No clock signal to the Shifter

[russellnash]

You shouldn't be worried about offending me, there is no such thing as causing it, only taking it.
I ask questions because I would like an answer and answers that are relevant and truthful are the best answers. I already know more now from your comments than I did before.
There is probably a good reason that you use an oscilloscope, otherwise you wouldn't use one. It is nice to have all the right stuff that you need. The reality is that people tend do all sorts of things without all the necessary tools and equipment. That is true from cooking in the kitchen to mixing concrete.
Thanks.

[russellnash]

After I replaced L22, R86 and R91, which are the three components between Y2 and Y3 on the photo, the input and output of the Shifter both oscillate on the logic probe, at the MMU, GLUE and CPU etc.

[russellnash]

With respect to the possibility of using either 2 or 6 chip TOS on this type of board I noticed something that got my attention. On the original 520ST schematic pin 22 OE of the TOS Roms is connected to UDS and LDS and also to the cartridge port. However, on the schematic for the 1040ST the TOS Roms are connected to ground with pin 22 via the A17 jumper. Is that correct that to run such a board with 6 chips they must no longer be connected to LDS and UDS?

[rubber_jonnie]

With respect to the possibility of using either 2 or 6 chip TOS on this type of board I noticed something that got my attention. On the original 520ST schematic pin 22 OE of the TOS Roms is connected to UDS and LDS and also to the cartridge port. However, on the schematic for the 1040ST the TOS Roms are connected to ground with pin 22 via the A17 jumper. Is that correct that to run such a board with 6 chips they must no longer be connected to LDS and UDS?

I've converted several 2 chip boards to 6 chip and all I ever did was add the decoupling caps and change the jumper pads that set the ROM size.

[russellnash]

Thanks for confirming that.

[russellnash]

Anybody know what makes IPL2 always low? Thinking that might be significant. I have got the board so that all data lines D0 to D15 are oscillating but all the address lines are stuck high. One thing that is noticeable is that IPL2 stays low from the second that I switch on and also by reset. The pull up for IPL2, pin 4 on RP4 measures 4.7K. All pull ups pass the resistance test.
The CPU, GLUE and MMU are confirmed working and the PLCC socket contacts pass a continuity test that I did using a pin to actually touch the chip contacts rather than just the socket contacts.

[russellnash]

It boots.

20240419_172350 (copy 1).png (823.4 KiB) Viewed 8228 times

What did I do? First of all I assembled the board completely, so that it had the YM2149, WD1772 and two logic chips U40 and U8 that were missing back in place.
That didn’t have any effect. I found posts about IPL2 and RDY on the GLUE being low, but none of them got an answer. I suppose nobody knows the answer. As far as I can tell it is a symptom of the GLUE reacting to some other, unknown problem.
The trace on RAS0, pin 8 on the MMU, was corroded through so I repaired it with solder and a bit of wire. No boot.
I tested the resistance on RP2 and got open circuit on every pin. I thought that was weird and that I would replace it when I eventually have the parts.
The CPU is not in a socket but in strips of round pins. The contact is not very reliable. I decided to solder all the address pins in place. I powered it up and everything that was stuck high or low was oscillating, so I connected up the monitor and got the GEM desktop. The picture quality seems very good compared to most ST's I have seen.
It was a lot of work but it was worth it to rescue a largely empty board previously used for parts.

[olivier.jan]

[rubber_jonnie]

Good stuff