SA801 bringing down to Shugart 34 pin interface issue

[ldkraemer]

The TRS-80 Model running Newdos 80 will handle 8" floppy drives. The PDRIVE command
just has to be correct.

Larry

 

[Eudimorphodon]

The TRS-80 Model running Newdos 80 will handle 8" floppy drives. The PDRIVE command
just has to be correct.

I feel like you're forgetting this part of the Newdos 80 manual:

Compare this to the fragment you posted: the 8" drives are type B, D, F, and H. Explicitly not supported by either the bare Model I disk controller or a Percom doubler.

 

[ldkraemer]

Yes, I did miss that.

THANKS.

Larry

 

[Eudimorphodon]

@Eudimorphodon - I didn't forget you, but I'm late for a doctor appt. But I'm aware, I'll let you know how I'm addressing that .

Still curious, I guess. Are you trying to set up PDRIVE so it's treating the 8" floppy like it's just a 5.25" floppy spinning at 360RPM instead of 300RPM? Not sure if that's going to work too well...

 

[Tuc]

Hi @ldkraemer , @Eudimorphodon and all

Bit of a delay due to family event this weekend.

I had worked out a whole reply for @Eudimorphodon and then ran into a glitch that totally invalidated it. So addressing the rotational speed difference (300 vs 360). I was under the impression from other posts elsewhere that it could be compensated for just running at a higher CPU speed. I installed a Sprinter II type board from Ian in AU. I also ordered 120ns 4116's to help ensure I could maintain the higher CPU reliably. I later found this wasn't totally true, but I'll come back to it.

Given that, I was initially just looking for the head to move to at least prove that the wiring was partially working. The NewDos/80 2.5 PDRIVE was straight out of the book. Couldn't understand why it didn't work until I started to pull apart the options. They assume if you're trying to use 8" drives, you have a Mapper II. I do not. (The errors I was getting was it saying I had incompatible controller types. ) So, I changed to my controller. Invalid. So RTFM, turns out they won't allow my controller to support 8" drives. So for the moment I'm at an impass.

I say the moment because of two things. First, given the realization that just the speedup wouldn't work, I was incredibly lucky to find (again from Ian in AU) an LNW Switch 5/8 as @Eudimorphodon quoted from. So why is there an issue? It was a kit. I'm terrible at soldering so sent it out to get put together. So that's a blocker. The 2nd thing is I think ND/80 2.5 is making my life harder than it needs to be. It's being a pain trying to work with certain HDs, where LDOS on a Model 4P make quick job of it. So, I'm starting to learn more about LDOS and see if I can make progress.

So that's where things stand. I'll probably be shuttling relatives around and to airports after today so it might delay me a bit.

But hopefully I'll see life.

Thanks!

 

[Eudimorphodon]

So addressing the rotational speed difference (300 vs 360). I was under the impression from other posts elsewhere that it could be compensated for just running at a higher CPU speed. I installed a Sprinter II type board from Ian in AU. I also ordered 120ns 4116's to help ensure I could maintain the higher CPU reliably. I later found this wasn't totally true, but I'll come back to it.

Yeah, that’s a big bum steer you got there.

The reason CPU speed upgrades are mentioned in relation to 8” disks on the TRS-80 is because these machines don’t use DMA to transfer bytes as they stream towards or off the disk controller, it has to use software PIO. And a 1.77mhz Z80 is *barely* fast enough to do that at the speeds either a single-density 8” or a double-density 5.25” disk runs at. (250khz bitrate) *Double Density* 8” drive operation, which runs at 500kbps, thus requires a speed-up board.

To be extra clear, the speed of the CPU has no direct effect on the bitstream to/from the actual drive, that’s independently driven by the oscillators in the disk controller hardware. (Which as I mentioned, are not set up to drive 8” disks unless you have one of those stated mods.) It’s just about being able to keep up with the controller.

Anyway, the rotation of the drive is kind of a red herring in this discussion; effectively it just means that at a given bitrate the faster rotating drive will store less per track because the track is effectively “shorter”. (But the faster disk might be more reliable because the bit written to the medium will be “wider”.) You might be able to compensate for that with PDRIVE, maybe, by specifying fewer sectors per track? (IE, my vague recollection is that a single density Model I disk has 10 sectors; if you took a drive and tuned its speed way outta whack up to 360 RPM maaaaybe you could tell PDRIVE you had an 8 sector disk and it would handle it?) Not sure, though, if the drivers will whine about the index pulses being too close together, or would erroneously add wider inter-sector spacing? But it would just be a software problem in this case.

… but that above is referring to a faster spinning 5.25” drive. I’m honestly not sure if a 8” drive would be able to handle the mini floppy data rates successfully. Drives have pulse recovery goo between the read head and the to-the-computer data lines and they might not handle a too-slow-by-half data stream correctly, depending on the drive? Anyway. There would be no point to this even if it worked, because the resulting disk would hold *less* data than a mini floppy and you wouldn’t be able to read it in another system that has the correct controller.

 

[Tuc]

@ldkraemer / @Eudimorphodon / et al,

So good and bad news.

Good News - I think with the help of some Arduino people I fixed the issue with my disk exerciser. Turns out I wasn't pulling Drive Select low. I've tested with a Gotek, I tested with a 5.25 and both were fine.
Bad News - The parts for the disk controller that could do 8" drive were incomplete. I'm trying to get them to the person working it and see if they can finish.
Bad News - I didn't fully realize it needed extra equipment once the boards done to get it adjusted. I was warned it needed adjustments, but didn't realize it needed -
* Digital Voltmeter 1% accuracy > 1megohm input impedance
* Frequency counter .1% accurany > 1k ohm input >5mhz
* Osilloscope > 15mhz bandwidth, triggered.
which I don't have any of. I'll need someone with a little Model I knowledge (I can send CPU if necessary) to align the board.
Bad news - I tried using the Arduino on the 8" drive now, and it does head load when selected, but the stepper isn't moving. Not sure why not, this disk was aligned/tested by someone and this is the first I'm using it. I'm just sending a DS, Direction, then step. Is there something I might be missing?


So any direction on the 8" disk and someone with those tools would be helpful.

Thanks, Tuc
Thanks

 

[ldkraemer]

Tuc,
Nope, Sorry I don't have any real knowledge on 8" Floppy Drives. I've never seen one up close.
If someone aligned it, then it should be working. I'd think if you had a 50 Pin cable to plug into it it would work.

Larry

 

[Chuck(G)]

Bad news - I tried using the Arduino on the 8" drive now, and it does head load when selected, but the stepper isn't moving. Not sure why not, this disk was aligned/tested by someone and this is the first I'm using it. I'm just sending a DS, Direction, then step. Is there something I might be missing?

I don't think so--how wide is your STEP pulse? Are all drive inputs terminated? Having the door closed and the head loaded with the drive selected should be sufficient. Some 8" drives don't even require the heads to be loaded.