This document last edited Nov 4 2012. Additions made July 29 2016 (c) Herb Johnson 2016.
Nick Papadonis contacted me in May 2011 about work on his CCS boards, to run in an IMSAI chassis, and a Northstar Horizon he was trying to get running. Here's our discussion of the Northstar Horizon, through November 28 2011. Issues are mostly about the drives and hard-sectored operation. I added some comments in Nov 2012.
A Northstar system I repaired in 2006 and later is on this Web page link. There's also a 2011 restoration described on that page. Here's a link to my Web page of Northstar documentation and notes, as part of my S-100 support.
- Herb
After he got the Horizon running, Nick gave me this background: "If you have a N* writeup I'm not sure if I mentioned this. My N* was is very bad shape. After checking out the power rails I discovered one of the bus power rails (8 or 16) was not delivering power. Also, 12V was not working on the floppy drive power molex connector.
Using the schematics, VOM and suggestions from you I traced the problem to a couple voltage regulators and tantalum caps. First I found a blown tantalum between the two heat sinked 7812 TO packages. I replaced that and the cap blew again at power on.
I then replaced ALL the voltage regulators and tantalum caps in the section of the motherboard. The parts were inexpensive enough and figured both would be under great stress with age. Problem fixed!
Note: I had to be extremely careful with the N* MB when soldering and removing solder. Even when I set my iron to a low temp, traces lifted up a little. They are not very strong or perhaps humidity caused them to separate from the MB." - Nick
Our discussion follows below. - Herb
May 8th, Nick asked: "In regards to your Horizons. Are you running the drives with the original NS FDC or are you using a different manufacturer? I'm asking because I tried the Shugart [SA400] drives in my NS Horizon with the ImageDisk PC. I put in a soft sectored floppy into the drive. I was able to format and read back fine. So it appears these drives can be used with a soft sectored controller.
I was wondering if it would be possible to just remove the NS FDC and use my CCS FDC (soft sector support). I would then use modified CBIOS that covered the non-NS FDC and NS-IO configuration. If the drives can be used in soft sectored mode, I think this will work. - Nick
herb says: "Hard sectored" is a function of the controller, NOT THE DRIVES, almost without exception. The drives don't "care". My Northstars run with their native controller and I believe they are Tandon 5.25" 40-track drives, but I've used a variety of 360K drives in Northstars.
Why remove the Northstar controller at all? The CCS FDC runs in its own I/O space. The Northstar floppy controller runs in upper RAM, around E800 or so, like the serial and parallel ports. No conflicts except the N* ROM boots the N* controller.
You could attach drives to BOTH controllers and run EITHER set of controllers and drives. Boot NSDOS or CP/M from the N* "drives" (and controller) OR CP/M from the CCS "drives" (and controller). Or physically move the drive cable from one controller to the other, notwithstanding any drive jumper or cable changes needed (probably none). The possibilities are obvious.
Nick says: "I am considering removing the N* controller because 1) it's hard sectored and I have to punch disks (I have a punch for this Dwight made though) 2) so far the FDC does not seem to work.
I'm using Dunfield's NST program. I upload Dunfields modified NSDOS image Intel hex using a ROM monitor option for the N* Z80 card. It is an image of NSDOS with his NST (imaging support he wrote) monitor running. I can then from my PC connected via serial port, move the disk head, however can't successfully write/read a disk image.
I think I either need a "known working" hard sectored disk to validate this or give the controller to someone with known working N* setup. I'm guessing the problem is either the controller or perhaps me incorrectly punching the hard sectored disk.
I worked on my N* and Kaypro tonight. I was able to get the Kaypro II to boot CPM after fixing the drive latch mechanism and creating a boot disk (using [Dave Dunfield's] ImageDisk)! Not as much luck with the N*.
I wanted to make sure my N*'s Shugart 400 drives somewhat worked. I placed a DD soft sectored disk in them and ran the cable to an ImageDisk PC. I was able to format the disk and read back data successfully. I then punched this validated disk (10 sectors) and connected the drive back to the N* and loaded up Dunfield's NST program into RAM. I then wrote a NSDOS boot disk, heard the drive clicking and was presented with success. I then tried reading it back and that failed.
At this point I'm trying to reduce the variables. Is the problem the floppy disk (I punched it myself with Dwight's punch), the drive or the controller. I'm wondering if I could send you a couple of my "punched" disks and if you could try them in your system? That way, I could at least validate the media is ok. Would this be possible?
Herb says:
I'm working with Northstar Horizons this week, with CP/M boot disks, hard sectored, and static RAM boards. I'll send you a boot disk and I'll note which floppy controller models it "boots" from.
I copied two versions of CP/M I have, and I have a few disks with some errors on them, so you will get five diskettes. The single-sided Shugart 400s seem to find more errors on the old hard-sectored diskettes, than the tandon TM-100's two-sided drives. So I upgraded to Tandon drives on that system.
One version of CP/M on my disks won't run the "configure" program, it gives the dreaded "synchronization error" that means the running CP/M has a secret serial number that does not match the copy that "configure" uses. The other version of CP/M has a configure program that seems to work OK. The CP/M's are set up for single-sided drives, that's why I was messing with re-configuring. One version of CP/M does not see the other version's disks as "formatted" correctly, for reasons unclear to me but possibly related to "sided"ness. But both versions boot up OK.
May 2011: Nick, it's interesting to hear about Dwight Elvey's hard-sectored disk punch. [I first heard about it] in SEBHC, the Heath 8-bit discussion group. (Nick later said he heard about the punch in www.vintage-computer.com) Do you mind telling me what happened with his plan to build several of them? - Herb
I think he built many. I recall it was about $30. He must have had access to a good machine shop. It came out very professional, but I would like to know if it punched the disk correctly. :) - Nick
I validated that I could use Dwight's punch on my disks and the N* would read and write them fine. I was able to copy your LifeBoat CPM disk onto a disk I punched and the N* booted with it. The only issues I have so far is that the punched disks have much noise [after punching]. It sounds like chaffing inside the sleeve. I let Dwight know about this today.
Update Nov 2012 by Herb: Kelly Leavitt responded to me in Nov 2 2012, to a comment to him about his use of the Elvey punch and raised holes: "I've only tried two [hand-punched] diskettes in my Northstar. They worked, without problems. I did punch the holes twice (once from each side) to reduce the raised area. I'll have to try some more and do a few more tests." - Kelly
Update 2016 by Herb: From what I heard from others who used the disk punch, and from my own use of disks sent to me; the punched holes often have a raised edge. That raised edge makes noise, and may impact diskette rotation speed, may make debris from abrasion. As of 2016, there are discussions about some improved version of a disk punch. These were, and may be, produced by a machinist, so that appears to limit progress. I'm just not comfortable discussing that work - I'm not involved, there are issues of use, it's someone else's product. A diligent Web searchwill likely find further information and current status from those involved. - Herb Johnson
May 15, Nick says: I inserted one of your real hard sectored disks and tried reading it with Dunfield's NST resident in N* RAM,(and PC w/ serial to talk to it). I was able to read back one of your disks perfect! Therefore I believe my controller and drives don't have a problem. [Details of working with his Elvey-punch produced disks were discussed.]
By the way, the N* CPM disks you sent me, do you know the configuration? Is it 64K RAM? Is motherboard IO jumpers at default etc? I tried booting off it and didn't see anything on the serial console.
June 4th, Herb asks:
So, what was the final outcome of the Northstar CP/M diskettes I sent you? Did any boot up? Were you able to make your own bootables? Were you able to format the disks I sent, or the disks you punched yourself? you don't have to send back my diskettes, but if you can make disks yourself it would be nice if you could send me some "new" diskettes.
I sold one of the Northstar systems I worked on, at the VCF-E show. Here's a Web link about the show and you can see the Northstar system I sold.
Nick replies:
My last results were that I could read your disks, however not write to any. Dave Dunfield and I had a discussion and he suggested that the write circuitry on the disk controller may have a problem. I hooked up a scope and it didn't look like a TTL shift register was working correctly. I'm going to continue debugging this this week and will let you know the results.
I just got a step further. I can write and read back to your disk successfully now. I used a scope to probe the parallel to serial shift 74166 IC at 6D. The data was not coming out serially on pin 13. I ordered a replacement IC off Ebay and replaced the IC. The card appears to work now. I had a logic analyzer read to go, but eventually the parallel data was converted to a serial signal. Not much use because it didn't have any sort of protocol analyzer built in. Thanks go to the analog scope.
[So now] I can at last read, write and boot your disks.Here's a photo of the boot screen. Now I'm off to trying my own hard punched ones! Thanks!
LAter he added:
I validated that I could use Dwight's punch on my disks and the N* would read and write them fine. I was able to copy your LifeBoat CPM disk onto a disk I punched and the N* booted with it. The only issues I have so far is that the punched disks have much noise [after punching]. It sounds like chaffing inside the sleeve. I let Dwight know about this today. [See notes above about the Dwight Elvey punch. - herb.]
Nick updated me June 9th: I connected the SA400 in question to my Imagedisk PC and tested it with a soft-sector floppy. I was puzzled why it wouldn't read the tracks the other drive wrote. I tried advancing the head track by track and watched the assembly. The head assembly was stuck! I moved the head to track 0 by hand, then stepped it again and it worked. I was then able to read data back from the disk. The drive turned out to ok after all. It must have been in between tracks or maybe when I formatted the head never moved. I used it as drive 2 in the N* chassis and was able to use COPY.COM to copy your hard sectored boot disk over. I was also able to copy to a "punched" disk.
[I advised Nick that drives need cleaning and lubrication of the rails and threaded drive shaft of the read/write head. This is the kind of work I do when I repair floppy drives. - Herb]
herb: Why would you say the Tandon TM100 you were sent "won't match the existing drives"? I'll grant it will look a little different but it will be black and full height. If his drive is working and the price is right, I suggest you take it. YOu can still develop on it, whatever the "look".
Nick: Correct. It should be a form factor fit and people indicate the face should be nearly the same. He already sent it in the mail. I think both my SA400s work, so I can use this on the IMSAI system with perhaps the CCS boards.
If I can get the hard-sector punch to work reliably I'll probably leave the N* as it was manufactured. The CCS boards may end up in the IMSAI (with 5.25" cabinet if I can find one) or perhaps I'll just leave the IMSAI original configuration with the Processor Technology aka SOL board set (and VDM-1). If I can't punch hard-sector disks reliably then the CCS boards will likely end up in the N* chassis. [So] I think I'm *ok* with the 5.25" drives for now. This was very good information and sent me in the right direction. - Nick
Nick also asked about wood covers:
By the way. Do you know what type of wood the NS cover is? My NS doesn't have a cover so I'll need to build one. Thanks!
Herb replies: The covers I've seen are plywood with some kind of veneer. My guess would be walnut stained dark. The weak part of the cover is the narrow slot inside around the front, where the front panel fits. The slot is close to the front edge, so the plywood tends to break off. If I built a cover, I'd reinforce that edge in some way.
