OK, on second thoughts, I don't think I'll be able to get much useful info of TestRam...
Seems more probable to me that VRAM has gone nuts. The VDP is not usually the culprit of this kind of faults.
Does TestRAM have a print option?
My bet is also that the VRAMs are defective.
You can find 4116 compatible DRAMs on eBay.
But, if it ends up being the VDP, you the TMS9929 can also be found on eBay.
I have repaired a similar problem on my Hit Bit 75P.
The problem in my case was in the power supply. There is a small fuse, with a transistor-like package (but with only two pins instead of three), that i managed to burn short-circuiting the -12V on the slot cartridge (i was a kid too and didn't realize it was dangerous for my computer).
This, burned out the fuse and also stopped the -5V line required by the VRAM. Check if you have a similar problem before desoldering the VRAM chips.
Here you can find a detailed analisys of the problem http://ilblogdiwakko.forumdeitroll.com/?p=78 but it is in italian.
The OP guy mentioned he inserted a cartridge backwards. That would break the -12v power supply.
That Motorola DRAMs are of the earliest type (nMOS) which need +5v, -5v and 12V to function correctly.
You need to have -5V on pin 1, 12V at pin 8 and 5V at pin 9 for it to operate correctly.
This certainly align with his story about inserting a cartridge backwards and making the computer stop working (the likely the power supply blew output fuses)
Edit: I just read BernardBernoulli's article and it's exactly what I was just talking on my reply. Good stuff and hopefully the OP will get it working again.
P.S.: There are (slightly more modern) replacements for the early nMOS 4116 DRAMs which do not need 12V and -5v to operate.
Checking back in for the first time in a couple of years, so happy to see some new ideas for a solution! I desoldered all the VRAM chips and put in some new ones. Without success, the graphics remained severely glitched.
Thanks BernardBernoulli and l_oliveira, I will definitely give the proposed PSU fix a try! Will check for correct voltages on the chips first.
Update: I am indeed getting the wrong voltage at pin 1 of the VRAMs, around .1V. Will order the right fuse and try to replace the old one!
In other words, I am observing the exact same condition as in the Italian article linked above:
This chip is rather particular because it to function properly it needs three distinct voltage levels: + 5V, + 12V and -5V. Both + 5V + 12V that were present ... however, when I went to measure the 5V voltage meter indicated a slightly positive (+ 0.1V).
You may want to refrain from powering up machine until psu is fixed. Read: disconnect psu from rest of computer, repair, check that all voltages are present again. If okay: reconnect, power up machine, and verify voltages a 2nd time with everything running. Only then go check individual RAM chips.
Reason being that these 4116 DRAMs tend to die quick when a supply voltage is missing (especially their -5V!). Meaning the old RAM chips may still have been okay, but dying while you're busy checking voltages in a powered-up machine. Same thing with the new RAMs put in. So better pay attention what you're doing, and in what order.
Thanks for that advice, RetroTechie! However, I already proceeded somewhat, let's see if I managed to make things worse or not...
I inserted a jumper in place of the ICP-N15 fuse. Powered up, and get the correct voltages on all the VRAM chips. Here they are in their sockets (all chips were already replaced):
And here is what's happening right now...
Hit Bit start screen:
It seems like every other byte/sprite is repeated in the graphic mode? BASIC becomes BBSSCC, for instance. And the Hit Bit logo seems to have some repeating patterns as well. Still, there is a definitive improvement since before I put in the jumper for the -5V (see images at top of page).
BASIC start screen:
Again, an improvement since before the -5V jumper was put in place. But it seems like about half of the characters are shifted one slot up (MSX -> LRX, Ok -> Lj, etc) when output to the screen. This also applies to the input from the keyboard (no screenshot of this). Most characters typed on the keyboard are printed as something else, causing syntax errors.
Here's a youtube clip of what happens when I insert the Yie Ar Kung Fu 2 cartridge:
https://youtu.be/4D1BcBeKlqM
Again, characters are shifted one slot (HI-SCORE -> HH-RBNRD), and there are some garbled graphics.
Summing up, this seems to be the observable status:
- Text mode seems to be OK, graphics wise.
- Characters are exchanged in input and output.
- Graphics mode seems somewhat garbled.
- The game cartridge isn't detected on every startup.
Probable causes?
- Is the ROM damaged somewhat (character handling via CHPUT/CHGET)?
- Or is there some RAM problem causing the character displacement, as well as the graphics repeats/glitches?
- Or are there actually two problems here, one causing the character munging (ROM or RAM?) and one causing the garbled graphics in hires mode (VRAM?)?