*** M E M O R U M ***

I can't seem to find much wrong on the reset circuit other than the 4070B getting hot, so I'm investigating everything that's connected to it.

I've currently removed all the socketed chips (see photo below) and I'm getting 303 ohms between the 0V and +5V rails, which seems rather low to me. If I install the TMS graphics chip then this decreases by 70 ohms. The trouble is that I don't know what the correct value should be. Does 300 ohms seem low to you?

Without either doing the same to one of mine, or working out an approximation from the schematic, it’s hard to say for certain.

However, turn that figure around using ohms law:

5V / 300 ohms = 16.67mA

This sounds reasonable to me.

If you have a multimeter that has a current range, put it in series (on the high current range, 10A to 2A range to start with) with the fuse. Then power it up and see what the meter says. Power down again, then select a more appropriate current range. Power up again and record the actual current.

Note, we start on the high current range just in case there is still a short circuit.

By ‘in series’, I mean, pull the fuse out of one fuse clip. Then bridge the gap between the fuse cap that’s now pointing into the air and the empty fuse clips with the meter and it’s leads. It’s a lot easier if you have croc clips

Mark

I've done that before with the MTX512... And discovered that the reason it was blowing fuses was that I'd bought fuses which were a tenth of the rating! Doh!

It's pulling about half an amp. Given that my MTX512 when fully operational pulled 0.85 amps that doesn't sound too much out of spec.

So, given that the only chip getting hot is the 4070B that may be the first target for replacement after I've put the sockets in the LS138 locations.

Of course, the 4070B is one of the very few chips I didn't get replacements for earlier in the week.

Now, with the LS138s replaced and 9C (4070B) out of the circuit I'm back to the black screen and beep. There's activity on the address bus too.

Checking out the voltages going to 9C I found that the output from 10E is actually negative, -0.8V! Surely that can't be right? I've noticed that the -12V line is actually running at -14V somehow, at least to the input of 10E.

I can't work out what the circuit including 10E is actually doing as I can't find any other references to its output, C/T3.

C/T3 is used for timing on the cassette data input.

But you don't need it for testing, you can remove the CTC and the system will run, albeit that no timed functions will work.
Serial data clock, cursor flash, keyboard repeat, clock etc will all be U/S, but the system should run

Should 6B pin 8 be mostly high but with lots of spikes down to about 2V? I'm guessing not. Having replaced 6B with no change, maybe 5B's faulty?

Why am I looking here? The ROM select lines don't seem to be changing and are always high.

Dave wrote: ↑06 May 2019 17:47 C/T3 is used for timing on the cassette data input.

But you don't need it for testing, you can remove the CTC and the system will run, albeit that no timed functions will work.
Serial data clock, cursor flash, keyboard repeat, clock etc will all be U/S, but the system should run

I tried that but the CPU didn't start, probably because the reset circuit is incomplete at the moment (4070B removed). The CTC seems to be giving enough of a nudge to start the CPU on power-up.

mmm, not sure that makes sense, but I think you're going to struggle to get defined behaviour from the CPU without a guaranteed reset at startup.

Try removing the CTC and manually grounding reset after you power on, through a resistor - say 100R and see if it starts

I probed the same pin on the MTX here, very similar looking waveform.
500ns between peaks, I wonder if I've captured an opcode fetch. So spike when the opcode read starts, then another spike before the refresh starts. Then the same with the next opcode.

The bit before the trigger could be a longer opcode.