forums.nabu.ca

As requested by klyball, brijohn, and others, I'm putting up some photos of the Hard Drive Interface for reverse engineering purposes.

Do you need picks of the MSU controller as well? or just the cable? (which is, as far as I know is 1:1 on wiring)

If you need any others or closeups, just ask.

lol love how simple it is - since it's really only an adapter from the bus to wd1001. I think they're gonna need pics of the underside near the connector to see where the traces go on the plug.

oh and here's the pinout on the back of the card - these simply connect on the other end directly to the wd1001 in the drive unit.

Here's the pinout for the wd1001 (not the connector of the drive unit). This is to show the connector of the hard drive adapter card has pins that connect directly to the wd1001.

WD1001 Datasheet/Manual for reference:
http://www.bitsavers.org/pdf/westernDig ... _Feb83.pdf

Greg

DJ Sures wrote: ↑Sun May 07, 2023 8:32 pm lol love how simple it is - since it's really only an adapter from the bus to wd1001. I think they're gonna need pics of the underside near the connector to see where the traces go on the plug.

Capture.PNG

Ok, does this get the connector well enough?

Hello, new to the Nabu but was dazzled when I found this piece of quirky vintage hardware had become available. Mine's still in the mail but that hasn't stopped me from researching it.

I'd love to be able to connect mine to an actual vintage MFM drive. I think the trend with a lot of guys is to try and circumvent this step in lieu of a soulless SD card (harsh!) due to the availability of working drives and the question of reliability, so I'm not surprised to see that there hasn't been a lot of action on reverse engineering the WD1001 and host adapter. That host adapter looks like a cake walk but the WD1001 itself is a potential handful.

It almost seems like the better solution would be to develop a host adapter to give the Nabu a compliant 8 bit ISA slot where a PC compatible MFM controller card with it's own onboard BIOS could be used. When combined that's basically all that proprietary Nabu host adapter and WD controller card seem to amount to.

Another option worth examining might be David Gesswein's MFM emulator board (not affiliated.) While designed to emulate an MFM drive, it's also an effective stand-alone disk controller board which can read and archive various and unknown drives through an embedded Linux distro. I might be flogging a dead horse here, and while David's MFM board is certainly not unknown in retro computing circles, employing this board would certainly require some coding on the linux side and would almost certainly have to interface with the Nabu via it's available USB or Ethernet port. That fact alone makes this pretty haphazard solution for the limited end result, but hey, bragging rights for having the Nabu with the most authentic/loudest computing experience.

HiFiasco wrote: ↑Tue May 09, 2023 4:34 amI'd love to be able to connect mine to an actual vintage MFM drive. I think the trend with a lot of guys is to try and circumvent this step in lieu of a soulless SD card (harsh!) due to the availability of working drives and the question of reliability, so I'm not surprised to see that there hasn't been a lot of action on reverse engineering the WD1001 and host adapter.

After suffering decades of painfully slow and tragically unreliable storage media (forever losing programs that took months to create), I just can't get that "storage nostalgia" that some people have. I have nostalgia for the computers themselves, but don't really care how data gets on to those computers - as long as it's reliable.

I started with tape - and I still cringe when remembering the dreaded "?IO ERROR" (usually followed by lots of loud swearing)

HiFiasco wrote: ↑Tue May 09, 2023 4:34 am but hey, bragging rights for having the Nabu with the most authentic/loudest computing experience.

For that, you'd have no drives and use the HCCA (network adapter) You can find out more about Nabu, and the intended purpose of being a client on a worldwide network at https://nabu.ca

1) Other than the development environment at Nabu HQ, customers did not have hard drives. No software was distributed on hard drives. A hard drive on the NABU is neat but not organic to the original configuration.

2) The floppy drive was an after-thought "fix/hack" because the cable companies could not be persuaded to upgrade their infrastructure to allow bi-directional communication. So they shipped a few floppy at the end of the days of NABU. The floppies were mainly used for Zork. You could spend extra and get a copy of CPM to run Wordstar. But that was rarely used by this point in the computer's lifecycle. The idea of the floppy drive barely aligned with the target customer demographic, who were not computer literate.

3) Many MFM emulators can be acquired without needing modification because they're manufactured products by companies. As you mentioned, the MFM is not an issue; it would be emulating the WD1001. I don't think emulating the WD1001 is as tricky as people think; it just takes time for someone to do it. The WD1001 documentation is well-defined. I considered going this route for my NABU 1600 (not the PC), but summer is here, and I won't spend much time inside. That would be a project for the fall

Thank you all for your replies. I decided to try and undertake replicating the SASI host adapter based on the photos that Leo posted.

Searching for a WD1001 turned up nothing, but recently a lot of purportedly new/open box Xebec s1410 SASI to MFM interface adapters recently became available. It appears that the board is simply Xebec's version of the interface the WD1001 provides, and also seems to be very well documented. I don't see why this board shouldn't serve as a suitable substitution. Like many of the early PC or XT HDD controllers, the board has a preprogrammed set of drive geometries for some of the more common MFM disk units available at the time. I have a handful of working examples of these drives so I should be able to make something work, and I don't see a reason why an MFM emulator shouldn't also work with it. Xebec documentation may be found here: http://matejhorvat.si/sl/slorac/delta/partner/s1410.pdf

I've attached an in progress version of the schematic for the Nabu portion of the host adapter, but I'm going to need some help to finish it as some of the traces vanish beneath ICs which are pretty much what's not drawn on the schematic from the P1/P3 interface connector, and the resistor shown next to U5 in the photos. Capacitance values shown are "acceptable" values in lieu of not knowing actual values.