Brett's STE - Repair thread

[stephen_usher]

It may be useful to 'scope the data lines on the ROM for activity and if you have a two input oscilloscope look at the ROM data line and same data line on the CPU socket for comparison. (i.e. triggering on the ROM data line and showing both on the screen) and do this for each of the data lines, just to make sure that there are no crossed wire... if there's any data activity at all, of course.

[rubber_jonnie]

It may be useful to 'scope the data lines on the ROM for activity and if you have a two input oscilloscope look at the ROM data line and same data line on the CPU socket for comparison. (i.e. triggering on the ROM data line and showing both on the screen) and do this for each of the data lines, just to make sure that there are no crossed wire... if there's any data activity at all, of course.

Yep, something I can do, and a good idea.

[rubber_jonnie]

So I quickly hooked up my logic analyser to my own good working STE with the original TOS 1.62 ROMs from Bretts machine which are known good.

The LA is connected to the address lines at the ROMs.

This is what I see during startup:

Screenshot 2022-01-25 at 09.59.16.png

So a flat line until power comes on then all lines go to 5v, and we see addresses on the bus and the machine boots.

In contrast, this is Bretts machine, with the same ROMs installed:

Screenshot 2022-01-25 at 10.00.12.png

So everything goes to 5v as expected, but then we don't see any proper address activity on the bus at all, and I am confident that even if I had made some wiring mistakes, or not connected something, I should at least see some bus activity.

Since I have buzzed all the address lines to their next hops on the mainboard, I am confident they should work, and they are all connected socket to socket as per the schematic, I don't believe they can be that far out. I haven't found any shorts either.

Since I know the following all work as tested in my STE:

CPU
RAM
Shifter

I know they aren't the cause. There is a separate Blitter which I can't test in my STE, so I need to find another way to do so.

So I need to start tracking across the address bus and see what's going on. My concern is a problem with the GST MMU/GLUE combo, but more work required.

[rubber_jonnie]

@BrettRogersUK I seem to recall you had a PSU problem on your STE, and it was emitting sparks or something, did it ever fail like this when fitted to the STE?

[stephen_usher]

Which of these are on the /OE and /CE lines?

[rubber_jonnie]

Which of these are on the /OE and /CE lines?

None of them, just the address lines, A1-A15.

I'd expect them to match my working machine.

/CE is pulsing identically to my working machine according to my scope. EDIT - I just checked and /CE has stopped pulsing :(

/OE is still pulsing on reset.

[exxos]

Whats on chan 12 ?

Can you post, RESET, BERR,HALT, AS, DTACK,

It looks like the CPU is not running.

[BrettRogersUK]

@BrettRogersUK I seem to recall you had a PSU problem on your STE, and it was emitting sparks or something, did it ever fail like this when fitted to the STE?

Not that I can recall. I did start hearing a few crackles when it was connected but instantly turned it off. It wasn't until I took it out of the case that it went bang. That was after I recapped it as I had to replace the transistor also found that it had a blown a resistor on the PS board completely in half. It was one next to the optocoupler. The fuse blew almost instantly.

[exxos]

Not that I can recall. I did start hearing a few crackles when it was connected but instantly turned it off. It wasn't until I took it out of the case that it went bang. That was after I recapped it as I had to replace the transistor also found that it had a blown a resistor on the PS board completely in half. It was one next to the optocoupler. The fuse blew almost instantly.

Might have been the switch as they are pretty terrible ? They spark and crackle etc that may well have caused the main transistor to fail. Though primary side failures don't generally cause problems on the low-voltage side. The primary has to be switching to generate voltage in the transformer core generates the low side voltages. So when something fails on the primary side, it's stops switching and the low side voltages will quickly drop to 0.

[BrettRogersUK]

Might have been the switch as they are pretty terrible ? They spark and crackle etc that may well have caused the main transistor to fail. Though primary side failures don't generally cause problems on the low-voltage side. The primary has to be switching to generate voltage in the transformer core generates the low side voltages. So when something fails on the primary side, it's stops switching and the low side voltages will quickly drop to 0.

That was when I decided to buy one of your power supplies. I couldn't remember what value resistor was next to the optocoupler and if that resistor had blown there was a good chance the optocoupler had gone too so made the decision to replace the PS with one of your own power supplies.the booting problem started before this though after I put the ROM switcher in.