Page 1 of 8

Paradroid STE DMA issue blog

Posted: 02 Feb 2021 19:43
by Steve
Hi, just doing some work on Paradroid STe with the infamous DMA issues writing to Ultrasatan. He has the 32mhz accelerator installed. I am looking to geogeo's recent thread for guidance as he managed to solve his problems: https://www.exxosforum.co.uk/forum/viewt ... 1&start=20

So far:
* P100 resistor array is 2.2k, so will swap this to 1.2k
* Remove the other resistor arrays as recommended
* Socket DMA to try different chips
PXL_20210202_160349956.jpg
PXL_20210202_160855176.jpg
Oh bloody hell how am I going to get these resistor arrays out with the 32mhz booster over the top and soldered into place...
PXL_20210202_161539903.jpg
With a lot of careful patience, desoldering gun & flux & wick I carefully managed to remove them with some tweasers.
PXL_20210202_162801111.jpg
Just a check to see if everything is still working:
PXL_20210202_175242620.jpg
PXL_20210202_173819747.MP.jpg

Re: Paradroid STE DMA issue blog

Posted: 02 Feb 2021 20:17
by exxos
Steve wrote: 02 Feb 2021 19:43 Hi, just doing some work on Paradroid STe with the infamous DMA issues writing to Ultrasatan. He has the 32mhz accelerator installed. I am looking to geogeo's recent thread for guidance as he managed to solve his problems: https://www.exxosforum.co.uk/forum/viewt ... 1&start=20
He didn't though ? He still has issues ?

geogeo_28 wrote: 01 Feb 2021 18:39 I can claim victory on the floppy read/write at 8/32MHz but for the UltraSatan I spoke too fast.
The irony is, with the UltraSatan, at 8MHz I still experiencing data corruption on write but no issue at all at 32MHz.

Re: Paradroid STE DMA issue blog

Posted: 02 Feb 2021 22:52
by Steve
Oh ok well... I'll just do what I can and see if my results are different to geogeo's. @exxos do you recommend running with the address bus resistors removed like this or shall I try 2.2k also.

Re: Paradroid STE DMA issue blog

Posted: 02 Feb 2021 23:55
by exxos
If the booster is in, then remove the MB roms...

*IF* They are all 10K they could be left in... I can't say for 4.7K ones.. Best to remove them...

As a side note, are people using a good PSU for ultrasatan ? as "chargers" can cause all sorts of chaos..

Re: Paradroid STE DMA issue blog

Posted: 03 Feb 2021 09:59
by tzok
I've also found the ACSI cable to be extremely important to the reliability of 3rd party ACSI devices. 20 cm ribbon cable seem to be more or less a maximal length. Also, the crappy fake Apple/Samsung USB chargers often bundled with various 3rd party devices are causing problems. Use a good quality PSU, or at leas a genuine Apple/Sony/Samsung USB charger to power your UltraSatan.

Re: Paradroid STE DMA issue blog

Posted: 03 Feb 2021 11:40
by Steve
This Ultra is internal and powered from the motherboard. So no dodgy external power.

Re: Paradroid STE DMA issue blog

Posted: 04 Feb 2021 10:36
by Steve
UPDATE ON TESTING:

TRIED WITH TWO DMAS:

DMA - CO25913-38 PH23-030 / 8M4 44

RESULT: 8Mhz write corruption / 32Mhz write corruption

DMA - CO25913-38 PH23-030 / SJ4 95

RESULT: 8Mhz write corruption / 32Mhz OK

Re: Paradroid STE DMA issue blog

Posted: 04 Feb 2021 10:56
by exxos
Cannot tell, but does not look like the DMA pullups are fitted?

Re: Paradroid STE DMA issue blog

Posted: 04 Feb 2021 11:04
by Steve
I'll try that next @exxos cheers

Just double check, you mean the 10k pullups on the WD1772 chip? https://www.exxosforum.co.uk/forum/viewt ... =17&t=1417

Re: Paradroid STE DMA issue blog

Posted: 04 Feb 2021 11:22
by exxos
Steve wrote: 04 Feb 2021 11:04 Just double check, you mean the 10k pullups on the WD1772 chip? https://www.exxosforum.co.uk/forum/viewt ... =17&t=1417
Yep.