By mid-1984 the stockpiled B128 inventory had to be liquidated, and Commodore began searching for a company that could do the job. As we now know, that opportunity went to Protecto Enterprises in the U.S. By this time, about an estimated 3,000 North American customers had already purchased a B-Series system from the dealer, and paid top-dollar and more for it and necessary accessories like disk drives, a printer, and software titles. It wasn't exactly customer satisfaction to have bought around $4,000.00 into a company’s product line, and then get the same treatment as if you bought the $199.99 VIC-20 from a retailer. And it would end up making a unique difference - something never seen before or since in the relationship between a computer manufacturer and their customers.
At around the time that Protecto was running their first ads, a group of B-Series owners in the Chicago area were being brought together by a motivated owner going by the name Norman Deletzke (pen name was Delezke). These early customers of the B-Series saw the potential in the machines' design, and wanted to pool their resources together to get effective return on their computer investment. In May of 1985, Norman formed the Chicago B128 Users Group (CBUG). CBUG put out it’s first newsletter, the CBUG Escape, in October of that same year. It was to become an international organization.
If you have the back issues of the CBUG Escape, volume eleven (1988) or volume thirteen and higher (1989-1990) I would purchase them to complete the library I have!
A lot of computer models get enough popularity to attract customer interest in forming and joining fellow users of the product in a user group setting. But the CBUG was unique from the beginning. Very persistent and leveraging the grass-roots approach long before the internet, they scored their first success in getting Protecto Enterprises to release the mailing list of the 11,000+ buyers of the B-Series equipment they had liquidated. An aggressive mailing campaign resulted in even more membership with a peak of 3,000+ members worldwide. Together, this very vocal group of consumers pushed Commodore into doing what no computer company had ever done, or has done since - turn over their engineering assets for a discontinued computer model! And here is a copy of the letter the CBUG Escape of Spring, 1986 showed to prove it! I have in my possession today the majority of those assets due to an arrangement with the former officers of the CBUG to preserve them.
Now all the true potential was coming to be. With the direct aid of CBUG to show developers of hardware and software that there was a market for the B-Series, the products began to flow. By the end of 1986, there was not one but two vendors marketing a 1MB add-on board for the B-Series, several commercial software titles, several repair and refurbished equipment vendors, and dozens of shareware and non-commercial software titles and ports from other Commodore platforms. Even Protecto got into the act, pulling together the programmers' reference that Commodore never managed to publish. It was becoming exactly what Norman had wanted to accomplish - a thriving computer-user community around the B-Series.
CBUG even managed to secure from Commodore US and UK most of the prototype Intel 8088 co-processor boards that had never escaped from R&D. One member even designed a variant that used the NEC V20 microprocessor instead so that it met power requirements for the lo-profile model B128-80. They received a master copy of the never released MS-DOS® 1.25 for the B-Series and for CP/M-86 1.1 for the B-Series. And they addressed one of the obvious hardware issues that Commodore never got around to doing - how to read a non-Commodore GCR encoded diskette format by publishing a way to connect a Commodore 1571 diskette drive to the DatassetteTM port so MFM formatted diskettes could be read, and the public domain software library of CP/M-86 tapped into. This was needed to read the MFM-encoded diskettes of the CP/M-86 and MS-DOS® world, as both CP/M-86 and MS-DOS® for the B-Series used Commodore proprietary GCR encoding natively so that standard CBMTM IEEE-488 peripherals could be used. Imagine working within CP/M-86 on your CBMTM 8050 dual diskette drive unit!
If you take a look at the Commodore Annual Report of 1982, you'll find that there was indeed a line of IEEE-488 diskette drives planned that could read MFM style MS-DOS® and CP/M-86 diskette formats - but these never even made it to the prototype stages AFAIK. But even this omission from the list of products which escaped Commodore did not deter CBUG members. They did complete the system Commodore originally promised with the B-Series line - but it took until early 1987 - a working Commodore co-processor daughter board with the Intel 8088 co-processor, CP/M-86 and a whole new set of software that ran on it. The MS-DOS® was never completed or released by Commodore either. But the internal development version CBUG came into possession of from Commodore U.K. was made available to a limited number of technical members to try and further the development. I believe I have the only known working copies still to exist of both the MS-DOS® and CP/M-86 for the B line.
CBUG ran from mid-1985 until late 1989 - shortly afterwards, in October, 1990, Norman passed away. The direct efforts of his user group is what has kept many a B-Series machine from the trash heap and made them more collectible today. It's interesting to note that while as early as 1986, CBUG had in its possession, and offered for sale to its members, some 40 Commodore-manufactured Intel 8088 co-processor boards (most were installed in U.S.-made CBM 128-80 or CBM 256-80 systems). Today, only seven 8088 co-processor boards are known to exist (and I own two of them!). If the internet had existed publicly only six or so years earlier, there’s no telling what the history of this classic computer system could have been with people like Norman out there.
After CBUG ended one of the most active vendors in B-Series hardware and software, including the development of some unique commercial software titles for the B-Series, they attempted to continue the tradition. Northwest Music & Computer published four issues of the Northwest Computer News, which focused on B-Series information for a short while longer, from 1990 until mid-1993. The final issue was never published. As this ended, too, the B-Series was left totally to the realm of the computer collectors, historians, and enthusiasts such as myself.
There are only a few resources for the B-Series that exist today (you're on the largest and oldest one of them now), but the amazing reality is that due to the internet, those few resources provide a wealth of information and support contacts. The best place to get started collecting is the auction site Ebay. I've listed here other websites I am aware of that have B-Series-specific content, as well as newsgroups where B-Series owners are likely to be watching. I'll leave you with a bit of irony. Read this letter from CBUG to Commodore from 1987, and then go to the VICE Emulator page. ;-)
View the hardware section of this
site for information on part numbers, and ROM levels and PLA levels that can be
helpful in repairing or restoring a B-Series computer back to factory original
condition. I've included some common failures during my restoring some dozen
odd B-Series systems and the probable causes below.
On Power Up System boots to monitor prompt instead of BASIC
The problem is the BASIC ROM(s). Either not seated properly or defective. Try re-seating first. Then replace with an EPROM using same revision images in the hardware section of this site as original for your model. If you have the production ROMs your best choice is a Motorola MC68764C or MC68766C EPROM. If you have the EPROM adapters as original then it is very easy, just use a common Intel 2764 EPROM. Nothing fancy, but make sure the part you use either matches or exceeds the original’s access speed. A slower part could cause intermittent timing problems.
BASIC programs known to be good give strange errors (e.g. next without for) or System boots normally but any BASIC command results in ?Syntax Error
The problem is most likely a bad DRAM chip if it can be reproduced right at power-on before the system warms up. Replace the Mostek MK4564N type with either 4164, 4864, or 2600P DRAMs. Debug the failed chip by removing the BASIC ROMs and booting into the monitor (TIM). Use the fill pattern 0xaa, 0x55, 0x33, 0xff on several locations in each bank to determine the bad bit(s). Use a logic probe to identify the matching chip (checking CAS in the appropriate bank). Each DRAM was responsible for one bit of the eight involved in each byte. Wherever you find a stuck or dead bit, the chip for that bit is the one to replace. You can also try the piggyback technique where you place a known working DRAM chip on top of each one (be careful you don't cross short a pin) and power on/off repeating for each chip until you notice "something different" in behavior. Most of the time that chip is the bad one (or one of the bad ones) and you can replace it and repeat. You will need the schematics from the service manual for this debug.
I have problems with running programs or characters changing after the system gets warm.
This could be a bad DRAM or even part of the refresh logic. But most likely it is a failing power supply. The electrolytic capacitors in these units are the first part to fail. You can check +5V at the power on LED connector for lo profile units. If it’s below 4.9 volts or more than 5.1 volts, you need to service the power supply. But first, try a simple adjustment; there is a small screw on one side of the power supply. This is R27, on the lo-profile power supply, and is used to trim the +5V feed. With a digital multimeter (DMM) try SLOWLY adjusting this with a non-conductive tool while checking Vcc and ground on one of the EPROM chips. Adjust to +5.05V as indicated in the service literature of the time. If that does not work, you'll need to remove the supply for service. Note: these are switching power supplies; do not test them disconnected from the motherboard without a load or you can permanently damage them. Also, these supplies work with HIGH VOLTAGE. DO NOT WORK ON THESE UNLESS YOU KNOW WHAT YOU ARE DOING - YOU CAN GET KILLED OR SERIOUSLY INJURED! In my experience replacing ALL of the original electrolytic capacitors followed by testing/replacing switching transistors makes for a lasting repair. The Motorola MJE 8501 switching transistor is the most likely to have failed if it is a lo-profile power supply. But it is ULTRA rare, even in the late 80’s. Go looking for the MJE 8503, STI8501 or BU508A instead. It will service fine as long as you get it in the same TO-220AB packaging. If you have a high-profile machine, then the part is a Motorola BUV48 transistor in TO218 packaging and a replacement can be the NTE2311.
On Power Up the system prints the power-on banner but is hung and a whining noise is heard
Only on hi-profile models, the power supply case is touching the voltage regulator by the cassette port. Bend it down more. This can also be caused by an 8088 board touching the power supply of early model power supply units which were bare steel (later models were insulated with paint to prevent this). Use clear packaging tape to form an insulation at the point of contact on the power supply case. Also watch out for early design power supply units in hi-profile models. These are sourced from CEAG as part 40071065905 and slightly too large and will make contact with the 8088 board. The problem ones are only painted brown on the back whereas the later ones are painted all around and are slightly smaller. These later ones may have been sourced from General Instrument like the lo-profile models. If so the model number is likely CE-75-502, whereas the B128 used GI power supply model CE-35-502.
System runs fine, but a whining noise is heard.
Only on hi-profile models, the cooling fan mounted on top of the power supply needs cleaning/replacement. Can also be a filter capacitor dying in the power supply unit. I highly recommend replacing all the electrolytic caps in the power supply. I offer a recap kit for this purpose on the forsale page.
8088 Board is installed but when booting CP/M-86 or MS-DOS the system hangs
Assuming the coprocessor board is okay, this is most likely either an incorrect PLA (only 906114-05 should be used in a hi-profile model. The -04 is only meant for the lo-profile motherboards) or a weak/dying power supply. If you test the power supply, do not forget it is a switcher so do not test without a +5V load! You can use a 1ohm 50W wire wrap resistor across the +5V and ground to provide this. I highly recommend replacing all the electrolytic caps in the power supply. I offer a recap kit for this purpose on the forsale page. Another possibility is that there is a known design flaw in the lo-profile models that makes them incompatible with the 8088 board, this is due to an unterminated DRAM data bus. The ECO for this is to solder a 10K ohm resistor onto any one bit of the bank of installed DRAM and +5V (this will have to be done for each populated bank of DRAM). Lastly, this can also can be caused by installing an 8087 math coprocessor into the coprocessor board socket for same. This was originally designed to be supported, but there is a defect in the PCB board that prevents this from working. If you want to have the 8087 work you can manually implement this ECO. Look in the valley between the 8088 and 8087. Find the silkscreen label "JP1". Near it will be two solder pads. One is connected to the 8087 pin 31 and then is linked to a trace to its left which connects to the 8087 pin 33. This is the error. This "link" should have been part of a two pin jumper on the board that was never implemented but called for in the original schematic. Cut this trace so 8087 pins 31 and 33 are not connected and it will work fine. The issue is that the 6509 needs to request the bus from the 8088 on its RQ/GT0 line. But when the 8087 is installed, the 8087 needs to arbitrate on this line (in the CBM design) and the 6509 instead needs to be connected to the 8087 passthru RQ/GT1 (pin 33). The jumper was originally intended to tie the 8087 pins 31 and 33 together when no math coprocessor was installed to complete this circuit and to disconnect them when installed. Hence the JP1 design intention.
The 8088 co-processor board doesn't fit my B128.
Working as designed. The board only reached the prototype stage, and would only fit the hi-profile systems with the General Instrument sourced power supply at the time the project ended. The lo-profile -04 PLA lacks a reliable implementation for the coprocessor board but it would function. But, the lo-profile power supply is not capable of supplying the additional power draw requirements for the co-processor board for any significant period of time.
My high profile system’s monitor has quite a bit of video noise.
If your unit is a very early B700 design, it could be due to a design defect that was corrected thru an engineering change order (ECO). The fix was to bend up pin 13 on U54.
My system is missing screws for the case.
These are metric machine screws and most Ace Hardware stores will have a replacement that not only fits, but has the exact same pan head shape and design. Make sure you get them zinc plated as original to prevent corrosion.
My P500 will not access a diskette drive.
If you type the BASIC 4 directory command and the system hangs then the problem is the 6526 chip which is socketed and is a fairly easy fix. You can confirm by peeking the first address location register and with no joystick or paddle installed on game port 1 you should read 255. If you read 0 it's a confirmed dead chip. But if the directory command instead returns immediately to the ready prompt without any disk drive activity then it's actually a bus grounding problem. I have come across two USA made P500s where the IEEE-488 cables braided grounding leads had to be connected on both the P500 and the diskette drive connector ends or no communication could be established. No doubt another factor of the very raw state of the prototype design.
My system is missing the rubber feet on the bottom.
These are VERY hard to find exact replacements for on the production models. The B500 and P500 used a dome type which is much easier to locate. The production models used a pancake type that is harder to find. The 3M Company makes a model with the right shape and approximate size to the original part but nobody carries it anymore - you can try to request "samples" from 3M though. I was able to get some this way. The part is 3M Bumpon SJ-5744. I also offer these for sale.
I cannot get a Commodore 1530 DatasetteTM to load/save to tape.
Unless you're trying this with a B500 or P500 with first (-01) revision ROMs, it is working as designed. The tape code was pulled from production ROMs after April, 1983 to make room for a machine language monitor. If you are trying this with a machine with the original ROM code, then run a tape head cleaner on the 1530 first, then a head de-magnetizer and try it again with a high-quality but non-metal tape.
I cannot get a commercial B-Series software title to load in my 8250/8250LP or SFD-1001 diskette drive.
Nearly all B-Series titles were released on 8050 diskette format only, most with copy protection as well. The 8250/8250LP and SFD-1001 drives will not function properly unless you program them as an 8050. Here's a short CBM BASIC program to run prior to running the application that will do this for a drive set as device #8:
10 open 15,8,15
40 print#15 ,"u9"
50 close 15
I cannot get a commercial B Series software title to run or it hangs.
Many B-Series models can be characterized as prototypes, even those that finally got into production. First off, if you have a P500 model then you are out of luck; the architecture is too different to run any of the commercial titles that got released. If not, you need to check if maybe you had a pre-production series ROM set. Type in the BASIC commands BANK 15:PEEK(65529) and press the return key. This address in bank 15 holds the ROM code version number. It should have a value of 01 at $aa. Anything else is too old and burning EPROMs with the latest (revision 04) firmware may be all you need. The images and technical info are on this site in the hardware section.
When I press a key on the keyboard, I have to press it really hard or more than once.
The problem is that the contacts are dirty, oxidized or damaged. Take apart the lo-profile case or the hi-profile keyboard enclosure and remove the keyboard assembly. Remove all the small Phillips screws attaching the base of the keyboard. Cut the two nylon ties holding the hi-profile keyboard cable (RadioShack has replacements). Unsolder the two connections for the shift lock key, and remove the bottom cover. Clean all of the contacts on the bottom cover with RadioShack contact cleaner, and blow out the whole assembly with canned compressed air. If you notice any wearing of the gold plate on the contacts, repair it using conductive paint. Take this opportunity to clean dust/debris from under the keycaps themselves too. This should fix the problem.
How can I convert my P500 from PAL to NTSC or back?
This is a relatively simple change. You will need to swap the MOS 6567 for a 6569 at U23 to go from NTSC to PAL. You need to open the link at LK5 and close the link at LK4. Finally, swap the crystal at Y1 for the 17MHz one. To go from PAL to NTSC you need to swap the MOS 6569 for a 6567, close LK5, open LK4, and swap the crystal at Y1 for the 14.31818MHz one. Lastly, make the color adjustment with the small trimmer until you get good color. Make the adjustments VERY SLOWLY with a non-metallic tool.
My lo-profile system comes up with no color on a color monitor.
This is normal for the B500 and B128, as they are 80-column monochrome business machines. But for the P500 this is not, and the most likely cause is the crystal at Y1. Either it is bad or wrong for your country (e.g., NTSC or PAL). Check by using both 5pin and 8pin video cables to eliminate the cables and connector first though. One last check for the P500 is the color trim adjustment R60, remove the metal lid/heat sink for the 6567 and adjust this screw EVER SO SLOWLY and see if that helps. If the problem only happens after the system warms up, suspect either C90, the 2114 color RAM U24, the 6567/6569 at U23, the MC4044 at U21, or 4066 IC at U26.
The power up banner displays and I have a blinking cursor but typing any commands causes a hang.
Sadly this is due to a failed 6509/6509A CPU. That or the 6525/6525A. And these being the two chips that are just about impossible to locate replacements for. The only source for a 6509 is another B-series computer. You may encounter either the 6509AR4 or 6509AR7. Rev7 was the last one. On the MOS 6525 or MOS 6525A there are several, though still rare, sources: another B series computer, the C64 magic voice cartridge, the C64 IEEE-488 cartridge from Commodore, the 1551 disk drive, the B series 8088 coprocessor prototype board, the Amiga A570 CD-ROM drive, the Amiga A690 CD-ROM drive, and the Commodore CDTV. Commodore also released service literature to their dealers in 1984 approving the substitution of the MOS 6523 / MOS 6523A (the 40pin chip NOT the 28pin 6523T) for the MOS 6525 / MOS 6525A in repairs, while an even rarer chip I include the 6523/6523A here for completeness. Note that the "A" suffix indicates 2Mhz operation and no suffix indicates only 1Mhz operation. So you can use an A suffix chip in a P500 for example, but you cannot use a non-A suffix part in another B series as it will be too slow. The "A" suffix is the more common chip to locate on average.
My 8050 frequently gets an error 75, Format speed error.
If you have already tried to readjust the RPM speed and it will not stay set or varies with temperature, then this could be due to a known design problem in the 8050 circuit in that the speed determination logic diode does not meet specification. It was fairly common in the mid-80s, but after 20 years these parts are no doubt much worse about leaking. There was a CBUG member fix published to field correct the problem. It involves switching CR25 and CR21 from a 1N5235 6.8v zener diode to an ECG 5014A (1N5235B) one. Follow service manual instructions for re-adjusting RPM speed.
I personally bought one of the last 1985-circa B-Series deals from Protecto Enterprises - I paid $495.00 for it. That's how it started for me. I got a very late model production B128-80 and a Matrix Printer Parallel (MPP)-1361 printer which was the best dot matrix printer I've ever encountered. I also got an SFD-1001 Single Floppy Double-sided (hence the "SFD") drive, as Protecto had ran out of the CBMTM 8050 dual diskette drives several months earlier. And I did not get a monitor either - I used my Commodore 1702 instead (though it had trouble with 80-columns). I still considered it a very good value, and used the unit quite a bit as I was just starting my undergraduate Computer Science degree at the time. I joined CBUG right from issue one and stayed on until sometime in early 1989 when, as I was starting my graduate degree program, I joined the IBM Corporation where I still am today (just have a thing for business computers I guess). ;-)
To this day, I still have most of the Commodore and CBUG materials as well
as numerous magazine clippings and other reference literature and material that
enabled me to develop this website - not to mention repair and maintain my B
computer collection and spare parts. Many years ago, I was most fortunate to
track down the last remaining assets of CBUG and given custody of them -
rescuing these assets before they were lost forever. As time permits, and interest
is shown, I will make this information available at this site and funet subject to various copyright requirements of the
I expanded my collection during the 1990’s with the addition of several additional B-Series systems. Today, I personally own the entire North American production family collection! Including an early production B128-80, one of the last B128-80s ever made, a B500, a P500, a CBM 128-80, a CBM 256-80 both with the 8088 board; peripherals of the correct vintage like the 2031LP, SFD 1001 and 8250LP diskette drives; the MPP-1361 printer, and the complete software Commodore titled library. Two prototype 8088 boards (one fixed by me to handle the 8087 math coprocessor), software including Microsoft MS-DOS 1.25 for the Commodore B Series, Digital Research CP/M-86 1.0 and 1.1 for the Commodore B Series as well as a prototype RAM expansion cartridge complete the accessories. None of this is for sale - so don't ask.
Though I'm still looking for a few things like a B256-80 labeled prototype, and one of the original 1982 engineering development prototypes for the B Series. And I still have several CBUG diskettes to find to completely restore the original library. But then that's half the fun of it. ;-)
As a Commodore hobbyist and purist collector, I am totally focused on the B Series systems made in the United States of America and distributed for sale in North American markets only. I am also interested in the peripherals of this period made in Japan for the 110/117V 60Hz North American market and any Commodore-branded software items. The rest of my free space is dedicated to my 1957 Chevrolet BelAir 2-door Hardtop and it’s spare parts. ;-)
When you do the digital archaeology and research, the capabilities of the
6509 architecture combined with the 8088 co-processing, the availability of
both MS-DOS and CP/M-86, integrated monochrome display, integrated dual 1MB
diskette drives, RAM expandable to 1MB - the Commodore B-Series was clearly the
most advanced and state-of-the-art 8-bit computer design ever realized,
if only in prototype form. It is tragic that for all the engineering
investment, it was never fully produced. Sadly the same executive management
forces that brought these designs, and Commodores major success, also were
their downfall. It is also worthwhile to thank the late Norman Deletzke, without whose CBUG there would have been a lot less
known about these computers and nothing remembered.
And so thusly, I was motivated to preserve this little slice of history for all to see and read about. I hope you found it of interest and value and would appreciate your signing the guest page B series registry.
© Copyright Edward D. Shockley, 1998-2021
All Rights Reserved