exxos's Atari, Amiga & retro forum

Good hunting! Yeah, it surely won't boot with 1.4K on those two data lines, so keep looking!

Thanks :D
Good thing there are only CPU two lines (D6 and D7) which give the wrong values now, so I'm going to see where they go according to the schematic and check for continuity. So far it looks like one of them going to the Blitter chip is OK while the other is supposed to but isn't present, so at least I have a place to start.

Pacman wrote: 07 Apr 2021 21:04 Good thing there are only CPU two lines (D6 and D7) which give the wrong values now, so I'm going to see where they go according to the schematic and check for continuity. So far it looks like one of them going to the Blitter chip is OK while the other is supposed to but isn't present, so at least I have a place to start.

Is the blitter socketed? Probably a good idea to remove anything socketed, to remove internal IC problems as a potential culprit.

It boots! :D
After painstakingly going through the schematic to figure out where the CPU D6 and D7 lines go I discovered that the pullup resistance suddenly showed the correct values (10K) after removing the 68901 MFP (U306).

The strange thing is that it appeared to be plugged in the wrong way round, but that should be impossible! I should have taken notes of its position while I was struggling to find its pinout/pin numbers as it's one of those PLCC ICs that are a bit unclear on where pin 1 is.
Anyway, I put it back in and the resistance stayed the same, so I powered up and the computer booted! With a few bombs mind you, but I believe that's only to be expected when neither the floppy disk drive or the internal hard drive interface is attached. Ruling out an incorrrect placement of the IC (it would probably short circuit it as well) it was probably a matter of bent PLCC socket pins and/or dirty pins.

Getting back to the original subject, I inserted the diagnostic cartridge and that worked too. I'll look closer into all those tests when everything has been put back in (Blitter (yes, it's fortunately socketed), disk drive etc.) but thought I'd see if I could get the STe to display its results via the nullmodem cable. Alas I couldn't get that to work. I've tried all 3 serial ports from the Mega STe, but nothing pops up in the VT52 emulator ACCessory in the STe. Have I missed something?
The VT52 default settings look OK (9600 baud, 8N1). I could of course have issues with the modem port(s) on either or both machines...

Another thing: how do you get back to the main diagnostic menu page when a test has been completed? I tried all sorts of key-presses to no avail. In the end I found out that I could press the reset button to reboot the cartridge, but that's not the correct way to do it, is it?

Pacman wrote: 08 Apr 2021 10:42 It boots! :D

:o :girldance:

The strange thing is that it appeared to be plugged in the wrong way round, but that should be impossible!

Right... Impossible... :mrgreen:

The VT52 default settings look OK (9600 baud, 8N1). I could of course have issues with the modem port(s) on either or both machines...

Another thing: how do you get back to the main diagnostic menu page when a test has been completed? I tried all sorts of key-presses to no avail. In the end I found out that I could press the reset button to reboot the cartridge, but that's not the correct way to do it, is it?

Your settings are fine, but as you say, any number of other things could be going on. At least you've got the machine booting now, which should make it easier to diagnose the other stuff. (You could, for instance, run the VT52 acc on the MSTE and see if you can produce output on the STE.)

Spacebar should take you back to the main menu. Are you sure the keyboard is working?

Thanks :D
And for confirming about the MFP orientation. It must have just been bad contact somehow.
EDIT: I thought you were kidding until I actually read the posting you linked to!
So it really is possible to put PLCC ICs the wrong way round. That's what must have happened.

Off my head, doesn't the MFP control the serial ports? Is so, it might make sense that the IC is blown and could be the reason why the serial connection to the STe isn't working. Fortunately Exxos has 68901 replacements available at the webstore if I need a new one. But it would be nice to confirm its status first.

Good idea to test with the cartridge the other way round. Are there other (simple) ways to test a nullmodem configuration between two STs? Could I for instance run the VT52 accessory on both computers, type something on say the Mega STe and have it display on the STe? A test like that would be very useful.

Spacebar should take you back to the main menu. Are you sure the keyboard is working?

Ah! I suspected that.
No, there are some known contact issues with the spacebar. I'll have to look into that.

Pacman wrote: 08 Apr 2021 10:56 Off my head, doesn't the MFP control the serial ports? Is so, it might make sense that the IC is blown and could be the reason why the serial connection to the STe isn't working.

Yup, that is correct. Luckily they are cheap and easy to replace.

Could I for instance run the VT52 accessory on both computers, type something on say the Mega STe and have it display on the STe? A test like that would be very useful.

Actually that is what I was suggesting, although I now see that wasn't entirely clear. :? If you run the VT52 acc on both machines, with identical port settings, you should be able to type on one machine and see it on the other. (And if you turn it to half duplex mode, you'll see it on both.) If you can't get this simple setup to work, I agree that the MFP is a likely culprit, given its recent unfortunate experiences.

derkom wrote: 08 Apr 2021 11:18 Actually that is what I was suggesting, although I now see that wasn't entirely clear. :? If you run the VT52 acc on both machines, with identical port settings, you should be able to type on one machine and see it on the other. (And if you turn it to half duplex mode, you'll see it on both.) If you can't get this simple setup to work, I agree that the MFP is a likely culprit, given its recent unfortunate experiences.

Ah! Nice and simple. Reminds me of UNIX "talk" back in the 90s :)
I'll give that a go once I've put the floppy drive back in.

Given the fact that I don't know if neither machine's serial ports work (and I don't have a serial-USB adapter for my Mac), is there a simple way to check if the Mega STe 68901 works or not? I ran the diag cart "serial" test but it failed as it obviously relies on a special "loopback" cable.
I had a quick glance in the Mega STe service manual but so far haven't found out what this is. Anyone know?

I've just put the blitter chips back into both machines (fearing that one or the other might be damaged while testing) but they both work (phew!).
Good thing, as those are harder to come by as far as I know :o
And I suspected a failed blitter in the Mega STe since I couldn't get the MonSTer board to work with it (reliably at least), but it must have been some of those other more trivial issues I guess.

To make things even worse I accidently put one of the blitter chips in the Mega STe CPU socket and powered it up! Checking the pinouts I saw that the 68000 voltage (+5V) pin went to one of the blitter's data pins, but apparetly that didn't damage it :shock:
The diag cart blitter tests (short and long) both told me the MSTE blitter was OK, so that should confirm it. On the STe I'm able to run the blitter on/off from the TOS 2.06 desktop menu -would that confirm it working, or could it be partly damaged?
I didn't bother re-configuring the diag cart for the STe, then back again for the MSTE, but I suppose a thorough test with it on both machines would be useful.

UPDATE: I don't know if this is because I've re-inserted the MonSTer board, but was surprised that the computer wouldn't even boot with the MFP (68901) removed! I was under the impression that the MFP only took care of the serial/parallel ports, but maybe not?
Or could it be that the MonSTer relies on it? I have struggled with the MonSTer for a long time -either the Blitter hasn't worked with it installed, and recently would crash when running stuff from the attached CF memory card (the IDE-CF adapter and memory card works fine with the MonSTer in my STe).
Then I got the crazy idea to try the secondary IDE port (and swap primary/secondary IDE with the on-board jumper), and it worked! I'm going to look closer at the MonSTer board concerning this as well as the missing Blitter (it works fine without the MonSTer), but would like to hear if someone knows about the MFP's role in all of this.
Even stranger is if the MFP is the cause of my faulty serial ports, but still prevents the computer from booting! Is it possible to have a half-working MFP?

derkom wrote: 08 Apr 2021 11:18

Pacman wrote: 08 Apr 2021 10:56 Off my head, doesn't the MFP control the serial ports? Is so, it might make sense that the IC is blown and could be the reason why the serial connection to the STe isn't working.

Yup, that is correct. Luckily they are cheap and easy to replace.

Which of the three serial ports does the 68901 MFP control? The service manual is a little unclear on the naming of the ports, and so is the schematic.
I suspect that "Modem 2" and "Serial 2" isn't working (more below).
"Modem 1" is confirmed working with the diagnostic cartridge and communications with the STe using the VT52 terminal and ConNect.

I got the pinouts for the 9 pin D-sub loopback plug in another thread, so I soldered up one, and it passes the diagnostic cart test, but only for "Modem 1". At this stage I don't know if the errors I'm getting with the other two ports are due to not having a differently wired loopback connector (this sounds strange, but nothing would surprise me) or if the ports are actually damaged.
I haven't yet come across anyone who has the pinouts for the other two loopback connectors (or confirming that they're the same as of the "Modem 1" loopback connector).

But there is definitely something strange going on with "Modem 2" and "Serial 2". When using ConNect and selecting either of them it looks like the cursor is flashing much faster (like when you repeat a character), but nothing shows up on the other computer's screen. With "Modem 1" selected the cursor flashes normally and when I type something on the MSTE it shows up on the STe screen (also running ConNect).

Could I for instance run the VT52 accessory on both computers, type something on say the Mega STe and have it display on the STe? A test like that would be very useful.

Actually that is what I was suggesting, although I now see that wasn't entirely clear. :? If you run the VT52 acc on both machines, with identical port settings, you should be able to type on one machine and see it on the other. (And if you turn it to half duplex mode, you'll see it on both.) If you can't get this simple setup to work, I agree that the MFP is a likely culprit, given its recent unfortunate experiences.

Once I re-installed the RSVE board inside the STe it worked right away! So now at least I know the STe's modem port is OK, as well as the MSTE "Modem 1" port.

UPDATE:: looking closer at the schematic it appears that "Serial port" is attached to U306 (MC68901) while "Serial port A" and "Serial port B" goes to UA04 (Z85C30).

Pacman wrote: 08 May 2021 15:25
UPDATE:: looking closer at the schematic it appears that "Serial port" is attached to U306 (MC68901) while "Serial port A" and "Serial port B" goes to UA04 (Z85C30).

The Z85C30 is the lan port and to my knowledge the only thing that accesses it is the Atari test cart with the loop back plug.

The attached document has a lot of info on the Atari in general and may answer some of your questions. It's a pdf file.