LC III - Will it ever run?

[stephen_usher]

I replaced the ROM socket and cleaned the pins... no change.

Fitted a switch to the reset pads to make it easier to reset and then went through all the data lines on the HI ROM and noticed that D20 and D21 looked strange so I lifted the D21 pin and that looked nice and sharp. So I then tested what the system saw on D21 with the pin lifted... and there was a signal... D20's nice signal! Oh, another short!

Initially 2 ohms... Took a knife and scratched between D20 and D21 on the CPU and... 6 ohms. Again... 20 ohms. Again... no short. Powered the machine up and we have constant CPU activity, though it's crashing and restarting constantly. I'm going to have to check every pair of lines out of the CPU now for phantom shorts.

I've never seen this before, has anyone else?

[stephen_usher]

I now have an inconsistency.

The machine is constantly resetting after around 10uS... or a multiple of this:

Short crash:

IMG_3640.jpeg (272.91 KiB) Viewed 3481 times

Long crash:

IMG_3641.jpeg (294.75 KiB) Viewed 3481 times

This is all determined whether the last bit in this sequence on this data line is a short spike or a full "1". If it's a "1" then the bus floats high it's it's a spike then it's a quick reset.

None of the address or data lines have phantom shorts any more (currently).

I see activity on the RAM data lines and RAS/CAS but nothing else. The CPU Read/Write line is constantly high.

I think I'm going to have to get at least a new HI ROM as the some of the legs which were corroded are very fragile and feel as if they're about to drop off. I don't feel that I can trust that the chip is working correctly.

[stephen_usher]

Oh. Found iffy connections on the address bus pins on the CPU.

[rubber_jonnie]

Oh boy it's turning into a right game of hunt the short/bad connection.

Quite the project you've got going, but definitely seems like progress at least.

[stephen_usher]

Reflowed and multiply tested (and reflowed again) all the pins on that corner of the CPU and...

IMG_3642.jpeg (229.62 KiB) Viewed 3471 times

We have a winnah!

I wouldn't have done it without the microscope. I really detest surface mount stuff, especially old, reused surface mount chips. Not helped by pads which are floating off the PCB.

[JezC]

[stephen_usher]

Fitted a CR2032 socket for the NVRAM and left the reset switch and hot glued everything.

IMG_3645.jpeg (453.79 KiB) Viewed 3463 times

Booted off a ZIP drive. As expected 4MB RAM

IMG_3643.jpeg (253.6 KiB) Viewed 3463 times

No FPU so the benchmark is lower than the one Speedometer has for the LCIII.

IMG_3644.jpeg (355.14 KiB) Viewed 3463 times

It seems to be keeping the time and settings when powered off so it looks as though the battery is doing its thing. However, as the Bomarc schematics have seemingly omitted the battery entirely I can't tell if there's a proper connection or not.

[stephen_usher]

OK. The NVRAM isn't keeping its settings. I need to know where the battery is connected and run a bodge wire. Unfortunately the Bomarc schematics leave this information out.

[stephen_usher]

Oh. Just discovered that the "RESET" on the motherboard also resets the PRAM and I've been using that. So maybe it's working properly.

[rubber_jonnie]

Success for you too!!