Mega ST4 - no video output at all

[asapreta]

As the equipment are the same, probably in the weekend I will swap these ICs. I think DMA and 68901 are soldered in 520ST.
I was very skeptical about those CAS signals, they really look like noise as you said.

Do a bad RAM (or more than one) can "lock" the CAS signal high?

Thank you.

Definitely not at the MMU end as there's a 74LS244 between the MMU and the RAM. Have you tested/replaced that 244. It's highly unlikely that it's that which is the problem as chip inputs hardly ever go short but it's something you can look at. 244s do go bad but usually they just don't output anything on one or more pins when the input changes. Having said that, with the line high it's unlikely to be a TTL chip as they have very weak pull-up ability and the MMU would have no trouble overpowering that.

Yesterday, I have removed and tested the four LS244 and the two LS373. All Ok. I didn't replace by new but used two different IC testers.

With a multimeter could you check what the resistance is between the CAS pin ON THE CHIP, not the socket, and the CAS input pin on the 244. It should be tiny if the MMU socket is OK.

Will do as soon as I get home.

Checking the resistance between the CAS line and VCC would be useful as well, in case there's a short somewhere.

Can I do that in the DRAM pins?

Again, thanks for your knowledge and support.

[stephen_usher]

Can I do that in the DRAM pins?

No, because they're not connected to the MMU, only the 244s are connected to the MMU. The 244s then relay the signals to the RAM.

Please ignore the RAM entirely as they're beyond the location of the fault, the other side of the bus transceivers/line drivers. i.e. they're on a completely separate circuit.

[ijor]

So that CAS scope was just noise ...

I would recommend going back to what I suggested above. In first place post a good picture of the MMU socket (without the chip, of course). I'm suspecting is MMU related because even with a RAM fault, there should still be CAS pulses for the video cycles.

And as I asked, please try to measure the approximate time from end of reset to the CPU being halted. And also for how long you have CAS pulses after reset. Or again, an rough estimate of how many AS pulses are generated before the CPU is halted.

With a multimeter could you check what the resistance is between the CAS pin ON THE CHIP, not the socket, and the CAS input pin on the 244. It should be tiny if the MMU socket is OK.
Checking the resistance between the CAS line and VCC would be useful as well, in case there's a short somewhere.

Stephen, unless I misunderstood, note that CAS is working fine for a small period after Reset. That should indicate that there is no electrical problem directly related to CAS. And that's why I suspect it is something related to MMU.

[ijor]

I would recommend going back to what I suggested above. In first place post a good picture of the MMU socket (without the chip, of course). I'm suspecting is MMU related because even with a RAM fault, there should still be CAS pulses for the video cycles.

Just in case, can you please scope DE and LOAD. They are both available at SHIFTER. Note that DE is a slow signal, it pulses (or at least, it should) once per horizontal line.

EDIT: I forgot. You never told us if chip select is working on the ROM chips between Reset and HALT. Most likely yes, but would also be better to confirm. Yes, chip select and CE is the same thing.

[asapreta]

I would recommend going back to what I suggested above. In first place post a good picture of the MMU socket (without the chip, of course). I'm suspecting is MMU related because even with a RAM fault, there should still be CAS pulses for the video cycles.

Just in case, can you please scope DE and LOAD. They are both available at SHIFTER. Note that DE is a slow signal, it pulses (or at least, it should) once per horizontal line.

EDIT: I forgot. You never told us if chip select is working on the ROM chips between Reset and HALT. Most likely yes, but would also be better to confirm. Yes, chip select and CE is the same thing.

DE, from Shifter:

WhatsApp Image 2024-10-23 at 18.26.50.jpeg

LOAD from Shifter:

WhatsApp Image 2024-10-23 at 18.26.54.jpeg

This is what I could get when power on or reset the computer in ROM PIN 20, /CE:

WhatsApp Image 2024-10-23 at 18.06.22.jpeg

I had to film and export the frame.

[asapreta]

As the equipment are the same, probably in the weekend I will swap these ICs. I think DMA and 68901 are soldered in 520ST.
I was very skeptical about those CAS signals, they really look like noise as you said.

Do a bad RAM (or more than one) can "lock" the CAS signal high?

Thank you.

Definitely not at the MMU end as there's a 74LS244 between the MMU and the RAM. Have you tested/replaced that 244. It's highly unlikely that it's that which is the problem as chip inputs hardly ever go short but it's something you can look at. 244s do go bad but usually they just don't output anything on one or more pins when the input changes. Having said that, with the line high it's unlikely to be a TTL chip as they have very weak pull-up ability and the MMU would have no trouble overpowering that.

With a multimeter could you check what the resistance is between the CAS pin ON THE CHIP, not the socket, and the CAS input pin on the 244. It should be tiny if the MMU socket is OK.

Checking the resistance between the CAS line and VCC would be useful as well, in case there's a short somewhere.

@stephen_usher

The resistance of both CAS signals are around 0,6 and 1,0 ohm from the MMU IC pins (which I had to use a wirewrap attached to the multimeter probe to make sure I wasn't touching the socket) to the LS244 pins 17 and 15.

From CAS to VCC, both are infinite, open, no shorts.

[asapreta]

So that CAS scope was just noise ...

I would recommend going back to what I suggested above. In first place post a good picture of the MMU socket (without the chip, of course). I'm suspecting is MMU related because even with a RAM fault, there should still be CAS pulses for the video cycles.

And as I asked, please try to measure the approximate time from end of reset to the CPU being halted. And also for how long you have CAS pulses after reset. Or again, an rough estimate of how many AS pulses are generated before the CPU is halted.

I will take the picture.
I need to put a wire on HALT, CAS and AS as I need to film this to get the frame where something is visible.

I will come back with this information.

[asapreta]

So that CAS scope was just noise ...

I would recommend going back to what I suggested above. In first place post a good picture of the MMU socket (without the chip, of course). I'm suspecting is MMU related because even with a RAM fault, there should still be CAS pulses for the video cycles.

And as I asked, please try to measure the approximate time from end of reset to the CPU being halted. And also for how long you have CAS pulses after reset. Or again, an rough estimate of how many AS pulses are generated before the CPU is halted.

With a multimeter could you check what the resistance is between the CAS pin ON THE CHIP, not the socket, and the CAS input pin on the 244. It should be tiny if the MMU socket is OK.
Checking the resistance between the CAS line and VCC would be useful as well, in case there's a short somewhere.

Stephen, unless I misunderstood, note that CAS is working fine for a small period after Reset. That should indicate that there is no electrical problem directly related to CAS. And that's why I suspect it is something related to MMU.

@ijor

/HALT:

WhatsApp Image 2024-10-23 at 20.10.54.jpeg

/AS:

WhatsApp Image 2024-10-23 at 20.12.00.jpeg

/CAS0:

WhatsApp Image 2024-10-23 at 20.21.58.jpeg

And pictures of the MMU Socket:

WhatsApp Image 2024-10-23 at 20.27.50.jpeg

WhatsApp Image 2024-10-23 at 20.28.26.jpeg

WhatsApp Image 2024-10-23 at 20.28.54.jpeg

As far as I know I am at least the third owner of this ST.
The previous owner said, as in the first post, sometimes when it had blank screen, he "gently slapped" the case and it would work again.

[ijor]

DE, from Shifter:
LOAD from Shifter:

This confirms, as I said, that there are still video cycles, only without CAS, and that is weird , if we consider that CAS works after Reset.

/AS:
/CAS0:

Unfortunately this is not very useful. I suspect we are not getting the full picture. What happens is that right after starting TOS executes a "soft" Reset. This is a Reset initiated by the CPU. It still drives the Reset signal and resets the hardware, that's the goal. The point is that during that time there won't be any AS or CAS pulses. But probably there were after that "soft" reset. It would be very unlikely that the CPU got halted at that point.

/HALT:

That could be useful. But could you estimate the time that HALT was high? Does the analog scope provide any kind of timing scale?

As far as I know I am at least the third owner of this ST. The previous owner said, as in the first post, sometimes when it had blank screen, he "gently slapped" the case and it would work again.

Precisely, this is usually an issue with these sockets ...

[asapreta]

/HALT:

That could be useful. But could you estimate the time that HALT was high? Does the analog scope provide any kind of timing scale?

I forgot to note it down.
I will redo the picture tô get the timming scale.

As far as I know I am at least the third owner of this ST. The previous owner said, as in the first post, sometimes when it had blank screen, he "gently slapped" the case and it would work again.

Precisely, this is usually an issue with these sockets ...

But do you see anything wrong with the socket?