Print

[rayarachelian]

I'd poke at it with BLU or NeoWidex and check the status of the drive as this will give you more detailed info, but yeah, reinstalling the OS seems like a good next step.

And most importantly, congrats!

[AlexTheCat123]

And most importantly, congrats!

Thanks! It feels great to get this lucky with an "untested" ProFile from eBay! Now it's just a matter of replacing the RIFA caps in the PSU and the ProFile will be fully repaired.

I'd poke at it with BLU or NeoWidex and check the status of the drive as this will give you more detailed info, but yeah, reinstalling the OS seems like a good next step.

Good idea. I'll give that a shot once I get the floppy controller fixed and can boot from a floppy disk. The tracking info says that the 6504 just departed from China this morning even though I ordered it a while ago, but hopefully it will be here soon!

Do you think it would be worth trying to recover any data off the ProFile before formatting it? This would be tough since I'm almost certain that my serial ports don't work due to corrosion on the motherboard so I couldn't image it with BLU, but it would suck to accidentally erase some sort of rare or prototype piece of software.

[rayarachelian]

Do you think it would be worth trying to recover any data off the ProFile before formatting it? This would be tough since I'm almost certain that my serial ports don't work due to corrosion on the motherboard so I couldn't image it with BLU, but it would suck to accidentally erase some sort of rare or prototype piece of software.

If your serial port is working, there is a way to type in a mini loader into service mode and transfer BLU over the serial port. You could use that to both probe the status of the ProFile *and* image it to see if there's anything there.

see: http://sigmasevensystems.com/blumanual.html "Lisa Simple Loader"

[AlexTheCat123]

If your serial port is working, there is a way to type in a mini loader into service mode and transfer BLU over the serial port. You could use that to both probe the status of the ProFile *and* image it to see if there's anything there.

Based off the amount of corrosion on the resistor packs on the motherboard that connect to the serial ports, I highly doubt that they're working. I can buy a serial cable to try it out, though. I would just need a DB9 to DB25 null modem cable, right?

[rayarachelian]

ard that connect to the serial ports, I highly doubt that they're working. I can buy a serial cable to try it out, though. I would just need a DB9 to DB25 null modem cable, right?

Yes, that (null modem cable) and a new Lisa

[AlexTheCat123]

Yes, that (null modem cable) and a new Lisa

I've had so many issues with this darn Lisa that I've often considered just getting a new one, but I feel like I'm finally on the home stretch with just the serial ports, composite video output (both of these are because of the corroded motherboard), and floppy controller to fix at this point. But of course, it would never hurt to have a second Lisa 

I just looked through one of my boxes of cables and I happened to find an appropriate null modem cable as well as a USB to serial adapter, so I can test the ports once I clean the corroded gunk out of them. If the serial ports don't end up working (they almost certainly won't), do you know of any good sources for a new motherboard? The resistor packs are so corroded that I doubt I'd be able to remove them from the board and the pads would most certainly be fully corroded away. I would rather have a new motherboard than run 30-something bodge wires to patch the traces and connect a new resistor pack. I know that John at VintageMicros has made some brand new unpopulated boards, but I don't really want to pay $250 for a bare PCB and I'd rather have an original for increased authenticity.

[Lisa2]

Sorry, folks.  While trying to split the topic,  I accidentally moved Ray's last reply/post here:

https://lisalist2.com/index.php/topic,191.0.html

This was his post:

"These days, it's either VintageMicros or eBay. I think we hit the top of the curve about two years ago for parts availability sadly, and now we're at the bottom of the barrel.
You can either keep struggling and repair the boards you have, and I know it's very hard, hence my flippant "and a new Lisa" - and indeed, I feel your pain, or save up for a new mobo, or wait until one becomes available.

The odds of a working 2/5 Mobo showing up on eBay is pretty low, most that will show up will be corroded.

There's another factor here, in that most of the 2/5 machines have had a battery leak at some point in the past, and while at some point they worked despite the leak, or were patched up, the alkaline contents got into the traces of the I/O boards and slowly over time ate those infected traces. So boards that once worked over time will, or have already failed.

Most of these types of Lisas that have been sold on ebay were chopped up into parts, and the parts sold, but if you see photos of them, most are very corroded.

2/10's don't have this issue at all, but 2/10 motherboards are incompatible with 2/5 chassis due to the wiring harness and require the 2/10 I/O board.
The only exception is that it's possible to use a 2/10 I/O board in a 2/5 if you use a modified Lite adapter.

If you try the reverse, a 2/5 motherboard in a 2/10 chassis, the internal widget cables would be shorted out when a Widget is connected, but you will be able to use the external parallel port - one of my two original Lisas was this kind of FrankenLisa which caused lots of other issues.

(I'm using 2/5 and 2/10 here to simplify - 2/5 ofc refers to a Lisa 2 that has a 5MB ProFile and 2/10 refers to a 2/10 Lisa that has an internal Widget, but I'm ignoring the storage medium here to be a bit more clear instead of just saying Lisa 2. AFAIK all the Lisa 2's were "upgraded" from Lisa 1s.)

2/10s are really nice if you replace the Widget with an Aphid/XProFile/IDEFile, etc. as Widgets are pretty flaky. There are also timing issues due to clock changes on the 2/10 board, and this causes incompatibilities with 3rd party devices such as some revisions of the FloppyEmu (and the in progress Aphid). The other advantage is that Xenix assumes that if you have a 2/10, the drive must be 10MB in size, which is a problem if you happened to snake a cable out the back to an external ProFile, etc.

But yeah, as in the other thread about the total number of Lisas, most of the existing working ones are at this point owned by collectors, the damaged ones have been either trashed long in the past (sadly) and there's very little stock now. So modern replacements are encouraged when affordable.

I wonder... I don't know if it's easier to repair an existing board than to design one from scratch based on the existing board and have some place print new boards and find replacement parts and then desolder the old board and solder the components to a new board (or use new components). Both are going to be a huge uphill struggle. So given that, I think what Vintage Micros offers is actually a great price. Sure, 10-20 years ago when we thought there were 100K Lisas out, and most were working, there, those prices wouldn't make sense, now they do.

Would it be less work to build/print a replacement PCB vs buying from Vintage Micros? How much is your time and energy worth? This guy seems to have done exactly this and even added SCSI to a Lisa 2 with a D-ROM (replacing the Serial B connector for some reason): https://www.youtube.com/watch?v=Z5DraRVMVFM - but won't released his gerbers, etc.

For some of us, this situation is a lot like owning a house for a very long time  in a city whose real estate has all of a sudden hyperinflated - like San Francisco and now the owners of said house are priced out by taxes and are forced to move - that is kind of analogous to not being able to find replacement parts to keep it working."

----------
Sorry,
Rick

[AlexTheCat123]

Would it be less work to build/print a replacement PCB vs buying from Vintage Micros? How much is your time and energy worth? This guy seems to have done exactly this and even added SCSI to a Lisa 2 with a D-ROM (replacing the Serial B connector for some reason): https://www.youtube.com/watch?v=Z5DraRVMVFM - but won't released his gerbers, etc.

Once you put it that way, I guess the one from VintageMicros does sound like it's priced reasonably. I was actually the person who asked that guy in the video for his Gerbers and it sucks that he won't release them.

I hooked up my serial cable and tried loading BLU over it and the Lisa actually seemed to receive the data somehow, despite how awful the motherboard looks in that area! After the entire file was sent, the loader returned control to service mode like it should, but I discovered that while the data at address 9c0 matched what the BLU manual says, the data at the entry point of c14 was very different from what the manual says it should be. I tried sending the file multiple times and I even lowered the baud rate to 9600, but the data at this point is still wrong. What could be going on here?

[rayarachelian]

I'd make sure you're using the latest BLU - beyond that, you could ask @Sigma7 who wrote it.
It's possible the data is not actually making it across properly, or that your cable lacks the proper lines for HW flow control (CTS/DTR/etc.) but you'd need other things working to test that (floppy/profile).

[AlexTheCat123]

It turns out that the problem was a combination of me being an idiot and the Lisa not liking my USB to serial adapter. Although I checked continuity on the RX and TX lines going from serial port B to the I/O board, I didn't think to check the flow control lines until just now and it turns out that they're faulty. I soldered some wires straight from my serial cable to the appropriate pins on chips on the I/O board and it received more of the data correctly than it did before, but there were still errors. I then tried it with an old Windows 98 laptop that actually has a built-in serial port and BLU fired right up!

I powered on the ProFile and tried to read its contents to the Windows 98 computer over serial and it seemed to work fine. After looking at the disk image in a hex editor, it seems like it was running SOS for the Apple III, which explains why it wouldn't boot on the Lisa. It appears that it was previously owned by a country club in Florida based off some of the other text that I found in the disk image. I then asked BLU to exercise the disk and that seemed to work fine too. To test booting off the ProFile, I selected the option that installs BLU onto your hard disk and now it boots straight into BLU off the ProFile perfectly!

This is some encouraging news, so I guess it's just a matter of recapping the ProFile and Lisa PSUs, getting the actual serial ports working instead of running bodge wires, and fixing that darned floppy controller at this point.

[AlexTheCat123]

Is there a utility that will convert a DC42 image of a ProFile to a BLU image that I could transfer over to my ProFile? Since the floppy drive still isn't working, I was thinking about trying to install LOS in LisaEm, converting the resulting ProFile image to BLU format, and then sending it over to my Lisa, but I can't find a utility that will convert from DC42 format to BLU format.

[rayarachelian]

Is there a utility that will convert a DC42 image of a ProFile to a BLU image that I could transfer over to my ProFile? Since the floppy drive still isn't working, I was thinking about trying to install LOS in LisaEm, converting the resulting ProFile image to BLU format, and then sending it over to my Lisa, but I can't find a utility that will convert from DC42 format to BLU format.

This might work: https://github.com/rayarachelian/lisaem/blob/unstable/src/tools/src/raw-to-dc42.c if not let me know and I'll add an explicit dc42-to-blu command - not sure if BLU checks for the BLU signature or not. BLU is the same format as raw, but with the ffffff block first and in there it has a BLU version #

Easiest way to build this is to build the unstable branch of LisaEm.

[AlexTheCat123]

The 6504 finally came in the mail and after putting it in, the floppy controller seems to be working slightly more than it was before, but it's definitely still broken. I don't get the error 57 anymore and the I/O ROM version is no longer FF, but the version is only reported as A8 every five or six times that I look at its memory address in service mode. Other times, it's seemingly random values. Sometimes, the Lisa lists two floppy drives in the Startup From... menu and other times it just lists one, so clearly something isn't right. It sometimes just hangs and sometimes gives error 39 (drive ROM can't keep up) when I try to boot from the floppy drive and BLU always gives floppy errors whenever I start it up. What should I try next?

[rayarachelian]

The 6504 finally came in the mail and after putting it in, the floppy controller seems to be working slightly more than it was before, but it's definitely still broken. I don't get the error 57 anymore and the I/O ROM version is no longer FF, but the version is only reported as A8 every five or six times that I look at its memory address in service mode. Other times, it's seemingly random values. Sometimes, the Lisa lists two floppy drives in the Startup From... menu and other times it just lists one, so clearly something isn't right. It sometimes just hangs and sometimes gives error 39 (drive ROM can't keep up) when I try to boot from the floppy drive and BLU always gives floppy errors whenever I start it up. What should I try next?

So either there's something wrong between the I/O board bus pins and the CPU board (i.e. motherboard connector) or the SRAM on the floppy controller is broken.

On initialization, the 6504 will write it's version to the shared SRAM. It will do this only once.

So if that value fluctuates, something is wrong either at the bus level (and we've seen this before with your Lisa) or at the SRAM level.
You can swap out the SRAM to be able to tell which.

One thing you could look at is to see the values that show up that are not A8 and map those in binary. This will tell you which bits are either broken in the SRAM or on the bus.

Another thing you could do is piggy back a second identical SRAM chip on top of the existing one. If the problem goes away that might indicate the SRAM.
However, if the existing SRAM is bad and reports "1" values where it should be "0", the 2nd SRAM will not help.

There's a tiny bit of danger to this since you're connecting two outputs together, but it's very small. Worst case you'll cause the new SRAM to burn out as well, but I don't think these are so expensive that it matters that much.

If the randomness issue doesn't go away, it's still either the SRAM or the bus.

If you have a large logic analyzer it may help you debug futher. i.e. the old Agilent/HP ones that you could interface directly to a chip. @Sigma7 introduced me to these early on in the days of LisaEm, but it turned out I didn't need to break out this big gun to work on LisaEm, and indeed I needed very little actual testing on a real Lisa in 99.999% of cases. So I'm not very well versed in how to use such test equipment. However, in this case, if you own one and have it handy it could immediately help you diagnose the issue.

All this said, the amount of time it will take you to diagnose, it might just be quicker to desolder to SRAM, solder in a socket, and then insert a new SRAM. If the problem goes away, you're done in ~30 mins. If it doesn't, you know exactly where to look next. (Well maybe some lines from the area on the I/O board where the SRAM and the logic gates that enable it to be visible to the 68000 - DISKDIAG?, to the fingers on the I/O board, to the motherboard connector, to the traces on the motherboard, to the CPU board connector, to the CPU board itself.

[AlexTheCat123]

I just wrote a simple Arduino program that will test these RAM chips by writing every possible value to each address and then reading the value back and both chips passed just fine. I guess this means that it's a bus issue, which will not be very fun to troubleshoot .

If you have a large logic analyzer it may help you debug futher

I have a USB logic analyzer, but the vast amount of data that it produces when connected to the Lisa makes it impossible to go through it all. The text and CSV files that it outputs are so large that I can't open them in any program without it crashing.