Print

[mactjaap]

I have a Lisa 2/10 with a somehow working widget. Not in best state, so most time it is not in the machine.
For these occasions I can choose for an X/Profile or Profile disk.

When I use the Profile disk (attached to the Lisa with a correct extension card and parallel cable) the disks start up with the ready light on. When I detach the cable the Profile does it normal boot procedure, stepper motor will do its work and ready light will be flashing until this procedure is ready. Only then the ready light is always on.

So for now I switch on the disk and 2 seconds later the Lisa. There will be a boot error but finally the disk check starts and after it is finished it can click on the parallel card icon and boot the Lisa. I made a short YouTube video to show it:

https://youtu.be/cjdD-nR7AYs

(a little bit edited..fits almost in one minute)

My question is. What is happening here? Is this normal? Can I do something so the Profile can do its normal procedure?

[stepleton]

Not sure, so let's try eliminating factors

Does it do this on both parallel ports?
Do you have a different parallel cable you can try? (How long is the cable?)
If you have another ProFile, does it behave normally using the same parallel cable on the same port?

[mactjaap]

Yes! That's the way to go:

Does it do this on both parallel ports?
> Same behavior on the upper port, but something worse.... no boot into LOS 3.1 but an error; 10741

Do you have a different parallel cable you can try? (How long is the cable?)
> I have only one, bought at VintageMicros. Must be good.
>The cable is 90 cm long (~35 inch) and is missing the ping number 7 in the upper row ( as it should).


If you have another ProFile, does it behave normally using the same parallel cable on the same port?
> No I have just one.....

Another test:
I  booted with the X/Profile attached. Then you notice it less. I will boot LOS 3.1 from the X/Profile and the Profile will be seen later.

Could it be something with an Lisa 2/10 and a external Profile only?

[rayarachelian]

Does it do this on both parallel ports?
> Same behavior on the upper port, but something worse.... no boot into LOS 3.1 but an error; 10741

This is normal behavior, either there's something in a configuration file or the boot loader blocks are different for drives attached to expansion slot ports. This is a known limit of LOS.
The boot block (AAAA) knows what device to boot from, but after loading the boot loader blocks (BBBB),

I think what happens is that the boot loader remembers that it was supposed to boot from the motherboard port and so it fails. Either that or there's some config file similar to config.sys on DOS that has hardcoded paths to get the the rest of the OS from, and that has the parallel port hardcoded to it (just my theory).

So basically this is "working as intended" even though you get that error. (You'd get the same if you try clone a Widget to a ProFile, even on the same port.)

Do you have a different parallel cable you can try? (How long is the cable?)
> I have only one, bought at VintageMicros. Must be good.
>The cable is 90 cm long (~35 inch) and is missing the ping number 7 in the upper row ( as it should).

Yup, that pin is supposed to be missing, you'll generally find that it's been plugged up on the female side of the DB25, but if you force a normal straight through DB25M plug, that filler will be pushed in by the pin and allow a connection, or it will fall out later, or it will bend that pin.

That missing pin was just a lame way to keep you from attaching a serial device to a parallel port.

If you have another ProFile, does it behave normally using the same parallel cable on the same port?
> No I have just one.....

Another test:
I  booted with the X/Profile attached. Then you notice it less. I will boot LOS 3.1 from the X/Profile and the Profile will be seen later.

Could it be something with an Lisa 2/10 and a external Profile only?

Wonder if it has to do with the CMD signal being enabled/disabled, I suspect that READY light is just mirroring the BSY signal.
There's another pin that the Lisa uses to detect a ProFile disk, and that's the OCD - Open Cable Detect pin. Not sure if that's of any help though.

So to clarify something, I know you said you have a Widget in that 2/10 that's marginal. Is that external profile hooked up to a Dual Parallel Port? Or are you rerouting the internal widget cable to the outside with one of these: http://vintagemicros.com/catalog/external-widget-connector-p-92.html

So as far as I'm aware, it shouldn't matter whether you have a 2/10 or a 2/5 - the boot ROM will set the VIA pins the same way, however, if you're using the Dual Parallel Port card, this might behave differently depending on the code on that card - there's both a status and boot routine in that ROM - the ROM is about 2K or so.It might be initializing one of the pins differently than the Boot ROM would, which would explain the light going out. But I'm not sure if that's actually true or not.

Actually, no, because at that point, you just powered the Lisa on, and it hasn't yet started testing the expansion slot cards, so it hasn't downloaded the test program routine from the expansion slot ROM, this is an electrical signal on the pins of the parallel port that the ProFile is reacting to. Most likely the CRES line (Controller Reset).

If I have time this weekend (and remember to), I'll try to test this by attaching one of my ProFiles to the dual parallel card and see what happens.
If you're using the Dual Parallel Port card, can you mention which port? Upper or lower you're using? I don't expect they'd behave differently, but you never know.

What I do see in your video is that the ProFile does its normal power on test which preens the surface for bad blocks. And then you turn on the Lisa. Immediately when you do that, the light on the ProFile goes out, and then it seems to do the same test again, which now since it is busy, times out the boot attempt from the Lisa since the drive isn't ready yet. But after the test is done, you can boot off the drive and it behaves normally.

So what I suspect, and I'm likely wrong is that when you power the Lisa on, the RESET signal from the CPU is sent to the expansion slot, which is reflected by the CRES (Controller Reset) signal on the parallel ports - this resets the ProFile and then it goes through it's self test. If that's what's happening, that's likely normal behavior.

If you want to test this immediately, you could go into service mode and write a small program to just execute RESET and RTS, and that might cause the same behavior (force the ProFile into a self-test.)
Incase you want to try this:

Code: [Select]

00000000                             1  ;------------------------------------------------------------------------------------
00000000                             2  ; RESET test
00000000                             3  ;------------------------------------------------------------------------------------
00000000                             4 
00000000                             5 
00020000                             6    ORG     $20000     ; starting address note spaces at start
00020000                             7 
00020000                             8  START:
00020000  4E70                       9    RESET              ; RESET I/O devices
00020002  4EF9 00FE0084             10    JMP $00fe0084      ; return to monitor (RTS might also work)
00020008                            11 
00020008     

So you'd do this:

• Power on your ProFile, let it finish the test until ready.
• Power on your Lisa as usual.
• do not insert a floppy but try to boot off the floppy.
• When it beeps with an error asking for the floppy, hit Apple-S to enter service mode.
• Select "2 Set Memory"
• Enter address 20000 (note 4 zeros)
• type in 4E 70 4E F9 00 FE 00 84 and hit enter.
• Select "3 Call Program" and type in address: 20000

Watch to see if the ProFile goes through a self-test again, the Lisa might return an error or not, but that doesn't matter, as soon as you hit enter after typing in the address in step 8, look at the READY light and see if it goes out and starts a self-test again.
If it does, I suspect everything is working normally, even if it might be unexpected. I'm pretty sure I recall the Dual Port parallel slots are wired somewhat differently than the motherboard parallel port.

It might also do the same thing if you go in the back and press the RESET button (or not).

[stepleton]

I don't think the LOS error is anything to worry about --- basically, if you install LOS with the drive on a particular port, then it will expect to start up on the same port forever. If you plug the drive into a different port and attempt to boot, you get an error. I don't remember if the error number is 10741, but it's definitely some five-digit thing.

The cable is probably fine, but a shorter cable would be a nice thing to try if we had the chance. No big loss if we can't, though.

At this point I'm not so sure. I feel like I've encountered this once before, but it must not have been as dependable as this or I would have investigated more deeply. I might try to do some experiments later with some real ProFiles that I haven't powered on in a long time.

I would be a little surprised if it were 2/10 specific. There is not much that a drive connected to a parallel port card can do to tell whether it's connected to a Lisa 1, a 2, or a 2/10, so far as I know. The parallel port card emits the same signals regardless. Also, the code that the Lisa will use to boot a ProFile on the parallel port card will be the same on any Lisa --- that's because the code lives on the parallel port card.

It's a 2/10 so there's unlikely to be any trace damage on the motherboard, but just in case, do you see the same behaviour if the card is in a different expansion slot? (Whether you do or not, you're sure to get a 10741 error or similar if you try to boot the Office System.)

If that doesn't provide any clues, then the answer may lie in the ProFile. Maybe one of the interface ICs on the controller board is not so healthy, or maybe a pullup resistor on one of the signal lines is not working to spec. (That would be a strange failure, though!)

ETA: this is a reply to mactjaap's post --- that's an interesting experiment from Ray!

[mactjaap]

Thank you very much for these answers. Even if we don't solve this completely I have learned a lot more!
Like, that if you install LOS with the ProFile connected to the lower parallel port, you cannot switch to the upper port with reinstalling the OS (error 10741)

To answer some questions.
@rayarachelian

>So to clarify something, I know you said you have a Widget in that 2/10 that's marginal. Is that external profile hooked up to a Dual Parallel Port?

Yes, it is hooked on an external parallel port. Connected in Slot 1 ( if you look at the back of the Lisa, far left). I use the lower port on the card.

About this:
>What I do see in your video is that the ProFile does its normal power on test which preens the surface for bad blocks. And then you turn on the Lisa.

Maybe I don't understand what this is:
"...the ProFile does its normal power on test which preens the surface for bad blocks."
As what I see the ProFile doesn't do its normal routine. If I turn the Profile on (with the cable connected and the Lisa OFF)... it will only turn on it ready light. It will never do its 'stepping motor stuff' with a flashing led. It will not do anything else. Therefore .. after a while I turn on the Lisa. Only then the Profile will do its test, but stops when the 'boot from' screen appears. If you then wait for a minute or so it will start testing and successfully give the red ready light.

About this:
>So what I suspect, and I'm likely wrong is that when you power the Lisa on, the RESET signal from the CPU is sent to the expansion slot, which is reflected by the CRES (Controller Reset) signal on the parallel ports - this resets the ProFile and then it goes through it's self test. If that's what's happening, that's likely normal behavior.

Yes, that sound exactly what I see! I switch on the Lisa, Profile red light will go off and the stepper motor begins.

About this:
>If you want to test this immediately, you could go into service mode and write a small program to just execute RESET and RTS

Very nice! I understand correct that I can enter the code listing direct in service mode? This question because I don't know to write a program for Lisa. Would be nice to learn. Absolute, but I'm afraid that will be hard for me.

@stepleton

Thanks for clarifying that there will be no difference in behaviour with Lisa 1, a 2, or a 2/10.

I will test the parallel card in a different slot. If it works properly I will get the error 10741 again

@rayarachelian + @stepleton
I will do you proposed test this weekend and will get back with the results in this post.

[compu_85]

I notice the same thing on my 2/10, when hooking a Profile up to a parallel card. The self test on the Widget takes long enough that the profile is ready by the time the OS starts to boot. If I want to boot from the Profile I just hit the spacebar and wait.

I am using a cable with all 25 pins.

[rayarachelian]

Maybe I don't understand what this is:
"...the ProFile does its normal power on test which preens the surface for bad blocks."
As what I see the ProFile doesn't do its normal routine. If I turn the Profile on (with the cable connected and the Lisa OFF)... it will only turn on it ready light. It will never do its 'stepping motor stuff' with a flashing led. It will not do anything else. Therefore .. after a while I turn on the Lisa. Only then the Profile will do its test, but stops when the 'boot from' screen appears. If you then wait for a minute or so it will start testing and successfully give the red ready light.

So, what I'm saying is that when you power on the ProFile drive, it goes through it's self test of reading from the surface of the disk to test for bad blocks to spare. This is its normal behavior at power on.

By the time you're ready to turn on the Lisa, the ProFile has finished this step.

However, when you turn the Lisa on, this sends a RESET signal for something like 100ms from the power on circuitry all over the Lisa's I/O devices, and the CPU, and also to the ProFile drive, and this causes it to reset, and restart its self test again, however since the Lisa finishes its own self-test before the ProFile does, the ProFile is now busy and can't be booted from.

If you look on page 48 of http://bitsavers.org/pdf/apple/lisa/hardware/Lisa_Hardware_Manual_Sep82.pdf you'll see that pin #21 is CRES - or controller reset.

If you then look at http://bitsavers.org/pdf/apple/lisa/hardware/050-4027-A_Parallel.pdf you'll see a signal from !RESET going from location C4 pin 7 to reset the VIA at U3D, this line then goes up a bit to U4E, to the right and down to location A2, through resistor pack U5E and out to the port on pin 21 labeled !LCRES - this is the same as on the motherboard and is the controller reset line for the lower port.

On the next page, you'll see something similar for the upper port.

So the power on signal is what causes the ProFile to reboot itself even though you've already powered it on and it's already done one self test.

So maybe the fix is to power on the ProFile at the same time as the Lisa, and then use the space bar or mouse to interrupt the boot from the Widget and this would prevent the 2nd reset.

[mactjaap]

So, what I'm saying is that when you power on the ProFile drive, it goes through it's self test of reading from the surface of the disk to test for bad blocks to spare. This is its normal behavior at power on.

No. There is no test at all. Nothing happens. Only a red light. No stepping motor like normal without  cable.

[rayarachelian]

Actually found something: https://apple3.org/Documents/Manuals/Profile_Level_2_Manual.pdf page 157:

"CRES - generated by the host - this is used to reset the Z8 back to its initial state. This usually begins at the start of communications, but whenever it happens, it depends on the host."

So there's also an output on the VIA outside the RESET signal that can trigger CRES, but CRES isn't used to reset the ProFile, it's to put the communications state machine back to the first listening state.

[mactjaap]

Thanks Ray, for looking so deep in this problem. Even if it is not possible to change this I really learn a lot of this!

I was planning to do your test last evening, but I ran into a stupid problem. The 2 on my keyboard refused to work. I managed to work around it last weeks, but if I have to type in commands I really need the two on my keyboard ( and I cannot use the 2 on my numpad...It will say ...wait... and nothing happens).
So today I will try to fix my keyboard.

[mactjaap]

Keyboard is fixed. 2 is working again!

About his service mode test...should I use space between the commands?

"type in 4E 70 4E F9 00 FE 00 84 and hit enter."


So like this:
4E 70 4E F9 00 FE 00 84
or like this:
4E704EF900FE0084

And..... an enter after?      Select "3 Call Program" and type in address: 20000   

[rayarachelian]

From what I remember you can enter spaces in between or not. It accepts it both ways. Generally I group it in two bytes, or a single byte. i.e. 4E 70 or 4E70

[mactjaap]

I did this test in service mode:

So you'd do this:

• Power on your ProFile, let it finish the test until ready.
• Power on your Lisa as usual.
• do not insert a floppy but try to boot off the floppy.
• When it beeps with an error asking for the floppy, hit Apple-S to enter service mode.
• Select "2 Set Memory"
• Enter address 20000 (note 4 zeros)
• type in 4E 70 4E F9 00 FE 00 84 and hit enter.
• Select "3 Call Program" and type in address: 20000 ( and hit enter..)

(I typed it in without spaces 4E704EF900FE0084, but tested also with spaces. Same result.)

It clearly make some reset. The video gets blurry and something is happening. But the Profile disk just stays running without a reset or disk scan.

I made a small movie:

https://youtu.be/HZepVikQ4UM

[rayarachelian]

love the Spongebob refs :-D - why didn't the ProFile do its self test on power on before the Lisa was powered on? that's weird? Something's not right there.

Yeah, sadly the CRES line doesn't mean reboot to the ProFile, it means go back to state zero and listen for commands as posted last night in reply #9.

Also note that there are now Mac(128/512/KE)2Lisa, ADB2Lisa and USB2Lisa keyboard interfaces now.
i.e. https://www.ebay.com/itm/ArcaneByte-adb2lisa-Apple-Desktop-Bus-ADB-to-Apple-Lisa-MacXL-Keyboard-Adapter/203234120754