PDA

View Full Version : Single Board Computers



Old Computers
September 23rd, 2012, 03:29 PM
I just acquired two SBC's. One of them is called The Bigger Board SBC-2001 Z80A manufactured by Business Information Systems. I have the manual with it, and it implies that a CP/M customized for it came with it. Can I modify CP/M myself?

I also have another SBC that does not have a name from what I can tell, but it is manufactured by Molecular Computers, and the numbers on it are 1000082 REV A FPX 80. 90272002 REV HI. Doing a search for it turns up nothing relevant. Do any of you know anything about this computer?

Thanks,
Old Computers

Chuck(G)
September 23rd, 2012, 04:10 PM
1) Definitely--DRI even published a "System Alteration Guide" for folks wanting to write their own BIOS. Check Gaby's CP/M site for copies of it.

2) If it's the same Molecular, that I'm thinking of, it might not be quite stand-alone. Molecular specialized in multi-user systems, so an SBC would seem not to belong to their portfolio. But there was more than one Molecular, so it's hard to say without seeing the board.

Old Computers
September 23rd, 2012, 04:23 PM
Thank you very much Chuck(G).

I can post some pictures of the second computer if you would like to see what it is. I did inspect the board, and it appears to be stand-alone. It has two rows of pins that appear to be serial ports, two PSU inputs, and two disk outputs. There is a Z80A on it, and a Z80ADART, Z80CTC. It also has a huge bank of what I am guessing are RAM chips, and an EEPROM with the sticker reading 50250130 G.

Chuck(G)
September 23rd, 2012, 04:46 PM
Well, @OC, here's the logo of the company that I recall:

http://logosdatabase.com/logoimages/73379822.jpg

They had the approach of using one Z80 as a supervisor with each user having his own Z80. Circa 1979-82 or so. You could have, IIRC, up to 16 Z80 user cards in a box.

Old Computers
September 23rd, 2012, 05:28 PM
I could not find the logo on the board anywhere. If it might help with identification, the board appears to be factory assembled. I'll go ahed and post a couple of photos tomorrow.

So if it turns out to be the type of system you described is it pretty much useless unless I find the other componets? That sort of system sounds very interesting.

patscc
September 23rd, 2012, 08:31 PM
The logo looks familiar, does anyone have any back issues of BYTE ? I wanna say there was a lengthy review in one, but mine are boxed up at the moment.
Old Computers, is there a FCC number on the board, by any chance ?
patscc

Old Computers
September 24th, 2012, 06:02 AM
I did not notice one, but I might have overlooked it.

Old Computers
September 25th, 2012, 07:40 PM
I checked the board, and there does not appear to be any FCC ID. I did look at the numbers on the board a bit closer and I think that it is the Molecular Computers you guys are talking about. The designation FPX BD makes me think that it is some kind of sub board to a system. What do you think?

Chuck(G)
September 25th, 2012, 08:37 PM
The name of the company that I'm thinking of is "Molecular Computer, Inc" (note singular "Computer"). If your board is labeled with "Computers" (plural) instead, I have no idea of what you've got.

Old Computers
September 26th, 2012, 06:24 AM
I guess that that board will remain a mystery for now. I searched for it, but all the results are things about DNA and computers. (Pictures will come soon. I have just been very busy lately)

I was able to test out the SBC-2001, and it appears to work perfectly. I ran its memory test for about 5-10 minutes and it did not return any errors. I also had it display its rom contents and it returned the bytes. One thing I noticed about it though is that the manual said that the ROM address was from C000 to CFFF. Around CE00 It started to return 00 for the bytes. Is that just because the monitor does not fill the EPROM?

The bad news is that its CP/M BIOS is not built in to the ROM. The manual mentions a BIOS disk and a distribution disk. I am guessing that CP/M will be easy to modify, but I am unsure about the BIOS. I have never written anything in assembly or machine code before. I am guessing that sample BIOS's that are in my CP/M programmer's book will not work since BIOS is system specific.

patscc
September 26th, 2012, 06:39 AM
That leaves a little over 500 bytes unused, which sounds reasonable.
Have you tried digging around on the web for any disk images for it ?
patscc

Old Computers
September 26th, 2012, 06:47 AM
I have done a basic search, nothing too major. I did look on Gaby's CP/M site and Dave's Old Computers website, but I did not find my system listed. I could probably search some more in those sites though. I tried to see if Business Information Systems still exists, but searching did not turn up anything. It was located in Johnson City, Tennessee.

patscc
September 26th, 2012, 07:07 AM
There's still a Business Information Systems listed in Tennessee. If you haven't, you might want to try giving them a call on the off-chance it's the same outfit.
patscc
p.s. if they are, and they offer you a bunch of old boards and docs they have no need for, I'd like some. :)

Chuck(G)
September 26th, 2012, 08:26 AM
While not unknown, CP/M BIOS in ROM was not common. Most systems coded it as part of the boot record, just as you see it in the "System Alteration Guide" . If you can puzzle out what's what port-wise on your SBC, you can write your own very easily.

There was a lot of copying from reference designs back then, so you may also want to see if there's an Ampro or Yasbec design that matches.

Old Computers
September 26th, 2012, 10:09 AM
@patscc: I looked them up, and the city they are located in is very close to the Johnson City TN, so it probably is the same company. I will look into contacting them.

@Chuck(G): I will look into those ideas.

Do you recommend any Z80 assembly/machine code tutorials?

Chuck(G)
September 26th, 2012, 10:18 AM
Do you recommend any Z80 assembly/machine code tutorials?

As strange as it sounds, I'd recommend learning 8080/8085 code first. The big reason is that the DRI coding guides all use 8080 examples, not Z80. The stock CP/M assemblers (ASM and MAC) are 8080.

As far as tutorials, no recommendations--I learned from the datasheets when the chips came out.

Old Computers
September 26th, 2012, 10:33 AM
Will a datasheet be easily found from an internet search?

As I start this project I will probably ask for some more help.

patscc
September 26th, 2012, 11:27 AM
You might want to try this:
http://www.textfiles.com/bitsavers/pdf/intel/ISIS_II/9800301-04_8080_8085_Assembly_Language_Programming_Manual_ May81.pdf
patscc

Old Computers
September 26th, 2012, 01:06 PM
That will be a great help. Thanks!

Old Computers
September 27th, 2012, 01:31 PM
Here are some pictures of the Molecular Computers board.
1040010401

Some more news about the SBC-2001: I have been running the test for the D command which outputs the contents of the memory to the console. It seems that if I enter parameters greater than 0 the computer freezes up. Could the EPROM have lost some of its data? I do have the monitor listing if I need to re-assemble it.

Chuck(G)
September 27th, 2012, 01:54 PM
The board photos show that the firm's name was "Molecular Computers", which is a different firm than the one I'm thinking of. Besides, I think Molecular bit the dust around 1983-84. Still, you mgiht dump the EPROM to see what this thing really is.

As far as the other board's monitor goes, you might see what the program listing says. It could be that you have the syntax of the command wrong (I've banged my head a few times on that one).

Old Computers
September 27th, 2012, 02:20 PM
Okay. To dump the EPROM I am assuming that I need an EPROM burner. Do I need a certain type of burner, because I have no clue what type of chip it is?

Here is an excerpt from the manual:

D (Start adr),(End adr (optional)) (cr)

Displays memory in hex from starting address to optional ending address. If no ending address is selected, 256 byte block will be displayed. A CTL-S freezes the output and a second CTL-S causes it to resume. Any other key depressed terminates the displayed output.

I have entered D0000 into the computer and it displays 2 lines of the memory contents. If I type something in with a larger argument it freezes (I first noticed this when I typed in D0000,FFFF, but it does it with much smaller ranges too and with only a start address specified). I have tried CTL-S and the no-scroll button on the terminal but that does not help. I also have tried pressing other keys and it does not return to the prompt. The reason why I think that the cause could be that the EPROM lost some data is because when I use the S command (it dumps one memory location to the screen and allows you to change it) without changing the value, I find some larger blocks of 00 in the returned data. I also double checked and the returned contents from the EPROM end around CA00, not CE00.

I have successfully executed other memory operations such as fill the range of memory with the specified byte. I have not tried the move contents command yet. I did try the run user program command, but it froze. I believe that the reason for that was because I did not have any program at the address I specified.

I have some more news. I was experimenting with the monitor program a bit more, and I found out that if I press the no-scroll button a couple of times, the monitor continues to list the memory contents. Some of the lines though are not completely dumped to the screen and are ended with the gibberish character. Sometimes though, the readout will continue and the console will display some numbers mixed in with the gibberish character. I am not sure if this is a terminal problem or a problem with the monitor program. After that the monitor may act a bit strange, but I believe that it has something to do with pressing the no scroll button. After a while the behavior returns to normal. I do know that it when it does display the memory contents the displyed bytes are correct. I also checked the PSU's output voltage, and it reads 15v. The manual says that it should be 12. That might be a cause for the improper output.

Chuck(G)
September 27th, 2012, 04:50 PM
Well, it could be that there's a bug. Another possibility is that your memory is parity-checked and you're running into a parity error trap. I assume that you've tried a space between the D and the address.

Most likely, the EPROMs on both these things are 2764, 27128 or 27256 types. That would be approximately correct for a 28-pin EPROM for the time.

Old Computers
September 27th, 2012, 05:03 PM
I just reviewed the ebay listing of that computer, and I found that it sounds like the seller might have had the same problem:

Recently tested, signed on with required baud rate. BIOS testing for several functions returned lines of code on each function 4,6 but no ASCII characters. As if baud rate was wrong on return but maybe picked up end of line character for displayed information. Numbers of lines of code returned depended on which BIOS functions such as memory dump, memory test, etc. No other testing was done.

Having the bug is not too big of a deal to me, I am more interested in preserving the contents of the EPROM so it is not lost. The monitor will not allow me to type a space after the command. It displays a new line and asterisk then the prompt on the new line.

What exactly is a parity error trap. I understand what parity is in relation to serial communications, but I never knew it existed in memory.

patscc
September 27th, 2012, 05:33 PM
Parity uses a bit(s) stuck on the end of a word to indicate the parity of the data byte(s). So, for an 8-bit byte that was even, you recorded a 1 in the 9th bit for parity. The IBM PC used it, but just halted if a parity error got trapped. More useful variants of this use more bits per word, allowing for automatic correction of 1 or even multi-bit errors. Real popular with the server crowd.
Most modern readers should be able to read them, but maybe not burn that type of EPROM, but that shouldn't really matter. If you don't have one, you can usually pick up basic ones pretty cheap on eBay. Or I'm sure someone here would be more than happy to archive them. I found a flyer on the web for the bigger board. Tomorrow I'll see if I can decipher the part no. on the EPROMS.
patscc

Old Computers
September 27th, 2012, 05:59 PM
Thanks. I have the schematics for the board too if you are interested in seeing them. I could scan them. I also found a contact info page for Business Information Systems. I think that I will try to contact them sometime this week or next.

I tried turning off the Auto XON/XOFF feature of the terminal, and the memory dumped continuously, but it still displayed the gibberish character. I think that it has to be a bug. I just changed the jumper settings for the baud rate, and the system is running perfectly when I do the D command. Although I have it set to 9600 baud, the system is using 1200 baud. I might have made a wrong connection.

Old Computers
September 29th, 2012, 05:07 PM
Another update: I did have the jumper settings set incorrectly when I ran the above test. I did try to use the monitor's auto baud rate detection, but that did not work too well (it still was at 1200 baud and the terminal at 9600 baud). I may have to check other jumpers to see if there is any conflict. I reset the jumpers back to 9600 baud and the monitor would not dump the RAM contents properly. I do believe that there is some sort of bug in the system.

patscc
September 29th, 2012, 05:18 PM
Yeah, it would be great if you could scan the schematic and post it. Feel free to PM me you want my email address to email the scan to.
patscc

Old Computers
September 29th, 2012, 05:35 PM
Okay. When I get it scanned I will let you know.

Old Computers
October 5th, 2012, 07:45 AM
I am still working on scanning the schematics, but I would like your guys' opinion on selecting an eprom programmer. I did an ebay search and tons of different types appear. I am aiming for a higher quality one because I do not want to trash the eproms and their contents. Here are the two that seem to come up the most:
http://www.ebay.com/itm/NEWEST-Willem-EPROM-programmer-BIOS-PIC-Designed-in-the-USA-ShipfromUSA-/290783060575?pt=LH_DefaultDomain_0&hash=item43b405625f
http://www.ebay.com/itm/TOP853-USB-universal-programmer-EPROM-MCU-GAL-PIC-NEW-/271059058202?pt=LH_DefaultDomain_0&hash=item3f1c60fe1a

What do you think of those, or would it be best to get a vintage burner? I am just really afraid of loosing the contents of the chips if I get a bad burner. Also just a quick question about eprom burners. Is there any way to write protect them so I do not accidently press the wrong button when I am reading the chip?
Thanks.

patscc
October 5th, 2012, 08:06 AM
I have a Willem, and it works pretty good. It's hard to do anything like accidentally burn a EPROM in it, so you should be good to go.
I don't think you can overwrite a EPROM, once it's burned, you can't access the gate electrode in the cell, which is why you need to erase them with UV first.
It's got so much more flexibility than older burners since it does FLASH and other types.
patscc

Chuck(G)
October 5th, 2012, 08:22 AM
I've had the best luck with used professional-quality programmers. A lot of the old gear is being ditched for good prices because of several factors:

(1) Nobody uses UV EPROMs any more, except to mainain old hardware--and that's vanishing.
(2) Modern PCs lack real parallel or serial ports used to interface to the old programmers.
(3) The old gear doesn't support more recent devices.

So, for instance, a Xeltek Superpro programmer might be had for $50, which makes it a better deal (if you have an older system to drive it with) than any new kit, with the benefit that the programming software (1) runs on DOS or at worst, Windows 3.1 and (2) has been thoroughly field-tested.

Much easier than trying to work your way through someone's Chinese. On the other hand, if you're looking for a decent modern unit and don't have unlimited funds, I've recommended the Wellon programmers as being a slight notch above the run-of-the-mill Far East stuff, followed by Genius, followed by the TOPxxx stuff. A glance at eBay shows lots of choices.

MikeS
October 5th, 2012, 09:20 AM
Okay. To dump the EPROM I am assuming that I need an EPROM burner. Not necessarily; do you have another computer with a compatible spare socket?

Old Computers
October 5th, 2012, 10:40 AM
Thanks for the suggestions guys. I think the Willem one is at the top because there is one for pretty cheap, and because it supports many different chips, but I will definitely look into some of the older professional ones too. I have several computers that could support one. An XT, and XT clone, an AT, and a few mid to late 90's towers that have parallel and serial ports on them.

MikeS, I do not know if the sockets in my computers would support it.

MikeS
October 5th, 2012, 11:40 AM
MikeS, I do not know if the sockets in my computers would support it.The XT and clone almost certainly have spare sockets and the AT probably does as well AFAIK. Did you ever establish what size/type the EPROMs actually are?

Old Computers
October 5th, 2012, 02:40 PM
For the SBC-2001 the EPROM is a 2532, according to the parts list (the label covers up the chip number in addition to the window so I have not read the number directly from the chip). There is also another chip on that board that is a 74S288 memory mapper ROM.

For the other board, the label completely covers the chip, and I do not wish to damage the label because it has some sort of number on it that might be helpful in the future. The printing on the board says 2732 next to an outline of the chip. There is a larger outline that surrounds the smaller outline that says 2?64. I think the ? is a 7, but there is a via right there, and the solder covers the number almost completely. There are also several smaller chips on the board with labels on them. I do not know if those would be ROMs or not.

patscc
October 5th, 2012, 02:59 PM
The smaller chips with labels might be PAL's. 2532's (should be a TI part) and 2732's are both EPROM's, though not interchangeable. The larger outline is probably for a 2764. The labels may, if the case is not plastic, cover the erase window, so you want to keep them on.
The Willems should read the 2532 and 2732, but check the manual for exactly what families are supported on their website.
The 74S288 is a TTL fusible link PROM. I don't think the Willem will read this, but I could be wrong.
patscc

Old Computers
October 5th, 2012, 06:32 PM
Thanks for the advice. I will be sure to check their website out.

I am not too worried about the PROM since those don't lose their memory over time like an EPROM can (unless I am mistaken). The manual has a listing for that PROM but it looks more like a table than a listing of code.

patscc
October 5th, 2012, 07:53 PM
They don't loose their memory, but they do go bad, just like any other chip.
patscc

Old Computers
October 6th, 2012, 12:18 PM
http://www.ebay.com/itm/Parhelion-8255A-EPROM-PROGRAMMER-TEXTOOL-SOCKET-PERSONALITY-MODULE-MINT-/160895436325?pt=LH_DefaultDomain_0&hash=item25761d8625

I found this one. What do you guys think?

patscc
October 6th, 2012, 01:18 PM
I don't see any software included.
patscc

Old Computers
October 6th, 2012, 01:30 PM
I'll let that one go then. Would something like that one be good for me?

I looked into the specs of the Willem ones on Ebay, but they did not support the chip numbers I need.

patscc
October 6th, 2012, 02:01 PM
If the specs support your chips. The big thing is available software. What chip(s) are you wanting to read ?
patscc

Old Computers
October 6th, 2012, 02:47 PM
One is a 2532 and I believe the other one to be a 2732.

When I read the specs for the Willem I found that it did support a 27(C)32 (is that the same or equivalent?). I might resort to using one of my PCs to read the ROM if those chips are supported.

Here are the supported chips of the Willem burner from Ebay:
EPROMS:
27C64, 27C128, 27C256, 27C512, 27C010, 27C020, 27C040, 27C1001
M27C1001, M27C2001, M27C4001
27C080 (A19) , M27C801, M87C257
27(C)16, 27(C)32
2764, 27128, 27256, 27512, 27010

This one just popped up on Ebay, and the software is on the net (I downloaded it): http://www.ebay.com/itm/RARE-Needhams-Electronics-PB-10-ISA-PC-Card-EPROM-Programmer-w-Opt-Prog-Bd-/110960920907?pt=LH_DefaultDomain_0&hash=item19d5c9494b

patscc
October 6th, 2012, 03:22 PM
The 5.0 reads the 2732's haven't tried it with the 2532's.
Some other Willem ones do, like:
http://www.mcumall.com/comersus/store/comersus_viewitem.asp?idproduct=4282
or (cheaper)
http://www.mcumall.com/comersus/store/comersus_viewItem.asp?idProduct=4225
For a supported device and adapter list,
http://www.mcumall.com/comersus/store/mcumall_TrueUSBWillemsupportICs.asp
These guys actually make them, so if you want any support, I'd buy from them.
I don't know how much you want to spend...
patscc

Old Computers
October 6th, 2012, 04:14 PM
The mcumall ones are very nice. I am considering those, but they are a bit beyond what I feel like paying for one since I probably won't need to use it too often. I might decide to buy one in the future because they support a whole lot of chips including the ones I need.

My price range is $50-60 and lower.

patscc
October 6th, 2012, 07:40 PM
You might be able to pick up a used one, but you might have a bit of a wait.
patscc

Old Computers
October 7th, 2012, 03:25 PM
Okay. I'll keep my eyes out for one.

I try to get the schematics and or manual scanned soon.

Old Computers
October 14th, 2012, 08:29 AM
I thought I would give a quick update for the SBC-2001. I changed some of the jumper settings to the default the assembly instructions mentioned. The auto baud rate detection now functions properly. I still have the problem with the D command at 9600 baud, but I have a theory that 9600 baud might be faster than the program runs. I also noticed that the board is missing a couple of ICs. I checked the manual, and most of them are spares, but one or two are not. I will try to source those and place them in the empty sockets.