Dave Dunfield wrote: Dave, thanks for mentioning your site. I knew someone had some material online, and it is probably you I was thinking of. As to the chicken or the egg, I took the approach of getting the bios code to run the horizon controller, and generated a horizon CPM that would run one of my Tarbell controllers in the box with the Horizon controller. it only took a bit of editing to boot the cpm with both, and it would sysgen and format either diskette, 8" or 5 1/4" and also run pip for copying data to and from each format. I just split it up so I had 2 drive letters for each type, so A,B was one, and C,D was the other. I flipped the code and merged the horizon code into the tarbell code later. I don't recall back then as to my chicken and egg problem which was how to get the sources needed to do the tarbell onto the horizon, but I think I captured a listing via a serial program and edited the tarbell bios stuff into the cpm on the horizon. After this bootstrap, which involved maybe a hundred lines of code, I could merge up enough to do everything via diskette, then finish. Now, if you have access to moving data from a simulated CP/M to 8" and 5 1/4" you could do the editing anywhere, and go back to the 5 1/4". but someone has to bootstrap you onto the 5 1/4" northstar, since it is pretty much one of a kind. many different ways exist onto 8" but the problem dave mentions means you have to start up somehow with a few 5 1/4" diskettes, and generate your boot, and I assume a way to assemble, test, then use the horizon bios, so you can adapt what you need. The only thing that a true horizon has is a 256 byte boot, and I don't know of anything that does anything but floppy booting exists. I remember the boot rom gets mapped into somewhere in the Exxx range, and reads in a 256 byte sector, then jumps to it, before going away. I don't think you could pull it and get a serial program into it to read one of the serial ports, but I won't say never to the crowd on this list. And my memory of 256 bytes may be off, I vaguly recall maybe on 32 bytes are visible and used by the processor at boot. People had to use pretty small proms in the early days of the hobby. Jim

On 25 Nov 2006 at 3:27, jim stephens wrote: I recall that the disk controller had a small bipolar prom and that it was indeed 32 bytes. It was miles better than toggling the boot code by hand... A 256-byte boot prom back then would have been an almost obscene luxury... ;) --Chuck

On 25 Nov 2006 at 13:29, Dave Dunfield wrote: My comment was directed toward your reference to the Tarbell controller, not the NS. A question crops up: why? What's the point--you need only enough to load a sngle sector. Was the NS controller so badly implemented that it required that much code to read a single sector? Cheers, Chuck

On 25 Nov 2006 at 14:20, Dave Dunfield wrote: My bad--I ignored the number of right-pointing doohickeys in the post- -it was Jim who mentioned the Tarbell. Probably so--and perhaps it was cheaper to do that way, given the soft-sectored alternatives. Intel used a two-board 3000-series bit- slice implementation on the MDS and IMSAI used a two-board 8080 implementation on theirs. (I never could get the IMSAI controller working reliably enough for everyday use.) Even the WD1691 and 1771 were very expensive items when they first appeared. Cheers, Chuck

Hi guys. I've got a mac classic ii analog board... but it's for 120v instead of 240v. The circuit board looks identical ... as do most of the components. Anyone know what would need to be changed to make it 240v? Thanks, Mark