• Please review our updated Terms and Rules here

Torch Triple X c1985.

Right. So in a fit of pique I switched the ATX supply on and left it on. Turns out, I've worn the on/off switch out... DOH! Shorting the switch and it's absolutely fine.

I have desoldered EVERY ceramic cap and tested them all. There are a group of about five 100nf that are reading less than 80nf so I will swap those out. And there's one that should be 100nf but it's reading a couple of tens of pf.

I do have a bench power supply so I will try what you suggested.

I am really keen to get this thing working despite throwing my teddy out of the pram the other day.. 🥴
 
Minor update. I connected the bench supply to the board with all accessories removed. Current limited to 2amp. Supply went straight into current limiting (as predicted). The voltage across all of the chips was 1.732v

I tried removing the three main chips (CPU, MMU and DMA controller) and the voltage increased to 2.04v but still current limited at 2 amp.

With the main (precious) chips removed the board is drawing 3.73amps at 5.1 volts. That seems pretty high for a bunch of DRAM and logic chips. That's about 19 watts. 😬

When I limit the current on the board and measure voltage is it right to measure it from gnd to 5v on each of the chips or should I select a single ground point - say near one of the corners - and then start probing the 5v pins from there outwards? 🤔

I'm puzzled why the voltage seemed the same for all chips when I just did it but I could well have done it wrong.
 
All capacitors that I removed have been checked and the out of spec ones replaced. No change.

I started to remove more of the chips to test them (and put sockets in) and I got a bit carried away and removed pretty much every chip (except some RAM that I had already removed and tested). At least this means I can use my cheapo IC tester to run through all of the logic chips.

As a result, so far, I have found a faulty 74LS148 (U193 on the schematics). This chip is part of the interrupt control for the 68010 and drives the /IPL0 to /IPL2 signals. Having read through some of the documentation for the 68K series it looks like if this chip sets low on all three /IPLx then this initiates a level 7 interrupt which is non-maskable - I assume non-ignorable - by the 68010. Since the system then looks for confirmation on what the peripheral was that asked for the interrupt, I am making a leap that no peripheral responds and so the processor stops at the interrupt. Hence why the board starts to make the beep pattern for 'Main CPU timed out'.

I need to hook this 74LS148 chip up to a breadboard and see what the resulting outputs are to check if I'm anywhere close with this theory.

I still have a lot of chips to test that aren't supported by the cheapo chip tester. These are predominantly 74LS244s, 245s, 373s and 374s with a couple of 795s too.
 
Latest update...

https://crashedfiesta.blogspot.com/2024/02/socket-to-em-torch-triple-x-part-5.html

In actual fact I'm now bouncing between two errors. There's the 'short long short short' error that means the video RAM is corrupted, and then there's the 'short long long long' error that means the 68010 CPU has timed out.. I think that the video RAM is just a dodgy connection in one of the many sockets I've installed and that the real roadblock is the CPU refusing to do anything.
 
You must log in or register to reply here.
Back
Top