www.jupiter-ace.co.uk

Hardware articles index > The ACF version 3

The ACF version 3


T
he experiments continue with ACF 3 in ACF 2 we had introduced the second IDE channel, well in theory we could have a CF cards as Masters and Slaves. Eagle Cad board files were sent to Futurlec for pcb production and to test out the schematic drawing, as it was the first time I had used Eagle Cad. After a short wait the pcbs arrived and straight away I realised that I had made it difficult for us to solder up the pcbs to test. In Eagle there was no library file for the type of edge connector used by the Ace. I had edited a PC 8-bit socket as a way of getting over the problem. What I should have done was to design one from scratch and designed a pad to which the edge connector could be solder easily. But, the pcbs were here now and when solder up seemed to work fine.

A Few ACF AceForth Words.
Select Drive, on stack 0 for master 1 for slave
: SD 0= IF 224 ELSE 240 THEN 1663 OUT ; 
Note: when selected slave the green light stays on?  unresolved as yet![now fixed in ACF 4]

8bitmode, puts CF card into 8-bit mode
: 8bitmode 1 383 out 239 1919 out ;

LBA word expect number 0-32767 (on stack,)
 used as Logical Block Address also selects 1 sector for count

: LBA 1663 IN 240 AND 224 OR 1663 OUT 0 1407 OUT DUP 256 
  / 1151 OUT 255 AND 895 OUT 1 639 OUT ; 

(Load sector to screen)
: LD 0 0 AT 32 WC BZ WD 512 0 DO 127 IN EMIT LOOP ; 

Use LBA to select a sector then LD to load it to screen like
: x 32767 0 DO I LBA LD LOOP ;


: WC 1919 OUT ; 			(Write Command)
: WD BEGIN RS 8 AND UNTIL ;		(Wait for Data request)
: BZ BEGIN RS 128 < UNTIL ;		(wait while busy)
: RS 1919 IN ; 			(Read Status register)




Yes! that really is two 8 meg Compact flash cards, Master and Slave.





Here are is the test PCB back, and populated first tests give us the results we expected.

Pull up resistors fix on the RD and RW IDE lines.

After some more testing it became apparent that the RD and RW signals where not stable enough using the CF cards internal pull ups alone. So, two 10K pull up resistors were added for better reliability. Now for a new pcb with the 96KRampack and the ACF on the same pcb, this would become the ACF Version 4. Edwin has also designed the ROMBoard v2 pcbs. This is where we hope to patch the Jupiter Aces Forth ROM