31st, 2013, 01:21 PM
This is hilarious. I dug out my Video 7 VGA 1024i card, which I know works well on 8088 machines, to try on the MPC 1600 instead of the Paradise card I tried before. Of course it still didn't do anything, but I grabbed the manual to look up the Video 7 dip switches, and it almost opened right up to a system installation page that reads:

"Columbia 1600: Must have Columbia BIOS ROM version 4.36 or later."

I never noticed that manual had anything about CDP computers before! :confused5:

And here I am with 4.34. So close! And all I need is a BIOS update. This is why BIOSes and manuals need to be archived.

SomeGuy
January 4th, 2014, 10:01 AM
16691

A couple more notes, DIP switch S2-6 enables a serial console mode.

In this mode, all BIOS I/O is redirected to the second COM port (at 19200-7-N). On the back, COM2 is labeled "console".

The 1600 uses female DB25 ports, instead of the typical male DB25, but a simple adapter fixes that.

It also uses an oddball parallel plug (female ~DB50, didn't count the pins), but fortunately mine came with the cable to connect to a Centronics parallel printer.

Also, the 4.34 BIOS images have been added to minuszerodegrees if anyone is interested.

Chuck(G)
January 4th, 2014, 10:29 AM
A DC-37 perhaps? A DD-50 has three rows of pins.

JOEZONE2000
January 4th, 2018, 11:57 AM
I have a Columbia MPC 1600=1 with the original software and boot disks, including the CPM boot disk, and Perfect Software. If you have any questions, let me know.

JOEZONE2000

SomeGuy
October 25th, 2018, 12:25 PM
After quite some time, the Columbia Data Products 1600 operations manual is now online. (https://archive.org/details/ColumbiaDataProduct1600MPCOperationsGuide)

The manual did reveal some surprising tidbits that I did not know before.

First, a figure in the manual suggests CDP OEM MS-DOS 1.25 R2.05 existed. What is out there now is 1.25 R2.11.

The manual also states that there were originally four 4k ROM chips. (16k of ROM instead of 12k?).

The manual does discuss operation via a serial terminal connected to the console serial port, so this was a supported feature.

It also discusses the use of CDP's optional 8" floppy drive ISA controller board. Apparently this is supported natively with CDP's DOS 1.25.

It has pinouts documented for the seemingly over sized "DB-37" parallel connector on the back of the unit. After thinking about it, Centronics plugs are "36-pin micro ribbon connectors", so technically the larger connector matches the plug, although only a few wires were ever used in the Centronics parallel interface design. Back then it may have been IBM that was the oddball.

Most surprisingly, the manual states that CDP DOS (apparently 1.25) supports a hard disk! This would have been with CDPs proprietary hard disk controller. Their controller was not an ISA card. The manual identifies a special connector on the motherboard as a connector to their controller board. The manual documents the hardware programming operation of their controller.

The "define" command - visible on the CDP 1.25 DOS disks but with no built in help - is used to set the hard disk type.

This suggests that this special hardware support may be built in to the later MS-DOS 2.11 and MS-DOS 3.2 CDP OEM software.

Yea, sometimes it helps to RTFM.

Chuck(G)
October 25th, 2018, 03:15 PM
It has pinouts documented for the seemingly over sized "DB-37" parallel connector on the back of the unit. After thinking about it, Centronics plugs are "36-pin micro ribbon connectors", so technically the larger connector matches the plug, although only a few wires were ever used in the Centronics parallel interface design. Back then it may have been IBM that was the oddball.

I don't follow--either this is a blue ribbon ("Centronics") type with 36 connections the 50-conductor variety or a "D-subminiature" with 50 or 37 pins (DC37 or DD50) or something else. There was also the "Dataproducts" interface. The "official" Centronics interface has 16 active signals, with corresponding returns.

Some Dataproducts equipment used the 50-pin "Winchester" connector:

http://www.surplussales.com/Images/ComputerAccess/Military/con-MRE50SG_2_lg.jpg

vwestlife
October 25th, 2018, 03:46 PM
I don't follow--either this is a blue ribbon ("Centronics") type with 36 connections the 50-conductor variety or a "D-subminiature" with 50 or 37 pins (DC37 or DD50) or something else.
Here's how it looks in the manual:

https://i.postimg.cc/kXPvCWd0/cdprear.png

It shows a DC-37 (colloquially "DB-37") connector for the parallel port.

https://showmecables-static.scdn3.secure.raxcdn.com/media/catalog/product/cache/c687aa7517cf01e65c009f6943c2b1e9/d/b/db37-female-crimp-connector-1237s-1.jpg

SomeGuy
October 25th, 2018, 05:09 PM
That's why I put "DB-37" in quotes. It actually calls it DB-37 in the manual.

Anywhos, now it would be interesting to get hold of a CDP hard drive controller. I think that is a big board that sits on top of the floppy drives and power supply. I've seen that in some pictures but I'm not sure one has ever been on eBay. Of course normal PC/XT ISA controllers work fine in a CDP 1600.