exxos's Atari, Amiga & retro forum

Have you checked the 16bit datsbus on the DMA all is connected OK ? Does the DMA port work OK with ultrasatan or something ?

What is the cable on the sound chip ? Have you checked drive A: drive select line is OK as you have jumpers on it ? Did the machine ever work properly ?

The cable on the sound chip is a select of drive A/B as master/salve.
I checked the jumpers... I can switch between external as A and internal as B or the opposite. The DMA port seems to suffer of corruption. After playing to Xenon2 on the UltraSatan I obtain garbage in the folder: I guess I will add the pull-ups 10K? on the 16 bits data lines of the DMA but not sure if it will fix also the floppy drive in write mode.

Pull ups are on the booster anyway.. I would suggest clean CPU with IPA and maybe resolder the booster..

Have you been testing the floppy with the UltraSATAN connected? If so then disconnect it and try again.

I had similar issues on my machine with the TF536. The UltraSATAN's design is a bit iffy and pulls way too much current on the data lines, which can upset the DMA chip. The issue got worse over time and replacing the DMA chip (and using 4.7K pull-ups on the bus instead of the original 10K or the Exxos recommended 2.2K) was the only way to fix it in the end. I think the extra power draw, out of spec for TTL, was slowly degrading the chip.

(The chips used to interface the UtlraSATAN to the ACSI bus don't have an open collector so the DMA chip has to fight those chips to drive the bus.)

The 1772/DMA databus I suggest 10K... CPU address & databus 2.2K. It is also worth nothing lower values than 2.2K can also cause DMA errors. This is why I suggest people with the booster to remove the pullups on the CPU bus.. Its possible that could be a problem...

I guess 4.7K could be tried on the 1772 DATABUS as well instead of 10K.

stephen_usher wrote: 25 Jan 2021 08:28 Have you been testing the floppy with the UltraSATAN connected? If so then disconnect it and try again.

I mostly tested the floppy without the UltraSatan.

exxos wrote: 25 Jan 2021 16:33 his is why I suggest people with the booster to remove the pullups on the CPU bus.. Its possible that could be a problem...

The pull-ups on the CPU side is an interesting direction.
To be sure to understand well, I remove P100 and let as is? The pull-ups of the booster will take over right?

Sorry, I do not see any mention of the pull-ups on the CPU bus of the booster installation:
https://www.exxosforum.co.uk/atari/last/V1STE/index.htm

geogeo_28 wrote: 25 Jan 2021 18:13 Sorry, I do not see any mention of the pull-ups on the CPU bus of the booster installation:
https://www.exxosforum.co.uk/atari/last/V1STE/index.htm

They are on the booster...

Normally you don't need to remove the motherboard ones, there never has been any need to.. but maybe in your case there is.. we don't know yet..

They are listed here https://www.exxosforum.co.uk/forum/viewt ... 235#p27504 But P100 remains! the others can be removed.

As for DMA pullups, thats in the mandatory fixes.. under item 4.

I removed all my mods. The only remaining part for now is the booster at 8 MHz.
No floppy drive, no UltraSatan, no Gotek.
I tried with DMA pull-ups, removing the SIL, removing the blitter, nothing help unfortunately. Replacing the CPU of the booster by the original one, same.
By probing D0 of the CPU line I get this, which is terrible: I think I am making baby steps.
The noise on the video signal is for sure correlated to my issue with the floppy write.

I probed still D0 without the DMA (of course white screen). We can still see some data on D0. We see the spike (same than the RDY on the DMA I measured in the past) of around 2.4V in the middle of a data. Something clearly generate this strong noise.

When playing audio with the diagnostic cartridge of the Yamaha, D0 has no noise anymore: Possibly the Yamaha or the audio circuit creates some background noise?