ST536 STE EDITION

[exxos]

Current benchmarks with TOS 206, fast ROM and BLTFIX. With and without blitter enabled.

IMG_3296.JPG

IMG_3297.JPG

[Cyprian]

nice,
ALT-RAM figures look better than earlier

[exxos]

curren12 is the latest firmware.

Steve's boards seem to need to slow slew to work at all. Otherwise TTram "vanishes" totally.

Later boards seem to be happier with fast slew. ( nobody has the later board yet other than me)

[PhilC]

Sounds like it'd be a good idea to forget about the earlier boards then

[exxos]

Sounds like it'd be a good idea to forget about the earlier boards then

It seems your board works fine with the fast slew. Its just steves which needs slow slew. But i can't figure out why.

EDIT:

OH thats odds, if I put finger on top corner of the PLD TTram starts passing :shrug:

[exxos]

@Steve's board is weird :lol:

If I give the left SDRAM chip a quick spray with freezer spray, it then starts working fine... :shrug:

IMG_3298.JPG

[Steve]

:oops: :oops: :oops: :oops: :oops: :oops:

[Badwolf]

Dry joints that pressure or temperature gradient compensates for?

BTW, I'm not sure there is a burst mode for write. It's not something I've ever come across anyway.

TBH, I've found burst mode to be pretty unremarkable in terms of general performance. If it's giving you headaches disable it and keep it in the back of your mind for later upgrade.

BW

[exxos]

It gets worse... Steve's board and it seems to work when it's cold, and my board only seems to work when it warms up.. Literally I've had the board running almost continuously for two weeks and it would pass fine.. Leave it off for a few hours and go back and is completely unstable again.

I'm not sure what else to try. I may just remove the SDRAM and put some SRAM on there to try things out, make sure the core system is working before trying to diagnose TTram faults.

Current12 which I turned BUST OFF, TTram isn't even found now. If I go back to when BURST was ON, it finds TTram.

I think the problem is what I thought originally, I change something ever so slightly in the code, and the timing is everything change just enough to break the timings of the SDRAM controller. Mixing temperature variations which are changing the timings yet again ..It's basically a never-ending battle..

EDIT:

I disabled the TTram stuff and with slow slew I get the odd corruption on boot back again. So maybe I should focus on figuring that out as obviously just changing the slew to fix it isn't a good idea.

[coonsgm]

@coonsgm just thought, have you hot the 33pf cap on the CPU ?

viewtopic.php?p=128013&hilit=Bad+clock#p128013

I don't have the cap on the CPU. I believe I removed it previously. Can you share a photo of which pins it goes between?