Diagnostic cartridge (dead Mega STe)

Problems with your machine in general.
User avatar
Pacman
Posts: 181
Joined: Mon Aug 28, 2017 12:56 am
Location: Norway

Re: Diagnostic cartridge (dead Mega STe)

Post by Pacman »

exxos wrote: Tue Mar 30, 2021 2:11 am the MSTE boosts to 16mhz which means other cpus won't likely work. But if the CPU works in a STE then its unlikely to be damaged. Even so, the CPU in the store may help, but assume not.

Have you checked the bus resistors? In that you get 10k pull up on most CPU pins?
No, I haven't checked that.
Are these the two resistor arrays on page 2 of the Mega STe schematic that go to "Data <15:0>"?
I'm guessing there are others too, connected the same way, hidden somewhere in the schematic as there weren't any surrounding the CPU itself (page 1).

Screen Shot 2021-03-30 at 12.47.56.png
Screen Shot 2021-03-30 at 12.47.56.png (59.62 KiB) Viewed 3518 times

So I check the resistance between the CPU PLCC-socket pins and +12V to see if there a 10K resistance?
Should I remove the CPU from the socket first, before measuring?
I'm also going to check for continuity from each CPU PLCC pin and to where the signals are supposed to go elsewhere. A cold solder joint or fracture isn't unlikely in this area since I've messed around with the MonSTer board a lot.
The inside pins (going to the CPU itself) have all been bent inwards to ensure a good connection. I've also checked for any damaged or bent pins on the CPUs I've inserted.

EDIT: I've done a quick check using the method I guessed above (checking for resistance between +12V on the PSU (of course with power turned off!) and the various CPU pins (but I didn't remove the CPU from the socket), and indeed: several of the pins read around 10K. But there were other lines reading 2K and 5K as well as completely different values. You said "most" CPU pins, so I assume this may be right, but when I have some more time on my hands I'll take a more thorough check, noting which pins have which value (and maybe I should have removed the CPU first as well).
STe | MonSTer with dual IDE-CF memory card adapter | NEC Multisync 1990SXi | ST_ESSC | RSVE | Link '97 | Sony HD floppy drive/AJAX controller | Exxos HD floppy module | Battery-backed Ricoh realtime clock module | Discovery cartridge | C-Lab Unitor-2 | C-Lab Export | C-Lab Combiner | C-Lab Steady Eye | C-Lab Human Touch | Unicorn USB

Are you a good person?
User avatar
frank.lukas
Posts: 660
Joined: Fri Jan 19, 2018 11:52 am

Re: Diagnostic cartridge (dead Mega STe)

Post by frank.lukas »

Between VCC (+5V) and the addresss (A1-A23) lines = 4,7K and between the data (D0-D15) lines = 10k
User avatar
Pacman
Posts: 181
Joined: Mon Aug 28, 2017 12:56 am
Location: Norway

Re: Diagnostic cartridge (dead Mega STe)

Post by Pacman »

Ah! Thanks. I will check those.
Would I need to remove the 68000 from its socket to get the correct readings?
STe | MonSTer with dual IDE-CF memory card adapter | NEC Multisync 1990SXi | ST_ESSC | RSVE | Link '97 | Sony HD floppy drive/AJAX controller | Exxos HD floppy module | Battery-backed Ricoh realtime clock module | Discovery cartridge | C-Lab Unitor-2 | C-Lab Export | C-Lab Combiner | C-Lab Steady Eye | C-Lab Human Touch | Unicorn USB

Are you a good person?
User avatar
frank.lukas
Posts: 660
Joined: Fri Jan 19, 2018 11:52 am

Re: Diagnostic cartridge (dead Mega STe)

Post by frank.lukas »

no
User avatar
Pacman
Posts: 181
Joined: Mon Aug 28, 2017 12:56 am
Location: Norway

Re: Diagnostic cartridge (dead Mega STe)

Post by Pacman »

Ok, thanks.
STe | MonSTer with dual IDE-CF memory card adapter | NEC Multisync 1990SXi | ST_ESSC | RSVE | Link '97 | Sony HD floppy drive/AJAX controller | Exxos HD floppy module | Battery-backed Ricoh realtime clock module | Discovery cartridge | C-Lab Unitor-2 | C-Lab Export | C-Lab Combiner | C-Lab Steady Eye | C-Lab Human Touch | Unicorn USB

Are you a good person?
User avatar
Pacman
Posts: 181
Joined: Mon Aug 28, 2017 12:56 am
Location: Norway

Re: Diagnostic cartridge (dead Mega STe)

Post by Pacman »

frank.lukas wrote: Tue Mar 30, 2021 1:04 pm Between VCC (+5V) and the addresss (A1-A23) lines = 4,7K and between the data (D0-D15) lines = 10k
I found something!
All the lines above check out OK except for four:
  • 68000 pin 32 (A1): 2.3K (instead of 4.7K)
  • 68000 pin 42 (A11): 2.3K (instead of 4.7K)
  • 68000 pin 66 (D7): 1.4K (instead of 10K)
  • 68000 pin 67 (D6): 1.4K (instead of 10K)
I thought I'd trace each signal line and check for broken traces etc. but I'm a little confused about the schematic. Here's the CPU with the abovementioned lines coloured in red:

CPU_pullups_wrong.png
CPU_pullups_wrong.png (390.49 KiB) Viewed 3381 times


I see that there's a "DATA" and "ADDRESS" line further down in the schematic, which I assume is for simplifying things/making the schematic more tidy, but what are the numbers on the outer right and left sides of the CPU?
STe | MonSTer with dual IDE-CF memory card adapter | NEC Multisync 1990SXi | ST_ESSC | RSVE | Link '97 | Sony HD floppy drive/AJAX controller | Exxos HD floppy module | Battery-backed Ricoh realtime clock module | Discovery cartridge | C-Lab Unitor-2 | C-Lab Export | C-Lab Combiner | C-Lab Steady Eye | C-Lab Human Touch | Unicorn USB

Are you a good person?
User avatar
derkom
Moderator
Moderator
Posts: 1208
Joined: Sun Jul 29, 2018 6:45 pm

Re: Diagnostic cartridge (dead Mega STe)

Post by derkom »

Pacman wrote: Wed Apr 07, 2021 10:02 am I see that there's a "DATA" and "ADDRESS" line further down in the schematic, which I assume is for simplifying things/making the schematic more tidy, but what are the numbers on the outer right and left sides of the CPU?
I'm not entirely sure I follow. The numbers just outside the IC on the schematic are the pin numbers, but you obviously already know that, since you've been referring to them. The A/D numbers on the inside are the address and data bus numbers, and the third set of numbers are the schematic bus numbers, to associate individual connections with the breakout on the other end of the bus on the schematic drawing. Conveniently, they've made these match the actual bus numbers.

As for your pullup readings, there's certainly something wrong with those, if the readings are correct. Double and triple check those, as it's easy to get the wrong pin on a PLCC. Since these are SIP resistors, it's a little odd that you'd have most of the pins working fine, and then just a couple with far too little resistance, but I suppose anything is possible.
User avatar
Pacman
Posts: 181
Joined: Mon Aug 28, 2017 12:56 am
Location: Norway

Re: Diagnostic cartridge (dead Mega STe)

Post by Pacman »

derkom wrote: Wed Apr 07, 2021 10:12 am I'm not entirely sure I follow. The numbers just outside the IC on the schematic are the pin numbers, but you obviously already know that, since you've been referring to them. The A/D numbers on the inside are the address and data bus numbers, and the third set of numbers are the schematic bus numbers, to associate individual connections with the breakout on the other end of the bus on the schematic drawing. Conveniently, they've made these match the actual bus numbers.
Thanks for explaining. Makes sense.
So with that info in mind I located the SIP resistor arrays and confirmed their connections to the lines with the wrong resistance.
First the 10K resistor (RP26):

CPU pin 66 (D7) to RP26 (10K x8) pin 9
CPU pin 67 (D6) to RP26 (10K x8) pin 8
RP26_10Kpullup.png
RP26_10Kpullup.png (57.92 KiB) Viewed 3365 times

I've measured the resistance of RP26 directly, between pin 1 (common) and pin 8, likewise between pin 1 (common) and pin 9 and can confirm their lower resistance than its other pins (2-7). I assume I would have to desolder RP26 to confirm that its faulty or not.
But before doing that I assume I need to figure out where else the abovementioned two CPU pins go to and see if there's something obvious there (a short circuit perhaps?).

Regarding the two other CPU lower resistance measurements I've located RP23 and RP24 (both 4.7K x8) and confirmed their connections to the CPU:

CPU pin 32 (A1) to RP24 (4.7K x8) pin 9
CPU pin 42 (A11) to RP23 (4.7K x8) pin 9
RP23_24_4K7pullup.png
RP23_24_4K7pullup.png (100.61 KiB) Viewed 3365 times

In the same way as with the 10K resistor array (RP26) I measured directly and both of those 4.7K resistor arrays show a lower resistance at pin 9.

As for your pullup readings, there's certainly something wrong with those, if the readings are correct. Double and triple check those, as it's easy to get the wrong pin on a PLCC. Since these are SIP resistors, it's a little odd that you'd have most of the pins working fine, and then just a couple with far too little resistance, but I suppose anything is possible.
Yes, nothing surprises me either.
Like when I couldn't get MIDI to work and it turned out that a filter (L306) had short-circuited to GND. Fair enough, but what made it so complicated was that the schematic only showed the filter to have two pins while in reality there was a middle pin (GND) as well, hence the name "T" filter I suppose ;)

Back to the resistor arrays. It's strange that they all show wrong values at just certain pins. Could it be a coincidence that they've all "gone bad", or do you think it's likely something else (like a short circuit as I suggested above)?
STe | MonSTer with dual IDE-CF memory card adapter | NEC Multisync 1990SXi | ST_ESSC | RSVE | Link '97 | Sony HD floppy drive/AJAX controller | Exxos HD floppy module | Battery-backed Ricoh realtime clock module | Discovery cartridge | C-Lab Unitor-2 | C-Lab Export | C-Lab Combiner | C-Lab Steady Eye | C-Lab Human Touch | Unicorn USB

Are you a good person?
User avatar
derkom
Moderator
Moderator
Posts: 1208
Joined: Sun Jul 29, 2018 6:45 pm

Re: Diagnostic cartridge (dead Mega STe)

Post by derkom »

Pacman wrote: Wed Apr 07, 2021 2:32 pm I've measured the resistance of RP26 directly, between pin 1 (common) and pin 8, likewise between pin 1 (common) and pin 9 and can confirm their lower resistance than its other pins (2-7). I assume I would have to desolder RP26 to confirm that its faulty or not.
Right, as you surmise, there are other opportunities for the path to VCC to include a lower resistance option, and the only way to test the resistor itself is by isolating it.
Back to the resistor arrays. It's strange that they all show wrong values at just certain pins. Could it be a coincidence that they've all "gone bad", or do you think it's likely something else (like a short circuit as I suggested above)?
If I were to hazard a guess, it's something else, but I hope that someone with a bit more knowhow than I have can provide some suggestions here, because this is the kind of problem I tend to haphazardly stab at in the dark until I strike gold, which isn't really the best diagnostic method. :)
User avatar
Pacman
Posts: 181
Joined: Mon Aug 28, 2017 12:56 am
Location: Norway

Re: Diagnostic cartridge (dead Mega STe)

Post by Pacman »

I believe I've made some progress!
I traced the discussed signal lines further, which led me to the two TOS EPROMs and furthermore exposed a short circuit which connected all the following together:

CPU pin 32 (A1)
CPU pin 42 (A11)
"Hi" TOS EPROM pin 12 (A0)
"Hi" TOS EPROM pin 23 (A10)
"Lo" TOS EPROM pin 12 (A0)
"Lo" TOS EPROM pin 23 (A10)
UA2 GAL pin 19 (I04)
UA2 GAL pin 8 (I8)

The reason for all that mess:
20210407-162313_P4070155.jpg
20210407-162313_P4070155.jpg (113.07 KiB) Viewed 3342 times
20210407-162325_P4070158.jpg
20210407-162325_P4070158.jpg (137.62 KiB) Viewed 3342 times
The red wire goes between the CPU pin 32 (A1) to UA2 (GAL) pin 19 (I04) in order to allow for HD floppy disk drive support. Upon closer inspection I noticed that the red wire had rested on top of UA2 GAL pin 8 (I8) puncturing the insulation and thereby shorting UA2 pin 8 to pin 19 :shock:
A positive side effect of discovering this is that now two of the pullup resistor readings have been corrected:

CPU pin 32 (A1) to RP24 (4.7K x8) pin 9
CPU pin 42 (A11) to RP23 (4.7K x8) pin 9

They now both read 4.7K !
The other two still give the wrong resistance though:

CPU pin 66 (D7) to RP26 (10K x8) pin 9
CPU pin 67 (D6) to RP26 (10K x8) pin 8

So I need to look closer into that. Unfortunately the computer still won't boot (and nothing is sent over to my STe's VT52 terminal via the null-modem cable and with the diag cart in the Mega STe -I tried all 3 serial ports to no avail).

The conclusion so far: sometimes hardware faults are just due to silly, trivial things! And at least those two resistor arrays aren't to blame.
STe | MonSTer with dual IDE-CF memory card adapter | NEC Multisync 1990SXi | ST_ESSC | RSVE | Link '97 | Sony HD floppy drive/AJAX controller | Exxos HD floppy module | Battery-backed Ricoh realtime clock module | Discovery cartridge | C-Lab Unitor-2 | C-Lab Export | C-Lab Combiner | C-Lab Steady Eye | C-Lab Human Touch | Unicorn USB

Are you a good person?
Post Reply

Return to “HARDWARE ISSUES”