REV 3 - REV 5 - The beginning (ST536)

coonsgm · Post by **coonsgm** » Sun Nov 30, 2025 1:28 pm

100% certain I burned BETA1SS.jed

prior to uploading the firmware I checked that the PLD was blank as well.

Post by **exxos** » Sun Nov 30, 2025 1:33 pm

coonsgm wrote: Sun Nov 30, 2025 1:28 pm 100% certain I burned BETA1SS.jed

Odd you got to run GB6 previously when TTram didn't have a clock..

What 4 buffers have you got next to the SDRAM? I did change them for faster types at some point..

I can isolate TTRAM in firmware to test the idea.. But can't do that until I get home tomorrow night..

coonsgm · Post by **coonsgm** » Sun Nov 30, 2025 1:50 pm

exxos wrote: Sun Nov 30, 2025 1:33 pm
coonsgm wrote: Sun Nov 30, 2025 1:28 pm 100% certain I burned BETA1SS.jed
Odd you got to run GB6 previously when TTram didn't have a clock..

What 4 buffers have you got next to the SDRAM? I did change them for faster types at some point..

I can isolate TTRAM in firmware to test the idea.. But can't do that until I get home tomorrow night..

No worries! Thank you for all the work going into this project! I'm happy to help test however I can.

The Buffers here:

: IMG20251130084823.jpg (378.84 KiB) Viewed 205 times

Post by **exxos** » Tue Dec 02, 2025 4:57 pm

Connected up my H5C1 to do some tests, and TTram test locked up about halfway in booting.. Made a small change and now it seems to be running fine again.. But the change I did is somewhat puzzling. I need to think about that some more.. But not much time for stuff today (or tomorrow).

: IMG_4095.JPG (330.4 KiB) Viewed 168 times

: IMG_4096.JPG (231.67 KiB) Viewed 168 times

Will upload BETA2SS shortly to the webpage..

Post by **exxos** » Wed Dec 03, 2025 9:27 am

Left the H5 on all night and seems fine...

: 20251203_092033.jpg (171.51 KiB) Viewed 140 times

Current problem is it seems totally unstable on @coonsgm and @PhilC machines.. The only clue so far is tristating TTram works.. But at worse, TTram should be corrupted and not cause the ST bus to become unstable like locking up or constant resetting etc. The ST bus is isolated during TTram access.. I've padded the timings out around the ST bus and DMA etc and nothing seems to help so far. It weird

Post by **exxos** » Wed Dec 03, 2025 9:20 pm

I was just poking around the data bus on a H5C1.. It never ceases to amaze... In particular as it has 2k2 pullups and one scope has it lurking around 2V for ages !

: IMG_4097.JPG (289.11 KiB) Viewed 117 times

: IMG_4098.JPG (282.66 KiB) Viewed 117 times

This is one of the address bus pins...

: IMG_4100.JPG (285.73 KiB) Viewed 117 times

Post by **PhilC** » Wed Dec 03, 2025 10:57 pm

Sorry. Had no chance to do anything, am I'll yet again.

Post by **PhilC** » Thu Dec 04, 2025 4:22 pm

And here's a bit of news with an unexpected solution from my point of view.

So I'd been testing all of the latest firmware versions that @exxos had made, including some he'd made for @coonsgm. I could get them to load and run Gembench some of the time, other times it would either reset the ST or come up file not found when looking for the .RSC file.

Now for the unexpected bit.

A CF card that runs quite happily with the original ST536 and even with this current version but without all of the mods, wouldn't consistently load Gembench and would cause a reset when trying to load Frontier.

I wanted to rule out an issue with the DMAchipsa I'd got IMP ones in my H5 and Mega ST, so swapped the H5 ones for normal chips and no difference. Then I wrote myself a new copy of my CF card, added my latest test stuff on to it and its only gone and bloody worked! EVEN IN THE MEGA ST!!!!!

So I'm happily leaving the Mega ST running loops Gembench now to see how things go......

Post by **exxos** » Thu Dec 04, 2025 7:40 pm

PhilC wrote: Thu Dec 04, 2025 4:22 pm Then I wrote myself a new copy of my CF card, added my latest test stuff on to it and its only gone and bloody worked! EVEN IN THE MEGA ST!!!!!

Really bizarre how a "bad" CF card could cause all those troubles. I have proper isolation on the IDE databus these days. IDE can only be active when the CPU accesses it..

The only other possibility is a glitching IRQ line which I have seen before.. But it would not really explain how formatting the card would fix such issues though..

Maybe the driver was somehow corrupted and caused TOS to go nuts at odd times.. very strange! But at least its running stable now.

I did document weird issues with a CF card from @Badwolf where it seemed to become more unstable the higher the 5V rail was. But it didn't show that behaviour on the later boards with the IDE fixes anyway.

I think some CF cards only work by voodoo magic half the time!

Badwolf · Post by **Badwolf** » Thu Dec 04, 2025 8:35 pm

PhilC wrote: Thu Dec 04, 2025 4:22 pm its only gone and bloody worked! EVEN IN THE MEGA ST!!!!!

Has doing that changed/replaced the HD driver at all, or is that loaded from elsewhere?

BW

REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)

Re: REV 3 - REV 5 - The beginning (ST536)