LisaList2

Advanced search  

News:

2019.06.07 fixed NChat for the "Curve" theme, will eventually move it to its own page and add it to the default theme as well. Other plugins are next. see post in the Meta board for details

Pages: 1 2 [3] 4 5 ... 10
 21 
 on: May 10, 2022, 09:37:09 pm 
Started by blusnowkitty - Last post by pintoguy
Yes, CopyIIMac, sorry for the confusion. And yes, Apple-Enter on startup. See screenshot pics.

I also thought about the FDD reader head, and whether a wider head would be better or worse. The cross-track width is probably similar for DD and HD (both 80 tracks per side), but it's the down-track reader length that changes, as HD floppies have twice the kbpi (kilobit per inch). My conclusion was that a SD drive with a longer reader would have a higher SN ratio, and hence be a better choice for scratched SD floppies. That would be a nice experiment to try, and it's on my list (I used an HD drive here on a Mac Classic). On the other side, BLU run on the SS/LD Lisa drive also threw errors.

Regarding IPA, since I'm a chemist by training, I always wanted to make my own opinion, since I know IPA is not a known solvent for most plastics. So I tried on dummy blank floppies from my collection (a few different brands), and then on program floppies of little rarity. I also checked a few youtube videos on the topic, and all used rubbing alcohol (aka IPA). So I guess my question to this board is: Would anyone have FIRST HAND data showing that IPA messes with the floppy, either the magnetic coating, or the plastic substrate ??

 22 
 on: May 10, 2022, 08:46:15 pm 
Started by blusnowkitty - Last post by rayarachelian
OK so I did some work on these floppies. I have three of them. One is an Install2 disk with a serial number starting with 7IN. The other two are Install1 and Install2 with a serial number starting with 8IN.

Excellent!

After IPA cleanup (yes IPA works very well for me - I have revived dozens of bad floppies using it), I was actually able to read the file "Install Omnis" from the 7IN-Install2 disk on my Lisa, and copied it on the Lisa 10MB widget HDD. Installation seems to have worked if done from the HD, but failed when done from the FDD (failure to display a window). Once installed, Omnis needs to be run in its own environment. However, when launching, it requires the "Master Disk", and none of the floppy disks get accepted.

So by own environment, do you mean press Apple-Enter on start up to get the Environments window and select it from there? That's a weird way to write an LOS app, but ok.  Glad the IPA worked, there's some danger of it damaging the media, but glad it worked!

I used DiskIIMac to make floppy copies, using the "sector mode". The "bit mode" did not work too well. After writing these diskettes, I was able to make diskcopy4.2 images, which I'm attaching.
Did you mean CopyIIMac?

Perhaps someone will be able to take it from here. Using fedit, I suspect that the the Install1 floppy is simply a Lisa 7/7 boot floppy, whereas the Install2 disks are the Omnis programs. Again, fedit tells me that IN7 is likely Omnis 3.2 whereas IN8 is version 3.3. Also 8IN-Install1 is the worst with about 40 out of the 80 tracks throwing errors, whereas In7-Install2 (one bad track #73) and IN8-Install2 (two bad tracks #51 and 79) are quite a bit better.

Finally, I was not able to successfully use BLU to read these disks. When checking them, it seems to get stuck on the first read error and requires reboot.

Surprised that BLU failed, but glad you got some of the bits off. If we get lucky in the future and find some more Omnis disks, maybe we can get a full set at some point! Awesome job! Hang on to the originals, never know when some other tool/technique will come up that might help get it further.

Obviously whatever media's worn out or wiped isn't going to be recoverable, but maybe there's other things that can be done, such as using a 1.44M Apple SuperDrive floppy to read them which has a narrow head. That might help (or hinder) depending on where the damage is, and whether or not the head is slightly off-center track-wise, and if you're lucky you might hit an area that hasn't been damaged, etc.

 23 
 on: May 10, 2022, 08:19:29 pm 
Started by sigma7 - Last post by rayarachelian
Many thanks for this excellent write up!

 24 
 on: May 10, 2022, 07:36:51 pm 
Started by blusnowkitty - Last post by pintoguy
OK so I did some work on these floppies. I have three of them. One is an Install2 disk with a serial number starting with 7IN. The other two are Install1 and Install2 with a serial number starting with 8IN.

After IPA cleanup (yes IPA works very well for me - I have revived dozens of bad floppies using it), I was actually able to read the file "Install Omnis" from the 7IN-Install2 disk on my Lisa, and copied it on the Lisa 10MB widget HDD. Installation seems to have worked if done from the HD, but failed when done from the FDD (failure to display a window). Once installed, Omnis needs to be run in its own environment. However, when launching, it requires the "Master Disk", and none of the floppy disks get accepted.

I used DiskIIMac to make floppy copies, using the "sector mode". The "bit mode" did not work too well. After writing these diskettes, I was able to make diskcopy4.2 images, which I'm attaching. Perhaps someone will be able to take it from here. Using fedit, I suspect that the the Install1 floppy is simply a Lisa 7/7 boot floppy, whereas the Install2 disks are the Omnis programs. Again, fedit tells me that IN7 is likely Omnis 3.2 whereas IN8 is version 3.3. Also 8IN-Install1 is the worst with about 40 out of the 80 tracks throwing errors, whereas In7-Install2 (one bad track #73) and IN8-Install2 (two bad tracks #51 and 79) are quite a bit better.

Finally, I was not able to successfully use BLU to read these disks. When checking them, it seems to get stuck on the first read error and requires reboot.

Thanks again for the help and great advice.

 25 
 on: May 10, 2022, 06:03:28 pm 
Started by dmark - Last post by sigma7
Anyhow, grateful for any pointers how to further troubleshoot this poor Lisa!

An attempt was made ....

Debugging Bus Errors: https://lisalist2.com/index.php/topic,266.0.html

 26 
 on: May 10, 2022, 06:01:27 pm 
Started by dmark - Last post by sigma7
Do you have advice where good test points for each voltage rail is?

I think this is still "current"  ::) ...

https://lisafaq.sunder.net/lisafaq-hw-pwr_checking_voltages.html

 27 
 on: May 10, 2022, 05:52:36 pm 
Started by sigma7 - Last post by sigma7
Here is a first go at troubleshooting bus errors... if it turns into something useful hopefully it will end up in the FAQ. -- corrections & suggestions requested!

Debugging Bus Errors caused by hardware issues v.2022-05-14-A

A bus error is generated if a timeout occurs when the CPU attempts to read or write to an address (memory or I/O).

There are circumstances when bus errors are expected, such as when checking to see if an expansion card is present in a slot. In these cases the software intercepts the bus error and carries on.

In the case where a bus error is not expected, it indicates a hardware or software fault, and so typically an error message is shown.

The bus timeout signal is generated by the 556 timer beside the CPU. It is re-started at the beginning of each bus cycle, so the timeout only occurs when a bus cycle does not terminate properly.

Bus cycles are terminated by an acknowledgement signal, which signifies to the CPU that the requested data is now available (for a read), or has been accepted by the hardware (for a write).

There are two acknowledgement signals, VPA and DTACK. VPA is used to provide compatibility with older 6800 peripheral timing and to do so, the CPU slows down the process of completing the bus cycle when it receives the VPA signal. DTACK is the 68000 signal that indicates the bus cycle can terminate more or less immediately.

In the Lisa hardware, VPA and DTACK are generated by different circuits depending on what hardware is accessed. ie. the I/O board generates DTACK or VPA for accesses to hardware on the I/O board, the CPU board generates DTACK or VMA for accesses to the various hardware control registers, MMU, and ROM on the CPU board as well as DTACK for slot RAM, and individual expansion slot cards generate their own DTACK or VPA.

So when a bus error occurs, the fault may be on any of these boards and so you may be in for some interesting troubleshooting.

Swapping boards with a working machine is helpful to isolate the problem board, but even having isolated a particular board, there is usually more than one source of VPA/DTACK to check.

If one knows the address that generated the bus error, that usually will narrow down the circuit substantially. While the Lisa is still operating from the CPU ROM, it will try to help provide the information needed...

Reviewing p25 of the Lisa Boot ROM Manual v1.3 Feb84, it indicates that when an exception (such as a bus error) occurs, the address is stored in the long word at $282. In the event this is not populated with the problem address that caused the bus error, one will need to use another technique to isolate the problem (see reply below).

The default memory map setup programmed into the MMU by the ROM looks like this:

    000000 - 1FFFFF* Slot Memory (a combination of both slots up to 2MB), with the video page using the highest 32K
                               *If there is less than 2MB, then the memory range will be smaller, eg. 1MB 000000 - 0FFFFF
                               Thanks to the magic of the MMU, the slot memory always starts at address 0 regardless of
                               which slot or how much memory is populated
                               A Lisa modified to support 4MB is special as the additional address line bypasses the MMU, but
                               roughly speaking the slot memory is mapped up to 3FFFFF

    200000*- FBFFFF This is unmapped space, ie. "nothing" is mapped in this address range, and so a bus error will occur if access to it is attempted. When an operating environment is loaded, this is likely to change.

    FC0000 - FC3FFF Slot 1  Note that it is up to each individual expansion card to
    FC4000 - FC7FFF Slot 2  determine which addresses to respond to in its space
    FC8000 - FCBFFF Slot 3

    FCC000 - FCCFFF Floppy Disk Controller shared memory
    FCD000 - FCDFFF I/O Board Hardware
    FCE000 - FCEFFF CPU Board Hardware

    FCF000 - FDFFFF Also unmapped (?)

    FE0000 - FFFFFF CPU ROM
Lisa addresses have 3 bytes as the 68000 is limited to 16MB of address space (68020 etc. have more). The high byte of a long word address is ignored by hardware and almost always ignored by software -- various components of Mac software use the high byte for flags eg. to indicate a locked handle. When entering an address into service mode it will pad the address with leading zeros, but if you are putting addresses into memory, likely you will need to provide the leading byte, eg. 00xx yyzz.

Once the problem board is isolated, the missing DTACK/VPA signal path can be investigated.

For example, by inspecting "Schematic System I/O Lisa" "050-4008-" page 2 of 5, we see that VPA can be generated by U10D-3 (when one of the VIAs is addressed) or U5E-8 (when the SCC is addressed).

In addition, page 3 shows that DTACK is generated by U5E-6 when the 9512 is addressed, and controlled by the 9512 "Pause" signal that will delay the termination of the bus cycle until it has finished its operation.

Once the problem device is determined, one can examine activity on the signal path at the time of the bus error to isolate where the problem lies at the chip level.

Additional addressing details that may be useful to know (but probably unnecessary) ...

Later versions of the Lisa Hardware Reference Manual have an errata page that indicates the I/O mapping in the manual proper is incorrect. However that errata assumes the I/O mapping will persist as initially set up by the ROM, which is not necessarily true. The MMU can be used to map I/O to many different places/pages in the address range of the 68000, and even to more than one page. The manual proper is written such that it doesn't assume where the I/O will be mapped and so provides offsets from the beginning of that page rather than absolute addresses. In most cases one can use the addresses provided in the errata, but there may be an operating environment where they are not correct (likely suspects are limited to LOS and the Workshop).

One simple example is the SCSI expansion card. On the Mac Plus, the SCSI port base address is $580000. For compatibility with MacPlus software, MacWorks +/II will map an alias of the expansion card slot that contains the SCSI card such that the 5380 SCSI chip also appears at $580000 (as well as in the $FCxxxx space).

The CPU and I/O board hardware addresses are not fully decoded. This means that the devices will also appear at other locations in the address space.

For example, by inspecting "Schematic System I/O Lisa" "050-4008-" page 2 of 5 we see that U7C, the Keyboard & Mouse 6522 VIA has chip select pin 23 driven by a circuit that decodes A10, A11, and A12 (as well as INTIO generate by the CPU board and asserted for eg. FCC000 - FCDFFF). Also note that A1-A4 go to U7C to select 1 of 16 addresses in the chip. As A5-A9 are not decoded, the VIA addresses will be aliased at any variation of A5-A9. eg. as well as the typical base address of FCDD81, U7C will also be found at FCDE81, FCDDC1, etc.

edit 1: 2022-05-10-A corrected VMA to VPA
edit 2: 2022-05-10-B fixed unmapped area below ROM, reference to 4MB modification
edit 3: 2022-05-11-A added SCSI port example
edit 4: 2022-05-11-B changed designation of I/O schematic as variant available on bitsavers differs http://www.bitsavers.org/pdf/apple/lisa/hardware/050-4008-H_IO.pdf
edit 5: 2022-05-14-A Corrected that DTACK for RAM is generated by the CPU board, not on the memory board

 28 
 on: May 10, 2022, 10:41:54 am 
Started by blusnowkitty - Last post by blusnowkitty
https://www.ebay.com/itm/124651364516

Just bring your own QIC drive, chassis, and high-speed parallel card!

 29 
 on: May 10, 2022, 10:40:37 am 
Started by AlexTheCat123 - Last post by AlexTheCat123
Quote
maybe you can use one of those WiFi-enabled ESP32 things for yet another emulator

I've actually been thinking about trying this for a couple of months! I bet it could make my emulator a little cheaper since an ESP32 costs around $5 less than an Arduino Mega. Also, I could probably get rid of the LS280 and do parity generation right on the ESP32 itself since it's way faster than an Arduino. Due to the ommission of the LS280 and the small size of the ESP32, I could probably make the whole thing a good bit smaller as well while still using only through-hole parts, which was a major goal of the ArduinoFile project. And plus, as you mentioned, the WiFi functionality could make it really easy to update the Lisa's clock to the current time!

 30 
 on: May 10, 2022, 10:29:13 am 
Started by blusnowkitty - Last post by AlexTheCat123
Nice work!

One little correction about the ArduinoFile: It supports drives of any capacity just like the Cameo/Aphid does, not just sizes of 5 and 10 MB.

Pages: 1 2 [3] 4 5 ... 10