MC3 - A DIY 8-bit computer
I have always been fascinated by early microcomputers. Ever since I painfully wrote my first machine code program on the 6502 based KIM-1 I have been hooked by these simple and lucid computers. I have constructed various designs in the past and this article is about my latest computer. The first one with some form of name actually. I call it the MC3 since it is a Micro Computer based on the HD6303 microprocessor :) The HD6303 is a variant of the Motorola MC6803 and is produced by Hitachi. The 6803/6303 is software compatible with the 6800 but contain on-board RAM, timer, serial interface, clock circuit and an improved instruction set. It's a lot of old fashion micro computing on a single chip and the HD6303 is still widely available on eBay to this date. Edit: Another good source for 65xx/68xx might be The reason for basing my design on the 6800 family rather than the more community friendly 6502 family is that I already have enough 6502 computers and I really like the structure of the 6800. Having a 16-bit stack pointer and index register is actually great. The 6801 improved the instruction set of the 6800 quite a bit and the 6303 improves it even further with functions such as the XGDX (exchange D and X registers). The MC3 is a small microcomputer that fits on a single 100x160mm board. It is my own design and the intention is that it should be able to run programs (with some modifications) made for the SWTPC6800, EXORciser or other similar computers. It should also be easy to interface to a modern PC for cross-development. This is an image of the CPU board also containing RAM, ROM, address decoding, serial console interface and bus drivers. It's a complete computer! As you see there is not that many components needed for a fully functioning system. For the bus connection I used a standard 40-pin flat-cable connector. The same connector used for IDE hard drives. It is simple, cheap, widely available and has enough pins for my bus. The back side view of the CPU board. I did not use wire-wrap this time as I did with my earlier designs. This board has been soldered. The main reason is that it is getting harder and harder to obtain wire-wrap supplies these days. I was also quite certain that I would not have to make many changes to the design. Soldering this way also makes the board much lighter and thinner than if it was wire-wrapped. This is the final schematic of the MC3 CPU board. Power supply is not included in the schematic. 100nF decoupling capacitors over every chips power supply are also present but not drawn (don't forget these!). Also all logic chips are HCT variants regarding of what the schematic says. I just did not have the correct symbols. All external connections are buffered. These CPU's are not manufactured anymore. Better be careful. Specifications - HD63A03RP CPU @ 1.2288MHz (4.9152MHz external crystal divided by four) - 32 kB RAM - 16 kB ROM - 128 kB addressable I/O space divided into eight 16 kB banks - 16-bit timer - RS232 console interface, 9600bps 8N1 Memory map $0000-$001F --- 6303 internal registers (timer, serial interface etc) $0020-$7FFF --- RAM $8000-$BFFF --- 16k*8 decoded external I/O $C000-$FFFF --- ROM There is also the 6303 built-in RAM placed at $0080-$00FF shadowing the external RAM and can be enabled/disabled when needed using the 6803 RAM control register at $0014. Selecting I/O bank Which one of the eight I/O banks that should show up at $8000-$BFFF is selected by writing to the bank select register at $0002. This is the built in 8-bit port (pins P10-P17) of the 6303 which I use for this purpose. - Bit 0-2 selects bank 0-7. These three pins are directly connected to IC8. - Bit 3-6 are not used - Bit 7 enables/disables interrupts from the bus. "0" enables bus interrupts and "1" masks bus interrupts. This is realized using IC6B. Bus interface All lines are buffered on the CPU board through '245 and '244 buffers. It makes it a little more complex and takes more chips but it's a good protection. D0-D7 --- Data bus. A0-A13 --- Address bus. S0-S7 --- I/O select pins. Active low. One for each page. R/W --- Read/Write signal. Motorola standard. High equals read and low equals write. E --- System clock. Motorola standard. Address and data lines are valid when E is high. IRQ --- Active low with pull-up. NMI --- Active low with pull-up. RES --- Active low reset signal. ROM and monitor program The 16 kB ROM area is actually a 27256 32 kB ROM that is divided in two. A14 is hardwired and divides the ROM in two halves. The high part and the low part. Which makes it possible to have two different monitor programs in ROM and selectable by pulling A14 high or low. I have written a monitor program that has the most basic functionality such as examining and changing memory byte by byte, dumping memory to console page by page, alter the stack contents and uploading S19 data over the console. > h MC3 monitor 1.1 Daniel Tufvesson 2013 G Go (RTI) J Jump to address L Load S19 from console MC Memory change MD Memory dump RR Print contents of stack RS Reset stack pointer RC Change stack CC RA Change stack A RB Change stack B RX Change stack X RP Change stack PC P Select I/O page > md 0000 ADDR 0 1 2 3 4 5 6 7 8 9 A B C D E F 0000 FF FF 87 4A 51 49 54 B2 60 86 9D FF FF 00 00 CA ...JQIT.`....... 0010 FF 0A 30 FF 3F FF FF FF FF FF FF FF FF FF FF FF ..0.?........... 0020 80 00 BF FF FF 55 BF FF 9F FE 95 2C 05 54 05 4C .....U.....,.T.L 0030 9A 65 50 CA 8A 20 25 E0 8E C5 82 23 A8 80 55 43 .eP.. %....#..UC 0040 04 4E A4 FB 25 50 6B E8 B6 CC 60 48 4C CB 57 58 .N..%Pk...`HL.WX 0050 80 AB 62 9E 4A F3 14 90 64 64 53 D9 EA 04 04 23 ..b.J...ddS....# 0060 5B 0A 61 92 9A 26 CD 45 54 08 12 18 48 29 97 94 [.a..&.ET...H).. 0070 50 42 23 00 98 02 0A 65 61 80 C6 0C C6 16 00 D3 PB#....ea....... 0080 7E 00 87 C7 60 81 F1 8E 00 7F CE 80 00 00 2A A7 ~...`.........*. 0090 00 08 8C BF FF 26 F6 CE 80 00 8D 1D A8 00 26 22 .....&........&" 00A0 08 8C BF FF 20 07 0D 0A 2A 20 04 00 87 00 EB 86 .... ...* ...... 00B0 2B BD C0 03 7C 00 83 20 CE DF 85 00 85 00 86 9B +...|.. ........ 00C0 83 39 97 84 CE 00 A6 BD C0 00 CE 00 83 BD C0 12 .9.............. 00D0 BD C0 12 BD C0 15 DE 85 20 C6 60 16 EC AB AA C0 ........ .`..... 00E0 28 AD A8 46 17 A8 01 15 60 0C 11 00 1A 04 AA 01 (..F....`....... 00F0 FF C2 54 8F 81 6C 15 0A 47 D9 A5 C0 47 50 80 FF ..T..l..G...GP.. > Printout of the help command and a dump of the zero-page View of the MC3 system with CPU board on top and two expansion boards underneath.
by Tom LeMense 2013-09-22 15:04 UTC
Hi! What a coincidence - I am in the process of making my own HD6303Y singleboard computer, for about the same reasons as yours. I was in the middle of porting a 68HC11 ROM monitor to the HD6303 when I found your site. I'll still complete the port but it's nice to have a "plan B" if/when my port doesn't work. My specs: HD63B03Y @ 1.8MHz, 64K of RAM, 16K of ROM (banks over upper RAM), FTDI UART-to-USB bridge, RTC7301 realtime clock, Z8536 CIO, and SD-card interface on port 6. I'm happy to share anything of mine with you, if you're interested (schematics, PCB layout, whatever). I'm curious to know what tools you use for code development. It's a bit painful under windows as the tools I'm using are DOS based (so I run them in DOSbox). cheers! -tom

by daniel 2013-09-27 08:13 UTC
Hi Tom! That's really fun to hear. I'm glad that someone else is building new things for the 6303. It's a nice CPU. Feel free to use the material I provide on my site. I would absolutely like to share some ideas and have a look at what you have done. The SD-card interface sounds interesting. Do you have any code written for it yet? I do all my cross-development under Linux using the Emacs editor and an ancient old-school open source assembler. No fancy stuff like macros :)

by Tom 2013-09-29 19:57 UTC
Hi Daniel! No code written yet for anything - I'm doing a bit of prep-work while waiting for my PCBs to arrive from OSHpark - mostly hacking on your MONITOR program a bit (goal is to include a disassembler). I'm planning on working with the old DOS tools in a DOSbox VM under Win7. I'd be interested in using Linux instead, as I would like to do more with Linux, so any tips you have on that would be welcome. Not meaning to sound too lame, but it sure would be nice if there were a freeware C compiler for the 6800. I keep hoping that there's a 68HC11 or 68HC05 C compiler around that could be ported to 6801. So far I have no luck. As for SD card, I don't have any bright ideas for it yet, but I think some block (sector) oriented loader would be easy and within my skillset/attention span. I have been looking a bit at MDOS which was the filesystem for the Motorola EXORsizer. Sure, FAT would allow easy use of SD card with modern OS's, but I'm not going to even think about that without a C compiler (more whining...) Send me email if you like and I'll send PDF of schematic and PCB info. -tom

by daniel 2013-10-03 12:16 UTC
Hi Tom! Feel free to use my monitor as a start for your project! The built-in registers are a bit different in the HD63B03Y compared to my HD63A03RP if I remember correctly and so are the reset vectors but besides that the code should run fine on your CPU. I can dust off the source code for my disassembler if you are interested in including it in your monitor. But sometimes writing it on you own is part of the fun ;) I also wish there was a C compiler for the 6801/6803/6303 but I have not found anything useful yet. The 6800 family is very assembly friendly making the lack of a proper C compiler a smaller issue but it would sure be nice to have one. I initially planned to include a Compact Flash interface for my system but since you have an SD interface now I think I'll go the SD route too so we can share code! Just have to finish my I/O back plane so I can attach an interface. I wrote a simple DOS for the FD1771 a few years ago. It was fun but I think it is time to leave the floppies behind on this one. At least initially.

by James Holland 2014-03-26 12:04 UTC
Hello I have played with a few car ECUs based on these chips and some Toshiba chips that are an evolution of these. For code development I use MiniIDE and TASM on Win7. I weote a brist guide to setting up the IDE here:

by Daniel 2014-04-03 12:09 UTC
Hi James! Thank you for reading. Really good work on the ECU's. It's surprising how many 6301 and derivatives there are out there when you start looking. Does this mean you have cracked the secret on how to extract the contents of a built-in ROM in a 6301?

by Heiko 2014-04-23 17:42 UTC

by James Holland 2015-11-07 22:53 UTC
The contents of the ROM can be read out by placing the micro in mode 0 and running a programme from an external EPROM that reads the addresses of the internal memory and sends them out the serial port.

by Eric Klaus 2015-11-22 16:16 UTC
Hello Daniel; I just started to build myself a simple DIY computer based on the HD6303YCP chip. I happen to have found a number of these, along with all the necessary EPROM, SRAM and interface chips in my basement so it seemed like an interesting challenge. Your schematic and monitor listing will be a HUGE help. Thanks for sharing it. A couple of questions if you have a moment: 1. What assembler program did you use? 2. Ever run across a 6801/6803 C compiler? 3. Where did you locate the HD63A03RP symbol for eagle cad? (searched all over and coule not find anything even close) Thanks Again Eric

by Daniel 2015-12-02 18:52 UTC
Hi Eric! That sounds like a really fun project! The HD6303YCP is a little different from the HD63A03RP I used but the MC3 monitor should be quite easy to adapt. If I remember correctly there are a few more ports, an extra counter and the serial interface is a bit different. If you choose your crystal frequency carefully I think you should be able to achieve a higher speed and baud rate than my MC3. For all code I use an old assembler originally written for the IBM PC long ago. I have the source code as well so I have adapted it to GCC for use under Linux. I will make it available here on the site since I've had a few requests for it. There is no functioning C-compiler that I know of for the 6303. There WAS the old HI-TECH C-compiler for DOS that generated nice code for the 6800 but I have not found it anywhere. It appears Microchip scrapped everything 68xx-related when acquiring HI-TECH years ago. Really sad... Another alternative might be to convert HC11 version of Small-C. I made an attempt a while ago but sadly never finished it. The HC11 has quite a bit in common with the 6301/6303 such as the D-reg. I really think it's doable and it would be a very nice thing to have! The Eagle symbol for the HD63A03RP that I used was hastily put together by me just to finish the schematic. There is only the symbol and no package I'm afraid. I really hope your project goes well. I'm here to help you in any way I can. Please share your result if you like :) Kind regards, Daniel

by Eric Klaus 2015-12-07 01:36 UTC
Hi Daniel; Well after quite a few late nights the MCU board is working! It's simple compared to your MC3 composed of: HD6303YCP with an external clock of 4.9152mhz 32K 27C256 EPROM, 8K HM6264LP RAM. Like you I chose to construct it on a breadboard with soldered wire-wrap wire connections. I found an assembler that works OK here: So far I have the serial I/O connected to a USB-TTL Serial adapter to my PC running the puTTY terminal program. Besides the basic examining, inputting, and executing functions I came up added functions to accept intel IHEX records from the serial port and load them into RAM, very handy. I've also got crude Dissemble and Assemble from/to memory functions. Once I get a chance I'll post the project to share. Thanks for the help and encouragement. Eric

by Eric Klaus 2016-11-28 02:41 UTC
After almost a year I picked this project up again. It was working pretty well but I still found some bugs. I used the BASIC you posted and even added SYS, PEEK & POKE command making it a bit more useful for a system like this. Details are posted here: Thanks again your postings really helped. Eric

by Daniel 2016-12-01 21:01 UTC
Hi again Eric! Great work! I'm really excited to see another 6303 build. I just played a set of your MasterMind game running om my MC3 system. Great fun! You really have feature rich monitor program. Fun to see a built in assembler and disassembler. Thank you for sharing :)

by Eric Klaus 2016-12-03 17:30 UTC
Hi Daniel; Glad to see you were able to port the game over. I just finished building another board! (crazy huh?). The proper chip sockets arrived a couple of days ago. Mostly the same as the first one, just a bit neater. updated pictures at:

by Steve 2017-03-12 01:53 UTC
Hi Daniel. I frequently check up on you site for progress of the MC3. No new updates for a while - hope all is well and you're still investing time in the good old 6303... I've lost your email address so I couldn't PM you... sorry.

by Daniel 2017-03-15 09:07 UTC
Hi Steve! Thank you for writing. I can assure you the the 6303 is not forgotten (and hopefully never will be). A lot of things have gotten in the way. It saddens me to see my web site in such state of neglect. I have a some upcoming things I think you will like though. Will try to get a few new posts online soon!

by skipp 2018-08-11 19:48 UTC
Hello there, I tripped on to this page from a Google search for anything related to the 6803 processor. Here it is in August of 2018 and I'd been reverse engineering the firmware from an old ham radio repeater controller using the 6803, both the Motorola MC and the Hatachi HD version, slight difference at the clock portion of the circuit. I found a lot of free and very functional resources are still available to those who really want to dig around. If anyone reads this, I'd be interested to know what type of Editor Assembler with Linker you might be using. I'd actually just like to find one that really works. I can run an old DOS program, no problem. I was very pleased to find enough free resources that run in a common Windows XP machine. There's a few really nice emulators and hex editors out there. Please feel free to Email me direct at skipp025 at jah who dot calm if you are able to provide some suggestions. I bought a few original NOS micro processors from a reliable source (not so easy to do these days). The basic controller I have is really simple, just a small number of interface buffer chips and the firmware stuck on a 27c32 Eprom. The original code runs entirely from the internal ram inside the CPU, just a marvel at doing something with a limited ram budget. I can now modify the original code, but I'd also like to find a real 6803 so I can generate some entirely new code. cheers, skipp

Write a comment

Name or handle

E-mail (optional and not visible to others)


Code from above