Resurrecting an MTX500.

[stephen_usher]

All the data lines float to 1.32 volts, which should be “low”.

It's the LCDx bus lines which are driving the Dx lines high via 2E. All the LCDx data lines are floating at 1.68V. With 2E removed all the data bus lines float to zero (though this can be VERY slow, as if a capacitor is discharging).

The system still doesn't get anywhere though, unfortunately.

[stephen_usher]

The A5 line looks a bit strange...

IMG_1418.jpg (441.19 KiB) Viewed 14759 times

Seeing that A5 only visits the ROMs, 6C and the Z80 and removing all but one ROM, placing it into the ROM A socket and changing that ROM makes no difference. Do I get the feeling that 6C is unwell? 6D has been replaced as that was definitely dodgy. Bad batch?

After the holidays that's the first chip to change.

Note: MA5, the other side of 6C looks sharp but intermittent and a lot of the time high, which would make sense.
Note 2: The image shows two traces from probe 1 overlain on the screen by the scope rather than two inputs.

[Martin A]

I've chopped up a bit of scrap matrix board and added a 32 pin and 20 pin socket plus some jumpers.

to come up with this mock up:

rom board idea.jpg (33.73 KiB) Viewed 14758 times

A 20 pin 16V8 GAL or CPLD has more than enough capacity to allow for paging in 4 different rom Images from a 32k, 28 pin EPROM or EEPROM.

However EPROMS are a pain for development because of the erase time. Electrically erasable memory is easier to work with, and it turned out only 1 pin has to change function when fitting a 32pin socket to use a 64K EEPROM like a 27E512 or a 128K flash like the 39SF010 as well as a 27C256.

Since the larger capacity device is going to be fitted at least for development, I decided to add some extra options. So the board has 5 jumpers.

One jumper deals with the pin that changes function. On 32k EPROMs where it's the program enable pin that needs to be pulled high and the larger capacity chips where it's an address line and needs to connect to the paging logic.
One jumper decides whether the diagnostic rom images are available to be paged in, or if the normal ones are.
The other 3 decide exactly what rom images are active.

With the normal roms setting It would be one jumper for each of the roms A B and C to individually selected to replace the on-board ones.

With the diagnostic rom setting the extra space on the larger devices can be used to test the rom paging system from within the diagnostics. Test with WINcupl suggest switching the 16 ROM images that would fit in a 128k FLASH chip is doable without needing to move to a larger GAL/CPLD.

Hopefully I'll get some time over the weekend to put some wires onto the mock up and see what happens when its connected to a real MTX!

[stephen_usher]

Wow! That’s fast work!

[stephen_usher]

I think it's dead jim!.

After removing the two 6C and 8A and putting in sockets I refitted the chips to check that everything was working as before and it was. I removed the chips and was testing the address and data lines when suddenly...

Flash! Crackle! Down around ROM C socket pin 1 (which has nothing connected into it) and magic smoke coming from the power conditioning part of the board.

I have no idea what could have shorted down there as there's nothing there.

That's probably the end of the machine.

[Dave]

ROM Pin 1 is +5VDC - perhaps you grounded it by mistake and blew something in the power circuit?

If you're lucky, it might just be a cap or something, if you're very unlucky, you might have melted a track, either way, the machine can probably be saved, don't bin it just yet

[stephen_usher]

I wasn't anywhere near it with the 'scope probe at the time. I may have to desolder the ROM C socket to have a look.

It'll have to wait until I get back from my holiday now.

[Martin A]

ROM Pin 1 is +5VDC - perhaps you grounded it by mistake and blew something in the power circuit?

If you're lucky, it might just be a cap or something, if you're very unlucky, you might have melted a track, either way, the machine can probably be saved, don't bin it just yet

I'd echo Dave's comments, because the MTX is all standard parts, you could blow pretty much every component on the board and STILL be able to re-build it!

[stephen_usher]

Don’t worry, I was never going to throw it away anyway.

I’ll have to investigate the dead short between the +5V and 0V lines. I’m guessing a stray bit of solder spatter from when I used the solder sucker has gone somewhere it shouldn’t underneath the ROM C socket.

[1024MAK]

Assuming it was the +5V that was shorted out, the most likely item in the power supply section of the board that will have been damaged is the current boost transistor, Q4 (TIP2955). That’s the large power device on the heatsink.

Also check it’s the associated resistor (R62, 10 ohms) and the 7805 voltage regulator (REG2). It is also possible that the fuse (FS1, 3.15A) blew.

See this page.

Mark

PS you have to be careful posting useful information, as very quickly it gets sucked up and then at a later date, appears on www.primrosebank.net