osiweb.org

Hi All,
As I was getting a little burnt out on fixing the OSI MPI floppy drive (and I am waiting for an alignment disk to arrive) I decided to shift my focus to getting a spare 502 board working. The board was very hacked up, and the first thing to do was to remove all the wires, and fix the cut traces to get the board back to what I hoped was it's original state. This took some time as I needed to replace some parts (mostly capacitors) that were obviously damaged, probably due to bad storage. With a lot of offline help from Dave I proceeded to test the board. Surprisingly the board fired up with the H, D, M menu, and I was able to drop nicely into Monitor. As the board only had a TMS2716 Monitor rom installed, one of my goals was to install the missing four BASIC ROMs. I actually spent a few days chasing a RAM error that wasn't even there...sorry about that Dave.

After a bit of initial failure I was able to burn and install four 2716 (not TMS type) EPROMs containing BASIC 1 thought 4. To install 2716 EPROMs instead of the 2316 ROMs I had to jumper pins 18 (on all EPROMs) from 5V to GND. Pin 20 was also an issue and I had to jumper these so that they fed directly into the 74LS138, skipping out the 7404 inverter. Once this was done I re-jumpered the address and BASIC worked like a charm.

Next issue was to get RS232 working. This turned out to be quite a challenge, and Dave's knowledge was invaluable here again. After some false starts, mostly because I was using my Logic Analyser for the first time and was learning the software as I went. first I installed the resistors and transistors needed, and using the LA I was finally able to confirm that the board should be running at 300 baud. I tuned the trimming pot so that a single serial pulse was exactly 3.3ms long and then I was able to receive data. The Serial out is not yet working, but I believe that is more an issue with my machine not having the -9V that is usually require. I have confidence that I can get this working sometime soon.

The final thing I have done was burn another 2716 EPROM with SYNMON code and replace the TMS2716. This also involved removing a jumper wire that was previously installed to hack in the TMS variant. I dislike the TMS 2716 EPROMS. Unlike the standard 2716's they are total oddballs, and it's very hard to find a programmer that can read and program them.

With serial-in finally working I was able to run RAM tests and confirm that all the 2114 RAM was working correctly, and finally load some basic programs.

Next I want to try to get my D&M Micro CEM-CM9 card working to combine it with the 502 card. My ultimate goal is to use this setup to test and run the data-splitter paddle board that Dave (and to a far lesser degree, I) designed. This config will make my OSI a 3 board machine, with plenty of internal space for the paddle board.

Philip

Now that serial data-in is working, I spent the evening tracking down the issue with data-out. It took some effort but I was able to find the problem.

Just on the off chance, I actually built a identical circuit comprising of the transistor and two resistors outside the the 502 board, then fed the TxData signal from the ACIA into it, and the TxData coming out I fed directly into the serial cable. I also made sure I had a jumper from the -9V line to GND, as my machine does not have a -9V power supply. Surprisingly it worked. I was able to receive data no problem. It was then a relatively simple task to track down what was different between the circuit I built and the 502 circuit. It turns out that on the 502 board the TxData signal as it leaves the transistors and resistor section splits, with one part going to the serial cable and the other going to pin 2 of U41. I just ended up cutting the trace to U41. I also permanently cut the -9V trace and jumped it to GND. Presto, it now all works perfectly.

Philip