5 MB Profile on the Apple II

[gary286]

I aquired a 5 MB Profile drive about 10 years ago that I intended to save for an Apple 3 but ended up getting priced out of that interest very quickly. When powered it looked like it was having trouble finding some tracks as the ready light hangs a few times on startup. Eventually, I found the Apple II card for it and threw it in my II+. To my surprise the drive works fine from what I can tell, it had Prodos on it already along with some Appleworks files. My understanding is there are not many utilities for this drive for the Apple II. Do the heads need to be parked on this drive, and can I do a low level format on it on the II+ or e?

 

[stepleton]

I've never seen a ProFile manual reference parking the heads on a ProFile. The disk itself is an ST-506 mechanism, so you could seek advice on where owners of those drives liked to park the drive heads before turning off their computers. You could then have a program running on the Apple command the ProFile to read a block that is likely to be located on an inner track of the disk. (You can't be certain if it's there because of sparing.) There is however no command that the Apple can ordinarily issue to a ProFile that specifically means "park the heads".

I'm also not aware of an Apple II program that will low-level format a ProFile. The programs I know are either Apple /// or Apple Lisa programs, though it would be possible to write an A2 program that did it. Note that you will need to replace the ROM (and potentially the Z8 microcontroller) inside the drive with a special debug variant in order to do an LLF. Once that's done, basically all the program does is send a command to the ProFile that says "hey, go format yourself". The computer does very little of the work.

Note that you can attach an inexpensive ESProFile device to a ProFile in "diagnostic mode" --- this device can tell a ProFile to format itself, and can also back up all the data on the drive ahead of time.

(ETA for picky experts: I guess you can be reasonably certain that a block is located on an inner track if you read the spare table on block $FFFFFF.)

 

[gary286]

Thank you, stepleton. That ESP device method is what I'll aim for. I'll try the standard reformat first to see if that will clear the startup hang. I probably should change out the filter capacitor and then image the drive before messing with anything, though.

 

[stepleton]

You are probably right that the pause during the self-test owes to some bad blocks. Use the ESProFile's diagnostic mode to query the drive's $FFFFFF block and find out how many blocks have been spared --- if it's just a few, you might consider not doing too much. Perhaps even just imaging the drive and then rewriting the image to the disk would be sufficient, particularly if you continued to back it up from time to time. Per its service documentation, Apple considered a drive serviceable even if it had a few (but only a few) bad blocks.

 

[gary286]

It took me a while to get back to the Profile but I got it fully working. Prodos did indeed find a few bad blocks on the drive, and the Prodos format is just a quick format. So I imaged the drive with Adtpro to my Linux machine. I zeroed a copy of the image with the Linux disk utility and then wrote it back to the Profile. That cleared up the startup hang and the drive ran for a good hour now without issue. All set up proper now with a Monitor III and stand. I never really see Apple II's with this complete setup.

 

[Wayne Stewart]

I’ve occasionally used a Profile on the Apple IIe. Nowadays I have one spot for Apple II computer use and I keep one set up for a while and occasionally change computers, mostly the IIe and IIgs. I also don’t have a static setup anymore and swap out cards and what not regularly so the Profile ends up sitting beside the IIe. Right now in the IIe I’m using a Rev C SCSI card with a SCSI enhancer though it still has a Vulcan HD in it but the card isn’t installed at the moment.

 

[resman]

It's that the Profile on an Apple II is so unbelievably slow. Like they took the 1541 approach to disk interfaces. How did they make a parallel connection to a hard drive so slow? Aesthetically it is a rare combination, so I keep it for that, but honestly a $35 a2pico with an SD card blows it away.

 

[olePigeon]

@gary286 Now add a DuoDrive for the ultimate in stacked Apple II goodness.

 

[stepleton]

How did they make a parallel connection to a hard drive so slow?

The ProFile port protocol is clocked by the computer and can shuffle bytes in or out at "up to 1MHz" per this resource; if it weren't for overhead like seeking and handshaking and all (not to mention somewhere to put the data), you could dump the entire disk in under ten seconds. Nothing goes that fast in real life, but the Lisa 2/10 does clock bytes at a pretty good clip. The slowness has to be on the computer side. (So perhaps it is like the 1541 in that respect...)

 

[resman]

The ProFile port protocol is clocked by the computer and can shuffle bytes in or out at "up to 1MHz" per this resource; if it weren't for overhead like seeking and handshaking and all (not to mention somewhere to put the data), you could dump the entire disk in under ten seconds. Nothing goes that fast in real life, but the Lisa 2/10 does clock bytes at a pretty good clip. The slowness has to be on the computer side. (So perhaps it is like the 1541 in that respect...)

The Profile on my Apple /// certainly feels more performant. The document you link suggest the Apple /// uses DMA. Perhaps the Apple II uses PIO for a much slower transfer rate? Oh well.

 

[djg]

The Profile on my Apple /// certainly feels more performant. The document you link suggest the Apple /// uses DMA. Perhaps the Apple II uses PIO for a much slower transfer rate? Oh well.

Can you change the interleave when formatting? Could be slower computer needs more interleave.

 

[resman]

Can you change the interleave when formatting? Could be slower computer needs more interleave.

From what I understand, the convoluted process to low level format a Profile only works on a Lisa or Apple ///. I couldn't find where the interleave can be changed, but I didn't look that hard. I use the Profile more as a visual aid to explain to the *kids* that this beast (5MB) is large enough to store only ONE song ;-)

 

[stepleton]

It's even worse: you need to swap the Z8 with its piggyback-or-mask ROM to a special "formatting Z8" with a different ROM (and also change a jumper) before you issue the format command from your Apple ///, your Lisa, your UsbWidEx, or (easiest to get your hands on) your ESProFile.

I don't believe that it is possible to achieve anything other than a 5:1 interleave for the ProFile. Checking the reverse-engineered code found on Bitsavers, it would seem to me that the interleave scheme is baked into the formatting-capable ROM.

 

[Istarian]

It's that the Profile on an Apple II is so unbelievably slow. Like they took the 1541 approach to disk interfaces. How did they make a parallel connection to a hard drive so slow? Aesthetically it is a rare combination, so I keep it for that, but honestly a $35 a2pico with an SD card blows it away.

The Pi Pico has a pair of ARM Cortex M0+ processors that can be clocked at up to 133 MHz and 264 KB of ram. So it runs way faster than the A2 itself and has more ram than most A2 systems. -- You can emulate the whole Apple II and some expansions cards with a Pi Pico.

On top of that, SD cards have an integrated controller and flash memory provides very fast random access reads+writes. There is no spin-up or seek time with flash memory.

So even if the Pi Pico were sluggish or inefficient somehow, it merely has to ask for the data it wants and catch what the SD card sends back.

It would be absurd if the modern tech didn't absolutely blow away any sort of contemporary Apple II hard drive going by transfer speed or storage space.

P.S.

You could probably have an SD card "hard disk" with a ROM and some interface logic where the Apple II itself does all or most of the work of talking to the card.

SPI is kind of timing sensitive though, so it makes sense to offload that interaction to some kind of processor/controller with a hardware SPI interface.

That's essentially how the Dan II card works, I believe. the ATMega328P is just there to interface with the microSD cards and emulate a ProDOS block device.

 

[shirsch]

From what I understand, the convoluted process to low level format a Profile only works on a Lisa or Apple ///. I couldn't find where the interleave can be changed, but I didn't look that hard. I use the Profile more as a visual aid to explain to the *kids* that this beast (5MB) is large enough to store only ONE song ;-)

I don't recall if there's an option to specify interleave, but Patrick Schaefer produced a disassembly of the formatter ROM. A bit of hunting around in there may turn up the logic (or possibly the table) responsible. I'm not sure if Patrick wants this file propagating without his permission. I'd suggest dropping him a note.

 

[Istarian]

@resman Ironically you can store more than one song in 5 MB, just not as high bitrate uncompressed audio (ala LPCM, WAV files).

So called CD-quality audio takes up a lot of space compared to MP3 files, MIDI encoded music, or even regular audio cassette tape.

Would be interesting if anyone has a definite answer on exactly how much Apple II style audio (speaker clicks) could be stored in 5 MB.

 

[stepleton]

Would be interesting if anyone has a definite answer on exactly how much Apple II style audio (speaker clicks) could be stored in 5 MB.

This will depend on the bitrate you can achieve on an Apple II, and I can't really speak to that.

On a Lisa, you can actually do better than plain old speaker clicks because the Lisa has three bits of programmable volume modulation: this allows you to use the volume setting to track the envelope of the sound, allowing you to use the clicky bits much more efficiently. Read more about this here: https://codeberg.org/stepleton/lisavox/src/branch/main/lisavox.x68#L50

That's from my demo Lisavox, made earlier this year. Anyway, the best bitrate to use for Lisavox is 31,836 KHz. A 5 MB ProFile could in theory hold 5,242,880 bytes * 8 bits per byte / 31,836 bits per second = 1,317 seconds or around 22 minutes of audio if I am not mistaken. There is some overhead in the way Lisavox encodes so back it down to 20 minutes to be safe. The astute will notice that I am assuming 512 byte blocks, leaving 20 additional tag bytes per block unused, so add more time back if you are feeling really determined.

I would guess that the Apple II figure will be the same order of magnitude.