Page 22 of 23

Re: Resurrecting an MTX500.

Posted: 19 May 2019 16:05
by 1024MAK
If the resistor is getting hot, yep, a new TIP2955 is needed.

Mark

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 15:53
by stephen_usher
Due to stuff not arriving for other machines I've had time today to get the machine out and replace the TIP....

And it seems to be working OK!

CP/M doesn't crash all the time (though reading the filesystem can sometimes fail). :D

Methinks we have another fully operational Memotech!

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 16:12
by stephen_usher
Bother! Spoke too soon. After half an hour it started playing up again.

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 16:32
by stephen_usher
OK, I'm back to almost instantly crashing in CP/M.

This image is what I manage to get when typing 'dir' just before the machine locks up:
IMG_1476.jpg
IMG_1476.jpg (185.53 KiB) Viewed 8637 times
Something is either heat sensitive or it's a capacitor which heals itself if the machine is left turned off but starts to degrade as soon as it's powered.

The problem is that when I've 'scoped in the past I've not seen massive amounts of noise on the address or data buses, well nothing large enough to corrupt data anyway.

Also, BASIC is rock solid, as are games which run only in the lower 48K of RAM.

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 18:29
by stephen_usher
Damn and blast! Back to square one due to tiredness and stupidity.

I noticed that the case was getting hot around the heatsink area, so I decided to test the current flow. So I reconfigured my multimeter to the 20A mode by plugging the probe in the other port. I checked the 5V line current draw and with the RS232 board disconnected it was marginally higher (0.9A) than my other MTX but nothing major.

I wanted to test to see if the TIP was getting proper power and the other power transistors were working, so took the video board out as that's the only way to get to the pins. I changed the dial over to the 20V range and the multimeter started beeping at me, but as it was also showing low battery I thought it was warning me of this. Hooked up the black lead to the modulator case and probes the left-most pin of the left-most transistor and.. *spark*! What the heck?!!!

Then I noticed the probe position.

I'd just sent 20V down the ground, i.e. -15V to all the chips on the board.

The 3A fast blow fuse had gone but obviously not before something or many things had fried. The machine just beeps again, so at least the sound chip isn't fried.

So, it's time to start checking stuff one by one again.

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 18:49
by Dave
Looking on the bright side, at least you have most of the ICs in sockets now :)

On the not so bright side, if you really have applied that voltage to all of the chips, there’s a good chance that all,of the VRAMs (as a minimum) will be blitzed, they are particularly sensitive to voltage levels and power up order

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 19:18
by stephen_usher
Dave wrote: 08 Jun 2019 18:49 Looking on the bright side, at least you have most of the ICs in sockets now :)

On the not so bright side, if you really have applied that voltage to all of the chips, there’s a good chance that all,of the VRAMs (as a minimum) will be blitzed, they are particularly sensitive to voltage levels and power up order
Yeah, I know. The CPU's definitely toast, but something is asking for things on the address bus. The data bus is just horrid spikes though.

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 20:37
by stephen_usher
Actually the CPU is fine. After pulling all the memory and fitting the Diag ROM I at least have something on the screen!

Video RAM is fine, as is the VDP. The chips running the keyboard seem to be dead, quite possibly both the buffer and the other chip.

One thing to note with the Diag ROM. If the RAM chips are truly bad, as in fried, even though the ROM doesn't use RAM the system doesn't boot.

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 23:09
by stephen_usher
Well, the address decoding's broken. If I have ROM B or ROM C in circuit at the same time as the Diag ROM then it doesn't start up. The EEPROMs test fine in the writer so it's the machine.

I feel that there's probably a full set of chip replacements in order... but not until September. It's going back into the loft in the morning.

Re: Resurrecting an MTX500.

Posted: 08 Jun 2019 23:59
by Martin A
The address decoding is 90% done in the PAL/GAL so that ought to be testable in the programmer.

The last 10% being done by the 74LS00 at J10 selects between rom B and rom C if the page roms are active.