Mega ST4 - no video output at all

[asapreta]

I've been thinking and I think we (or at least me), might have been misguided by the exact symptoms on the CAS signals.

Can you please confirm:

- You have activity for a short period of time in CAS0 (CAS1) after reset.
- CAS1 (CAS2) are constantly high.

Is that correct? If so, where are you probing the CAS signals? At the MMU chip itself, or at some other place like the LS244 buffer? If you didn't probe directly at the MMU chip pins, please do it.

In most schematics these signals are named CAS0 & CAS1, but in the Mega ones, at least in some of them, they are named as CAS1 & CAS2 for some reason.

Hi iJor!

if not mistaken the last time I probed the CAS1 signal was directly in MMU pins and was this reading below:

CAS in MMU

cas0.jpeg (147.19 KiB) Viewed 508 times

But, of course, can redo it.

And the lines, yes, they are high.

[asapreta]

I captured again the both CAS0 H/L and CAS1 H/L, as below:

WhatsApp Image 2024-10-25 at 18.06.15.jpeg (196.42 KiB) Viewed 496 times

WhatsApp Image 2024-10-25 at 18.06.15 (1).jpeg (142.07 KiB) Viewed 496 times

WhatsApp Image 2024-10-25 at 18.06.15 (2).jpeg (127.17 KiB) Viewed 496 times

WhatsApp Image 2024-10-25 at 18.06.15 (3).jpeg (140.92 KiB) Viewed 496 times

Also I tried to use the Logic Analyzer and I don't know if it makes sense what I captured, still learning:

CAS 0L and CAS1L

cas0-1.png (859.09 KiB) Viewed 496 times

Hope this helps about the CAS signal.
Will prepare myself to capture the other data previously discussed.

Thank you all!

[asapreta]

To begin with I'd look at /AS /DTACK and /HALT on the CPU plus A1, A2, A3 and /UDS and /LDS. This is just to see how many reads are happening when the CPU comes out of reset. The three address lines should be counting up as it reads the first four words of the ROM before starting to execute code.

Next I'd go to the MMU and sample /RAM, /DMA, /DEV, /RAS1, /CAS1H, /CAS1L, /LOAD and /CMPCS. This is to make sure that the lines from the GLUE and to the Shifter are doing their thing and to determine when CAS1* stop working. At some point the MMU should be telling the Shifter to read the bus and the chip also putting the address of the lowest word of video RAM on the bus when it does this.

Next I'd look at the data bus and look at the first four lower bytes that the CPU is reading from the ROM after a reset and then compare them to the first four bytes of the lower ROM. This may be able to tell you is the CPU is getting the wrong values due to an address or data line error.

You may want to try capturing the same signals on your working ST to compare.

Is this what you asked for, please?

7 PRIMEIROS.png (769.72 KiB) Viewed 494 times

7 PRIMEIROS_.png (1009.18 KiB) Viewed 494 times

[stephen_usher]

OK. The good news is that the CPU is trying to run code and is doing so for some time.

This means that using the diag cartridge is likely to be useful.

[asapreta]

OK. The good news is that the CPU is trying to run code and is doing so for some time.

This means that using the diag cartridge is likely to be useful.

The EEPROMs are on their way to be delivered today.
So then I can burn the diagROM and see what we got next.

[Badwolf]

The EEPROMs are on their way to be delivered today.
So then I can burn the diagROM and see what we got next.

Just a note to highlight DiagROM and the diagnostic ROM cartridge are not the same thing. DiagROM is a work-in-progress port from a bit of Amiga software and requires a fair degree of understanding of the system to use effectively.

The diagnostic ROM cart is more likely to identify specific problems in a human-readable way. I would consider this a more useful first step if you're able to repurpose a cartridge.

BW

[Badwolf]

BTW, whilst you're waiting and now we suspect RAM to actually be functioning to some degree, have you tried the finger test?

Power the machine on. Let it run for a minute or so and then put your finger on each RAM chip at a time, paying attention for any noticeably hotter than the others.

It's more common in the old RAM chips that had a 12V pin, but even 5V RAM can run really hot if it gives up the ghost.

BW

[asapreta]

The EEPROMs are on their way to be delivered today.
So then I can burn the diagROM and see what we got next.

Just a note to highlight DiagROM and the diagnostic ROM cartridge are not the same thing. DiagROM is a work-in-progress port from a bit of Amiga software and requires a fair degree of understanding of the system to use effectively.

The diagnostic ROM cart is more likely to identify specific problems in a human-readable way. I would consider this a more useful first step if you're able to repurpose a cartridge.

BW

Hi Badwolf!

Yes, sure. I got the difference but atm I can't get the cart right now.
I will make the PCB and have it handy.

The diagrom is more a another way to try to help the diagnostics while the cart isn't available.

[asapreta]

BTW, whilst you're waiting and now we suspect RAM to actually be functioning to some degree, have you tried the finger test?

Power the machine on. Let it run for a minute or so and then put your finger on each RAM chip at a time, paying attention for any noticeably hotter than the others.

It's more common in the old RAM chips that had a 12V pin, but even 5V RAM can run really hot if it gives up the ghost.

BW

Yes I did and probed the RAMs but couldn't see any differences in the signal.
BUT... I do think I have one or more bad RAM in the lower banks.

Can this machine run with only 1 or 2 Mb?
If so how can I disable the other RAM banks?

I don't have any spares of these RAM ICs here and will order some to arrive next week.

[stephen_usher]

Badwolf was suggesting testing the RAM chips' temperature using your finger to see if any are getting overly warm.

Answering your last question, it'll need one bank of 16 chips. I can't remember what the layout is for that board but I know it should run with half the RAM banks as that was how it would have been sold as a Mega2 ST (2MB).