exxos wrote:Similar they blame "DMA chip" as bad. Nothing to do with it. Just noise on the bus.
Well. I think it is clear that there is a problem with the IMP DMA chip. If it is really "bad", or "too god" (and then too sensitive to noise), it is just semantics. That matters only if you want to blame somebody, which might have been an issue back at the day, not now. For the purpose that most people care, the IMP chip has more problems, and the ST has less problems than the STE, even with the IMP DMA chip.
Yes, I think we all agree that the -001 chip doesn't solve all the problems. And that the ST (non STE) also has problems. But it is a fact that the IMP chip is the one with most chances to have problems. And more so on the STE than on the ST.
I am just curious about what makes the difference between the STE and the ST. It would be interesting if somebody with an STE motherboard that has problems, tries to "convert" the problematic ACSI interface to "ST sytle". That should mostly be removing the TTL buffer that the STE design added on the ACSI bus.
Regarding ground and earth problems with the PSU. It is definitely a problem with floppy drives. I have a floppy emulator that doesn't work reliably with an external PS. It sounds as could be an issue with ACSI devices as well. But here, soviet992 tried using internal power without noting any difference.
It can only take just 1ns for something to work and not work.
Can't say I am 100% sure, but that doesn't seem to be exactly the problem here. Seems to be more an issue of noise creating an extra edge on the DMA chip. As I posted in another thread sometime ago, from the point of view of synchronous logic, the DMA chip has the worse internal design.