Poke 11895 to change stepping rate in realtime on 8" drives

Post Reply
lowrybt1
Posts: 212
Joined: Sun Mar 08, 2015 3:42 pm
Location: New York State

Poke 11895 to change stepping rate in realtime on 8" drives

Post by lowrybt1 »

The Dragon Byte Technical Manual for OSI disk interfaces indicates that disk stepping rates under 65D (at least for 8" drives) can be changed on the fly by poking a value into 11895. Poke 11895,3 sets stepping rate to 3ms. Poke 11895, 9 sets stepping rate back to 9ms. According to the manual, this might be necessary when using 'later edition' D-series and above Siemens 8" floppy drives. I guess this might be useful if a system with two 8"drives mixes older or newer Siemens or Shugart drives. Has anyone tested this before?
C8PDF w. 48K, 2x 520 24K RAM boards and Glitchworks 64K board
OSI 567 Telephony board
Spare 8" drives
Klyball D-13
Mark
Posts: 297
Joined: Tue Sep 16, 2008 6:04 am
Location: Madison, WI
Contact:

Re: Poke 11895 to change stepping rate in realtime on 8" drives

Post by Mark »

I'm not sure which version of OS65D was referenced for that poke. At least with OS65D V3.2+ that address seems to be the high order byte in the command lookup table for the SE command. See OS65D_Dissasembly_Manual

The head step seems to happen around $2683 with a delay value stored at location $EF, but you can't POKE that when running BASIC. Zero page is constantly swapped in and out by BASIC, so that location changes meaning. Step delay at $EF seems to be set once at $22B9 during OS load, and swapped in & out with BASIC's zero page storage at $2F79+ with $EF being stored at $3068. BASIC swaps it back when making OS calls.

So a poke at 12392 might be the address you are looking for on OS65D V3.2 & 3.3 with BASIC running. I don't have older OS versions for 8" systems to see what they did.

I wonder why a slower step rate would matter to newer drives and make them operate "ineffectively"? Maybe I misunderstood their statement. Certainly setting it faster for drives that can handle it would make seeks faster.

(My OS65D version seems to use a value of 8.)
lowrybt1
Posts: 212
Joined: Sun Mar 08, 2015 3:42 pm
Location: New York State

Re: Poke 11895 to change stepping rate in realtime on 8" drives

Post by lowrybt1 »

Thanks Mark. This is not the first time something in that technical references seems off. Another example: the reference illustrates certain trace cuts and jumpers for Siemens FDD-100-8D/E drives that will be used as B drives with OSI machines. Looking at 3 FDD-100-8D drives used as drive B on C8 and C3 systems, not a single one used the same trace cuts and jumpers shown in the Dragon Byte book.
C8PDF w. 48K, 2x 520 24K RAM boards and Glitchworks 64K board
OSI 567 Telephony board
Spare 8" drives
Klyball D-13
Post Reply