Previous | Next
| From: Crown (all posts)
Date: 1998-07-06
Subject: Re: Turbo emulating
|
Hi,
>Are you sure, that it works on 1 Mhz? Ok, I know you wrote turbos in
>the past, but as I know it is faster than 1 Mhz.
Yep, I checked on one of the transfer routines. On the 1541 side there was 10 cycle between two bus change, and the Plus4 was reading on every 9 cycle, actually the last read was later. The routine on the Plus4 run in Slow mode (886.7Khz), so the two thing was pretty close to each other.
>Ok, maybe not, but it is not necessary an exact 1 us timing, because
>it needs at least 8 cycles to put two bits on the bus (if both DATA
>and CLOCK is used as data).
Yes, if you could syncronise the startup of the transfer to a 1 us error range then you only have to sample the lines every 8-10 usec,
but this sampling must be perfectly timed.
Ok you might get away with 3us delay on syncing, as you have at least 6usec safe area on every double bits.
To make it a little faster you might try to transfer more than 1 byte after a sync. This might be a little hazardous, as the crystal can have +/-some% error, so relying on it for extended time could introduce transfer errors.
Of course you definitely have to forbid multitasking for this to work....
You might also try to leave the GCR decoding for the Amiga, and just simply transfer the uncoded bitstream over.
Speaking about GCR, as far as I remember the Amiga can read GCR coded disks, and external 5 1/4 drives are existing for it. Why not using that for reading Commodore disks?
BTW anyone used GCR in fast mode on the Amiga? It could offer more than 800K on a 3 1/5 disk! I always wanted to try if its working,
but was too lazy to try... :(
Tibor Biczo / Crown of GOTU
|
| |
Copyright © Plus/4 World Team, 2001-2024. Support Plus/4 World on Patreon |
