http://www.6502.org/users/sjgray/projects/microuk101/
PBC for the UK101 project by Grant Searle.
About small SBC systems
http://www.6502.org/users/sjgray/projects/microuk101/
PBC for the UK101 project by Grant Searle.
Somehow during the migration I managed to loose the OSI Model 300 trainer page.
Restored and with some additions!
The Micro-KIM is a SBC designed and built by Vince Briel. First sold in 2007 and alas not available anymore.
Based upon the ideas of Ruud Baltissen how to replace the 6530 with a 6532, it was the first KIM-1 clone that was running the unmodified KIM-1 ROM.
– Manuals, circuit diagrams, single step fix, tape devices, support Cd images
– Photos
– Videos
Specifications
See the KIM-1 page for more KIM-1 info and programs.
First impressions
Most of the First Book of KIM programs work. Some require the presence of the second 6530 (6532 here).
See the KIM-1 software and manual pages. Also see the photos section for TinyBasic and KB 9 Basic.
See the PC utilities page for papertape and other conversion utilities.
Micro-KIM | KIM-1 | ||
$0000-$03FF | 1024 Bytes of RAM | $0000-$03FF | 1024 Bytes of RAM |
$0400-$07FF | 1024 Bytes of RAM | $0400-$07FF | Optional Memory Area |
$0800-$0BFF | 1024 Bytes of RAM | $0800-$0BFF | Optional Memory Area |
$0C00-$0FFF | 1024 Bytes of RAM | $0C00-$0CFF | Optional Memory Area |
$1000-$13FF | 1024 Bytes of RAM | $1000-$13FF | Optional Memory Area |
$1400-$16FF | Optional Memory Area | $1400-$16FF | Optional Memory Area |
$1700-$173F | Optional 2nd 6532 I/O, Timer | $1700-$173F | 6530-002 I/O, Timer |
$1740-$177F | 6532 I/O and Timer | $1740-$177F | 6530-003 I/O, Timer |
$1780-$17BF | 64 Bytes RAM from 6532 | $1780-$17BF | 64 Bytes from 6530-003 |
$17C0-$17FF | 64 Bytes RAM from 6532 * | $17C0-$17FF | 64 Bytes from 6530-002 |
$1800-$1BFF | 1024 Bytes of EPROM | $1800-$1BFF | 1024 Bytes of ROM in 6530-003 |
$1C00-$1FFF | 1024 Bytes of EPROM | $1C00-$1FFF | 1024 Bytes of ROM in 6530-002 |
$2000-$FFFF | Unused memory | $2000-$FFFF | Unused memory or 32K RAM baord |
* The 6532 has 128 bytes of RAM vs. only 64 bytes on the 6530. The Micro-KIM utilizes all 128 bytes from
the single onboard 6532 so all original memory locations are available.
65C02 Single step fix by Timali
My Micro-KIM shipped with a 65C02, and apparently there is a timing issue which prevents single-step from working with the 65C02. I tried an original NMOS 6502, and single-step worked ok with it, but not with any of my 65C02s. I did some debugging with my scope and determined that there is a small timing difference causing the SST signal to be erroneously asserted (pulled low) for 100-200 ns during EEPROM accesses with the 65C02, which is just enough to cause a problem. The easiest way I could think of to fix this was to delay the SYNC signal briefly with a small RC circuit, which prevents the glitch in the SST signal. I cut a trace on the back side of the board and added a small resistor and capacitor, and single-step is now working correctly with my 65C02’s. It still works with the original 6502, also. Click on the image to see a larger picture.
Tiny Basic
Microsoft Basic KB9
https://www.youtube.com/watch?v=R_zD5T_khKs
The base for all ‘modern’ KIM clones, KIM reproductions, MICRO KIM, PAL-1 and more.
Credits to Ruud Baltissen for the idea and details. ‘I’ means Ruud in this page!
RRIOTs are mask programmed for address selection, choices are:
Pin 18 PB6 or CS1
Pin 19 PB5 or CS2
Pin 17 PB7 can have a pullup
ROM selection on RS0, CS1 und CS2 s
RAM on RS0, CS1, CS2, A9, A8, A7 and A6
I/O on RS0, CS1, CS2, A9, A8, A7 and A6
See the various 6530/6532 datasheets for (some) more detail.
The KIM-1 has two 6530s on board. For more info about this IC, please read the datasheet. Anybody who is a little bit familiar with the hardware market can tell you that you cannot buy the 6530 anymore. Happily enough there is another IC available which you could call its brother: the 6532. The 6532 has 16 I/O-lines, an internal timer and 128 bytes of RAM on board, but no ROM. The internal ROM of the 6530 can be selected independently from the I/O. So for this project we’ll use an external EPROM as replacement. The pin out of the 6532 is completely different but that should not be a problem.
The next difference is the fact that the 6532 has a separate IRQ and PB7 line. As we will see, the functionality of both lines is the same as with the 6530. To create the same circumstances we only have to connect them together.
The third difference is the availability of PB6 with a 6532. See it as a bonus as I haven’t found any reason how it could jeopardize our project.
The fourth difference is that it is possible to generate an interrupt depending on the behaviour of PA7. But this is an option, which is out of function by default after a reset.
The last and major difference however lays in the way the registers are selected:
function: RS: A6: A5: A4: A3: A2: A1: A0: R/W: RAM 0 x x x x x x x x DRA 1 x x x x 0 0 0 x A DDRA 1 x x x x 0 0 1 x B DRB 1 x x x x 0 1 0 x C DDRB 1 x x x x 0 1 1 x D PA7, IRQ off, neg edge 1 x x 0 x 1 0 0 0 F PA7, IRQ off, pos edge 1 x x 0 x 1 0 1 0 G PA7, IRQ on, neg edge 1 x x 0 x 1 1 0 0 H PA7, IRQ on, pos edge 1 x x 0 x 1 1 1 0 I read interrupt flag 1 x x x x 1 x 1 1 E read timer, IRQ off 1 x x x 0 1 x 0 1 J read timer, IRQ on 1 x x x 1 1 x 0 1 K Clock / 1, IRQ off 1 x x 1 0 1 0 0 0 L Clock / 8, IRQ off 1 x x 1 0 1 0 1 0 M Clock / 64, IRQ off 1 x x 1 0 1 1 0 0 N Clock / 1024, IRQ off 1 x x 1 0 1 1 1 0 O Clock / 1, IRQ on 1 x x 1 1 1 0 0 0 P Clock / 8, IRQ on 1 x x 1 1 1 0 1 0 R Clock / 64, IRQ on 1 x x 1 1 1 1 0 0 S Clock / 1024, IRQ on 1 x x 1 1 1 1 1 0 T In total 5 address lines are used, meaning 32 registers. But 11 of the 19 registers have one or more mirrors. Read: J E J E K E K E J E J E K E K E Write: F G H I F G H I L M N O P R S T R/W: A B C D A B C D A B C D A B C D
As we can see, the last 16 registers equal the 16 of the 6530 itself.
So now we have to develop some logic which will do the following:
Conclusion:
Here we have a luxury problem. We only need 2K of (EP)ROM like the 2716. The problem is that the 2716 is hard to find and more expansive then the 2764 or its bigger brothers. When we use a bigger EPROM we only have to tie the unused address lines to GND. The same problem occurs with the RAM.
If we have to use bigger RAMs or EPROMs anyway, it is quite easy to use other parts of that chip by OR-wiring the CS-line with more Kx-outputs of the main 74145. (Don’t forget the address lines!) In case of the EPROM we also can tie switches to the surplus address lines and have the advantage of a multi-KERNAL system.
What are the major differences with the original circuit:
You may notice that Ruud did not change things which are more or less obvious like replacing the clock circuit by a module or replacing the various 74XX TTL-ICs by their LS or HCT equivalents.
(Thanks to J Coville)
Back in the day, a designer ordering a custom 6530 would have to specify certain parameters in addition to the ROM contents. During my KIM-1 restoration effort, I found a datasheet for the Synertek version of the part. The datasheet has quite a bit of information describing how to provide the ROM contents. There is also a form for “Additional Pattern Information” (tables to describe the desired chip-select and addressing information). They look like this:
Chip Select Code (Check one square in each block)
|
|
|
|
ROM/RAM/I-O SELECTS (Specify H or L or N (don’t care) in each box.)
RS | CS1 | CS2 | A9 | A8 | A7 | A6 | |
ROM Select | N | N | N | N | |||
RAM Select | |||||||
I/O Select |
As best as I can determine, the designers of the KIM-1 filled out the tables for the RRIOTs like so:
6530-002:
|
|
|
|
RS | CS1 | CS2 | A9 | A8 | A7 | A6 | |
ROM Select | L | H | N | N | N | N | N |
RAM Select | H | L | N | H | H | H | H |
I/O Select | H | L | N | H | H | L | H |
6530-003:
|
|
|
|
RS | CS1 | CS2 | A9 | A8 | A7 | A6 | |
ROM Select | L | H | N | N | N | N | N |
RAM Select | H | L | N | H | H | H | L |
I/O Select | H | L | N | H | H | L | L |
In a KIM-1, K5 is asserted low by addresses 1400-17FF. It is connected to the CS1 lines of both the 6530-002 and 6530-003. A9 and A8 need to be 1 for all I/O and RAM, and A6 and A7 determine which chip and I/O or RAM you get:
0001 0111 00XX XXXX = 1700 – 173F = 6530-003 I/O
0001 0111 01XX XXXX = 1740 – 177F = 6530-002 I/O
0001 0111 10XX XXXX = 1780 – 17BF = 6530-003 RAM
0001 0111 11XX XXXX = 17C0 – 17FF = 6530-002 RAM
The RS lines are connected to different select lines, K6 for 6530-003 and K7 for 6530-002, this gives:
0001 10XX XXXX XXXX = 1800 – 1BFF = 6530-003 ROM (1k)
0001 11XX XXXX XXXX = 1C00 – 1FFF = 6530-002 ROM (1k)
All these values agree with the KIM-1 memory map documentation.
K0 $0000 – $03FF 1024 bytes of RAM (8*6102)
K1 $0400 – $07FF free
K2 $0800 – $0BFF free
K3 $0C00 – $0FFF free
K4 $1000 – $13FF free
K5 $1400 – $16FF free
$1700 – $173F I/O, timer of 6530-003
$1740 – $177F I/O, timer of 6530-002
$1780 – $17BF 64 bytes RAM of 6530-003
$17C0 – $17FF 64 bytes RAM of 6530-002
K6 $1800 – $1BFF 1024 bytes ROM of 6530-003
K7 $1C00 – $1FFF 1024 bytes ROM of 6530-002
K0..K7 = output lines from 74145
Have a KIM-1 with faulty 6530? Look here for replacement with 6532 and some.
On this page two KIM-1 replacements are presented.
Based on the same design principles of 6530 with a 6532 + ROM and glue logic, as also used in
Both solutions are made up of a PCB with 6532 + ROM and glue logic. The defective 6530(s) need to be replaced with an IC socket, the replacement board is fitted in the IC socket on top of the KIM-1.
Which one you choose depends on the state of the KIM-1. If only one of the 6530s is damaged, the Corsham solution is the right one. If both 6530s need replacement the Eduardo Casino design is the appropiate solution.
The DOS65 information is now complete. Last year I did document most of the system. And now I scanned the large pile of paper with manuals and source listings that were waiting to be added. Source listings of ROM, the Operating system and the Monitor and other documents. Most are in English! Enjoy!
I have added the excellent html formatted manuals by Erik van den Broek to the KIM-1 Manuals page.
Groepaz from the VICE emulator team and the C64 demo group Hitmen just released an updated PDF of 6502 illegal or perhaps more accurately called unintentional opcodes.
I would say this document and the use of illegals are for advanced programmers but they can come in very handy for generating smaller or faster code at times.
Some illegal opcodes can be unstable on certain chips
NMOS 6510 Unintended Opcodes no more secrets (v0.91 – 24/12/16)
San Bergmans has released SB-Assembler 3
Now written in Python 3, runs on Windows, Linux, Max OS.