IDE (2)

From: Ruud Baltissen (
Date: 1998-12-12 21:38:07

Hallo allemaal,

It seems there are 2 groups which each their own idea how to connect the IDE to 
your C=: the Q&D-boys (Quick&Dirty) and the CbS-gang (Compatible but Slow). As 
I once said before: most of the SW can be used by both groups. The SW needed to 
create a file, or even better, to create a file inside a image, will be the 
same for both systems. The only difference will be the way the SW is started 
up. Using a C64 you'll probably save some parameters in RAM and fire up the SW. 
Using a 1541 some more actions are needed but you end up firing up the same SW.

For your information: I'm planning to test everything using my C64 in the first 
place. This is much more convenient then reprogramming a EPROM for your 1541 
all the time. 

I hope this brings both groups more together.

The hardware:

At this very moment I can only read and only 8 bits wide. And that's all for 
the moment :-( The main problem is/was the way the C64 behaves. As base of my 
HW I used a mix of several designs but it still looks like Johns design. The 
time between an address is valid (and that is after the rising of CLK2) and the 
pulses generated for the IDE proved to be to short. I could write 16 bits but 
read only 8 bits. And the reading did not work out fine: from every cycle of 
reading 256 bytes I kept on missing bytes varying from 0 up to 6 bytes. I 
didn't know for sure until I changed the design this very morning. The 
IDE-addresslines and CS-signals are generated using a 74ALS573 to latch them. 
Disadvantage: I need 3 writecycles to program an IDE-register (first one). 
Advantage: I only need 3 bytes in the memory map. Reading a buffer or sector 
still goes at the same speed. 
At this moment I need 7 ICs: 3 74als573, 74ls138, 74ls32, 74ls245 and 74ls14.

Rereading the above I wonder if I could have the same result with delaying the 
start of pulses. It will make them shorter but as IDE runs fine on a 8 MHz ISA 
bus, this won't be a problem.

After sending this I resume testing....

Groetjes, Ruud

This message was sent through the cbm-hackers mailing list.
To unsubscribe: echo unsubscribe | mail

Archive generated by hypermail 2.1.1.