Here is the github page (thanks, Wing, for pointing me to it):
https://github.com/btb/SmartDiskII
The front page claims that they replace the P6 PROM with a "2716 - compatible" EPROM. Which may not be fast enough to make the SmartDiskII work reliably in any Apple II.
In post #31 of this thread I've shown the timing calculations which call for a 150ns access time for such an EPROM substitute for P6:
https://www.applefritter.com/content/ibs-ap53-floppy-disk-controller-pcb-marked-ap41-why
... but this includes some reasonable safety margins. If these are taken out then 200ns EPROMs would fit the bill, but only barely, and no reserves. From the mid 1980s on, some semiconductor manufacturers made the venerable 2732 on newer and faster NMOS processes, to cut costs of course (smaller die, more yield). ST had the M2732A-2F1 whose fastest speed grade was 200ns. So with a little bit of luck and no safety margin left, this part might work in the SmartDiskII.
AMD had the AM27C191 which is an almost pinout compatible 2k x 8 EPROM with 25ns, 35ns, and 45ns access time grades, certainly fast enough to replace the P6, but these are as rare as hen's teeth.
So here is the question:
Does anyone reading this have a SmartDiskII that really works reliably even as a DISK II controller substitute ?
(I have a hunch that if you code the SmartPort routines in a special way, the timing problem with the slow EPROMs could be dodged, but this is only a conjecture, and does not apply to Apple's P5 bootstrap code, or Apple's RWTS code in general).
Comments invited !
- Uncle Bernie
I do have an early SmartDisk II board. It does work, but when I first received it there was an issue that caused it to wipe out any disk in the drive! On boot, it would permanently go into WRITE mode. This did indeed turn out to be a timing issue, but between the 245 and the 9934/LS259.
The board was shipped with a 74ACT245 and 74LS259. After some testing, I discovered that the board would work if the ACT245 was changed to a LS245 OR the LS259 was changed to a real 9934. I discussed this with the designer Brad Bell and I think he stopped using the ACT parts.
In post #2, 'jeffmazur' wrote:
" This did indeed turn out to be a timing issue, but between the 245 and the 9934/LS259 "
Uncle Bernie comments:
Note that issues with the mode register (the 74LS259) have nothing to do with the EPROM P6 substitute possibly being too slow (unless the EPROM is fast enough, of course), it's a different problem lurking in all 6502 systems I wrote about several times on Applefritter (I got burned by it some 38 years ago and had to redesign one of my products). This is caused by the 6502's address outputs crumbling very quickly after the end of PHI2 and if some write gate is still open at this time, bad things can happen.
Note that Woz intentionally ignores the PHI2 offered by the 6502 in all of his Apple computer designs (beginning with the Apple-1), which, as the story is told, once led to an argument with Chuck Peddle, who had brought the only existing 6502 ICE to help Woz to get the Apple-1 going, and after Chuck accused the Woz to have "no idea about proper 6502 system design", Chuck left the premises in anger. And Woz continued to use that trick. Which gained him a little bit of more timing reserves (some 10's of ns) to dodge the crumbling address issue. But if you add enough further gate delays to make some write gate signal, then this eats into the budget and eventually reaches into the crumbling addresses. And the poor 74LS259 will do what it believes is right, changing its internal state based on these crumbling addresses. Not good !
During development of the 'IWMless' I again ran into this crumbling address problem and had to add latches at the address inputs, gated appropriately to keep the addresses stable until the write gate of the mode register is turned off.
This thread is looking for P6 substitution-by-EPROM timing issues !
What I am after with this thread is to find out if some of the SmartDisk II work (I would not expect this based on the fact that all 2716 and most 2732 are too slow to replace the P6 PROM, at least from theoretical calculations) and if some are found to work fine, which type / manufacturer this miracle EPROM has, because these would be fast enough.
If all SmartDisk II do work, fine, no further action needed to help SmartDisk II owners with fixing that issue - I could do this as I've recently resurrected a 20 years old P6 substitute design of mine based on a GAL22V10 (the hard part was finding my source code, took longer than doing the original design, so I could have started from scratch, creating a new source).
So, still looking for user experiences with the SmartDisk II !
Comments invited !
- Uncle Bernie
I have a couple SmartDiskII cards that don't work, and I've never taken the time to look at why. They were built with chips that may be a little dodgy, so I suspect the 74LS259s or even the 74LS245s to be questionable. Out of the same batch of PCBs I have a couple others that seem to work, at least as Disk ][ controllers. I've never gotten anything to work as a SmartPort controller, not even my real LiRON card.
I could send you the non-working SmartDiskII cards if you want to test them.