Mega STe stability issues.

Problems with your machine in general.
User avatar
exxos
Site Admin
Site Admin
Posts: 23499
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Mega STe stability issues.

Post by exxos »

Check the schematic https://www.exxosforum.co.uk/forum/viewt ... f=31&t=100

Looks like 10k on the address bus and 4.7k on the data which is the same as later STEs. I can't say if changing the pull ups will make any odds as I've never done research on the MSTE. but for the time your spending on it, just grab some 4.7k sils and tag them on the bottom of the board and see if it helps or not.
https://www.exxosforum.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxosforum.co.uk/atari/store2/ - All my hardware mods for sale - Please help support by making a purchase.
viewtopic.php?f=17&t=1585 Have you done the Mandatory Fixes ?
Just because a lot of people agree on something, doesn't make it a fact. ~exxos ~
People should find solutions to problems, not find problems with solutions.
atari030
Posts: 315
Joined: Mon Feb 12, 2018 12:43 pm

Re: Mega STe stability issues.

Post by atari030 »

I'm picking the sockets are going to be more of an issue. The ROM's, for example, trace back to the CPU VIA the Blitter socket. Dry joints there caused mine to white screen on boot. I've had to repair about four traces besides the one I cooked.
Dbug
Posts: 71
Joined: Sat Sep 11, 2021 2:49 pm

Re: Mega STe stability issues.

Post by Dbug »

Guess I'll read through all the BAD DMA Myth page, Mega STe schematics, and probably also investigate if I have dry joints.
User avatar
exxos
Site Admin
Site Admin
Posts: 23499
Joined: Wed Aug 16, 2017 11:19 pm
Location: UK
Contact:

Re: Mega STe stability issues.

Post by exxos »

Dbug wrote: Fri May 06, 2022 8:58 pm Guess I'll read through all the BAD DMA Myth page, Mega STe schematics, and probably also investigate if I have dry joints.
The "bad DMA thing" I assume is not relating to the PLCC DMA anyway. I have not researched that one but I assume it may well need pullups like the other DMA chips do. Of course its probably not a good idea to try soldering resistors on the PLCC DMA chip. so they would have to be soldered elsewhere probably.
https://www.exxosforum.co.uk/atari/ All my hardware guides - mods - games - STOS
https://www.exxosforum.co.uk/atari/store2/ - All my hardware mods for sale - Please help support by making a purchase.
viewtopic.php?f=17&t=1585 Have you done the Mandatory Fixes ?
Just because a lot of people agree on something, doesn't make it a fact. ~exxos ~
People should find solutions to problems, not find problems with solutions.
Dbug
Posts: 71
Joined: Sat Sep 11, 2021 2:49 pm

Re: Mega STe stability issues.

Post by Dbug »

Ok, so I ordered some GAL, 9pin resistor bridges, and some new IDC connectors to shorten the UltraSatan cable (I never managed to remove one of these connectors without breaking the plastic retainers!).

I also started writing a small program to do some corruption checking, so far it does not do much and has not found anything, so probably what it's doing is not actually triggering any issue, but basically what it does:
  • It has a large memory buffer (right now it's a 3 megabytes BSS section but I could simply check in the base page the allocated size)
  • Loops forever (need to CTRL-C to exit)
  • Each loop generates a 32 bit random number
  • This number is then written to the buffer to fill it with a DBRA loop
  • A second loop checks that all the values in the buffer still match the random value
  • The buffer is written to disk in a file (FCREATE/FRWITE/FCLOSE)
  • The buffer is read back from the file (FOPEN/FREAD/FCLOSE)
  • The read check is done again
  • Start next loop iteration with a new random number
My expectation was that the problem would at least trigger just by doing the writes and read back through the disk, but that did not happen.

Obviously doing writes with a DBRA and a move.l will probably not stress the bus, so my other ideas were to do the write using multiple methods:
  • Switch between 8mhz without cache and 16mhz with cache
  • Repeat the test for both ST RAM and ALT RAM
  • Test with the current move.l/dbra
  • Test with a partially unrolled movem.l to write at the maximum 68000 speed
  • Test with fast writes done with the blitter (with the values in the internal pattern memory for maximum throughput)
For the disk, I don't see any other way of doing things, other than:
  • Do plenty of small reads and writes instead of just one FWRITE or FREAD
  • Possibly test file copies between the two UltraSatan SD cards by doing small reads and writes.
  • Testing the writes on different partitions (but I don't see why that would be different from a disk point of view)
As a side note, it seems that having the Mach32 outside the MegaSTe case seem to reduces the cases of crashes and glitches, could it be some kind of electromagnetic radiation caused by the card or all the cables?
User avatar
viking272
Posts: 130
Joined: Mon Aug 10, 2020 11:32 am
Location: Reading, Berkshire, UK

Re: Mega STe stability issues.

Post by viking272 »

Dbug wrote: Wed May 11, 2022 6:57 am As a side note, it seems that having the Mach32 outside the MegaSTe case seem to reduces the cases of crashes and glitches, could it be some kind of electromagnetic radiation caused by the card or all the cables?
Could it be an overheating issue?
Steve
Posts: 2570
Joined: Fri Sep 15, 2017 11:49 am

Re: Mega STe stability issues.

Post by Steve »

Dbug wrote: Wed May 11, 2022 6:57 am As a side note, it seems that having the Mach32 outside the MegaSTe case seem to reduces the cases of crashes and glitches, could it be some kind of electromagnetic radiation caused by the card or all the cables?
No point speculating, this is why I suggested having it tested on another machine to rule the nova out. Plus it is more likely to be bus loading than 'radiation'.
Dbug
Posts: 71
Joined: Sat Sep 11, 2021 2:49 pm

Re: Mega STe stability issues.

Post by Dbug »

Steve wrote: Wed May 11, 2022 1:32 pm
Dbug wrote: Wed May 11, 2022 6:57 am As a side note, it seems that having the Mach32 outside the MegaSTe case seem to reduces the cases of crashes and glitches, could it be some kind of electromagnetic radiation caused by the card or all the cables?
No point speculating, this is why I suggested having it tested on another machine to rule the nova out. Plus it is more likely to be bus loading than 'radiation'.
How can having the card inside the case vs outside the case (still connected and in use of course) have any impact on the bus loading?
Steve
Posts: 2570
Joined: Fri Sep 15, 2017 11:49 am

Re: Mega STe stability issues.

Post by Steve »

@Dbug I didn't see you meant outside. Maybe you have a faulty ribbon cable.
Dbug
Posts: 71
Joined: Sat Sep 11, 2021 2:49 pm

Re: Mega STe stability issues.

Post by Dbug »

Steve wrote: Thu May 12, 2022 11:03 am @Dbug I didn't see you meant outside. Maybe you have a faulty ribbon cable.
Np :)

By the way, I remember having seen a couple weeks ago, somewhere, somebody who posted a photo of a replacement cache for the VME area, with a modified Nova that had the ISA connector properly setup without all the big cables so the ISA card could be safely put inside.

Does anyone remember having seen the project?
Post Reply

Return to “HARDWARE ISSUES”