When I recieved this MSX2, I had concerns as it had certainly already had a reapair attempted on it. What if this person had tried everything and discovered that a custom chip had blown and it was unrepairable?
Well, I decided to carry on regardless and work through the issue myself. To start off with I checked the basics on the CPU. Clock signal, reset circuit, data and address buses. They all looked fine.
The video circuit all seemed to be operational, I could see the 15khz sync signal but the composite video output was just black.
I noticed that the CAS signal for the main RAM was missing, which suggested that the CPU wasn’t successfully executing code, even though there was a bunch of activity going on. My first assumption was bad memory. I socketed both the video RAM and the main RAM and swapped chips around but it made no difference.
I decided to add a useful tool to my inventory and purchased a mini DRAM tester off Ebay.
One by one I inserted the DRAM chips and they all tested good. Not quite as simple as a RAM fault then.
After reading through the schematics for the Sony HB-F1XD (The closest MSX2 I could find to my model) I noticed the VDP read and write signals came from a custom IC on the board labelled as a MB64H444.
I checked with my thermal camera and noticed this chip stayed completely cold when powered on. My first concern was maybe this chip was completely dead which would be game over for this computer as they are virtually impossible to purchase.
I found the pinout for this chip and probed every pin with my scope. Everything seemed to be in order.
It was at this point that I put the scope back on the composite video pin and noticed it was showing a signal!
How could this be? I’ve not changed anything!
Even though I don’t have the replacement hic board yet, I hard wired the composite video output from the vdp chip into my little monitor and there it was, a perfect working image.
Although I was happy that the computer was working and that all of the custom chips were good. I don’t like things just fixing themselves as it can always break again.
And that is exactly what happened. But I discovered that by putting some pressure on the board I could break and fix the computer at will.
My next job was to find the bad connection, so I started re-flowing the solder joints on the chips in the area where I was bending the board. I also reflowed the solder on the MB64H444 chip. At this point the computer was non working and bending the board wouldn’t bring it back to life. Back to square one 😞
Only this time, I knew all the components were working and I was looking for a bad trace or solder joint.
I went back over the MB64H444 IC as this was the last thing I touched, and that is when I noticed no activity on the A13 address line. Could it really be that simple?
As a temporary test, I soldered a bodge wire from A13 on the MB64H444 up to A13 on the ROM chip.
I crossed my fingers and powered on the machine which greeted me immediately with the MSX logo. And no amount of flexing the board would cause it to not work.
The entire computer brought to a halt because of a bad trace on one of the address lines to this MB64H444 chip.
I’m still not 100% sure what role this chip plays in this computer. The schematics list it as a speed controller. So I will be doing some more reading up now to see why this issue caused the machine to not boot.
I have alsp sourced a modern replacement for the HIC board now, so once this arrives, I will be able to put this computer fully back together and start using it.
There is something so satisfying when you finally find the issue with these old machines.
If you have anything sat in your loft broken and you fancy donating it, I will do my best to bring any computer back from the dead and give it a good home 😀
Be the first to comment