DRAM control can be done in several ways (early write, hidden refresh.. it's complicated). But usually, /CAS signal(s) work like a chip select. So remaining high just means that DRAM isn't asked to write or output (read) anything. Which can happen if eg. a faulty ROM causes the machine to get stuck in a loop. Or some other issue makes the machine not get to the point where it searches for / uses the RAM.
But wouldn't a healthy Z80 always be fetching instruction data if it's getting a clock signal? Even if it's in a loop, it should be constantly fetching from memory, right?
Just to give this some closure: I ended up tracking down a working Toshiba HX-10 and started reverse-engineering the circuit diagram and checking things on both. Everything checked fine and I didn't see any obvious problems with the faulty one. I also swapped all socketed ICs just in case, but nothing.
So I decided to desolder and socket the TCX-1007 (which is the big custom IC) and... bingo. That was the problem :-(
Anybody knows where I can buy a TCX-1007 or am I out of luck with that machine?
This chip seems to be the MSX engine of Toshiba. The only way to repair the machine is to get a broken motherboard with the working chip. I could buy the broken Toshiba from you if you decide to sell.
I would like to add to this:
I also had a black screen HX-10 on the bench to fix for someone. Similar symptoms as above: no main ram CAS activity. In my case it also turned out to be a corrupted BIOS ROM, i.e. CPU executing garbage instructions, never actually selecting the correct RAM slot, and hence no activity on CAS. A replacement EPROM fixed that.
I also have a non-functioning Toshiba HX-10. Black screen,...
I have a few questions:
- What is IC16 on the board: it has a heatsink and the pins are coloured black...
- What is the purpose of the toggle switch in the cartridge slot. if i press it I loose my video output...
- how come the bios rom can be corrupted (MB83256). It doesn't have a window, so light can't erase bits, so how can this be...
O.K, quick update, made myself an arduino based DRAM tester and tested all the D4164C chips(IC7 - IC14) and it turns out two of them were bad (IC8 and IC12). I have bought some new ones and put in chip sockets for all the DRAM.
Now all is well and I have a perfectly functioning HX-10, thanks for all the help and suggestions.
Could you tell something more about how this arduino tester works?
- What is IC16 on the board: it has a heatsink and the pins are coloured black...
The HX-10 PCB differs between models; but it must be the VDP or the PSG. If it's the european model, most probably, the VDP.
Quote:
- What is the purpose of the toggle switch in the cartridge slot. if i press it I loose my video output...
It mechanically resets the machine when you plug in a cartridge. If the machine is on, this serves to momentarily switch-off the power and avoid shortcircuits.
Quote:
- how come the bios rom can be corrupted (MB83256). It doesn't have a window, so light can't erase bits, so how can this be...
ICs always degrade with time, only one "rotten bit" in a mask ROM and your mileage may vary, from not noticing anything bad, to making the machine unusable.
