A Lisa Inside An FPGA

[Lisa2]

When I received my LisaFPGA board the first thing I wanted to try was get it connected to the modern internet.  I was hoping to use Localtalk (MacIP), but then I learned that the hardware for Localtalk was not implemented on LisaFPGA. 

As a side note, how did you get your PFG working in the SCC socket? Did you revert to my old code that uses the external SCC, or did you modify the new code to make it switchable between internal/external? Either way, it's really cool to see someone use a PFG with the board; I think you're the first person besides me to actually try it!

I am using your last bitstream before you switched to FPGA based SCC.  I plan add Localtalk support using a daughter board plugged into the SCC socket.  I Will not need the PFG to use Localtalk.
Rick

[AlexTheCat123]

When I received my LisaFPGA board the first thing I wanted to try was get it connected to the modern internet.  I was hoping to use Localtalk (MacIP), but then I learned that the hardware for Localtalk was not implemented on LisaFPGA. 

As a side note, how did you get your PFG working in the SCC socket? Did you revert to my old code that uses the external SCC, or did you modify the new code to make it switchable between internal/external? Either way, it's really cool to see someone use a PFG with the board; I think you're the first person besides me to actually try it!

I am using your last bitstream before you switched to FPGA based SCC.  I plan add Localtalk support using a daughter board plugged into the SCC socket.  I Will not need the PFG to use Localtalk.
Rick

Definitely share the result with us once you get Localtalk working. That's going to be really cool to see!

[stepleton]

I have been thinking that my LisaFPGA would be a good place to store my spare SCC, so I'm also interested in setups that use the real McCoy...

[sigma7]

If you've got a logic analyzer, that would work too.

This might be an opportunity to see if the low cost 8 bit logic analyzers are useful for this sort of issue.

eg. SparkFun TOL-18627
https://www.digikey.com/en/products/detail/sparkfun-electronics/18627/15842546

In addition to the device itself, you'll need the pulseview software suited for your modern computer, and some clips to connect to your target device such as https://www.digikey.com/en/products/detail/digilent-inc/240-137/9916326 (as that's a set of 5, you might want two sets to have 10 clips).

Having used it for a few minutes, I expect most will have some questions getting started.

[sigma7]

As a side note, how did you get your PFG working in the SCC socket? Did you revert to my old code that uses the external SCC, or did you modify the new code to make it switchable between internal/external?

I am using your last bitstream before you switched to FPGA based SCC.

Is the external SCC interface (to use a physical Z8530) still included (albeit disabled) in the LisaFPGA design files on GitHub? If not, is it practical to put it back in so it becomes an optional configuration, or if not, add a branch with the old code?

[sigma7]

probe the PWM line on the Sony floppy connector (J3), which happens to carry MT0
...
I would just probe it myself on my board, but the Twiggy floppy controller won't emit pulses on MT unless it detects a disk in place, and since I don't have any real Twiggy drives
...
And @sigma7, if you could confirm whether or not both of your Twiggies spin with the new LisaFPGA code, that would be great too.

I'm observing a problem with controlling rotating floppies which makes me wonder if there is a problem with my particular LisaFPGA.

I previously reported problems formatting or writing to a Twiggy drive and a 400k drive.

This is the sort of scenario I'm seeing with build 555dcf7, Sony 400K drive, using full 20 conductor cable, A8 I/O ROM, stock CPU speed:

• Switch on USB power. If a disk was already inserted, the drive runs briefly then stops (which is normal).
• Press Lisa Control Power on, arrives at Startup From menu (normal)
• Apple-3 to boot from ESProFile, boots BLU 0.90, all normal
• Insert floppy, runs briefly then stops (normal)
• Select Floppy - Format - Yes erase
• Drive starts, head steps, speed changes, seems normal, completes normally (sometimes).
• Assuming formatting completed: Select Floppy - Read
• Drive starts, progress counts normally, completes normally.
• Select Floppy - Write - Yes overwrite
• Drive starts, a few sectors appear to write properly, then a floppy drive error is reported
• The drive continues to rotate (it should stop)
• Select Fail to return to the BLU menu, drive continues to rotate
• Select Floppy - Eject
• Floppy is ejected, but spindle motor continues to rotate.
• Insert disk and eject again, spindle continues to rotate.
• Press Lisa Control Reset, sometimes self-test shows I/O Board error 57 (probably because FDC is non-responsive)
• Boot into BLU again
• can still insert floppy and eject and motor continues to run.
• Selecting Floppy - Read and first sector generates error
• Press Lisa Control Power off, Powers down but drive continues to rotate
• Lisa Control Power on, Lisa powers up, but floppy drive remains unresponsive aside from ejecting on command.

Switch off and on main power, and drive is again responsive until the hiccup is triggered again.

On my setup this is reliable, so I think unlikely that others would not have seen such floppy failures, making me think it has something to do with this particular unit.

Given the inscrutability (or at least un-probe-ability) of an FPGA, I wonder if there is a hardware test suite for the LisaFPGA to help confirm or disprove this suspicion?

[Lisa2]

As a side note, how did you get your PFG working in the SCC socket? Did you revert to my old code that uses the external SCC, or did you modify the new code to make it switchable between internal/external?

I am using your last bitstream before you switched to FPGA based SCC.

Is the external SCC interface (to use a physical Z8530) still included (albeit disabled) in the LisaFPGA design files on GitHub? If not, is it practical to put it back in so it becomes an optional configuration, or if not, add a branch with the old code?

Yes it's still on GitHub.  To use a physical Z8530,  I used the bitstream from this commit: https://github.com/alexthecat123/LisaFPGA/commit/d179067bed131ce858645aeda29335a92021a05c

Rick

[AlexTheCat123]

probe the PWM line on the Sony floppy connector (J3), which happens to carry MT0
...
I would just probe it myself on my board, but the Twiggy floppy controller won't emit pulses on MT unless it detects a disk in place, and since I don't have any real Twiggy drives
...
And @sigma7, if you could confirm whether or not both of your Twiggies spin with the new LisaFPGA code, that would be great too.

I'm observing a problem with controlling rotating floppies which makes me wonder if there is a problem with my particular LisaFPGA.

I previously reported problems formatting or writing to a Twiggy drive and a 400k drive.

This is the sort of scenario I'm seeing with build 555dcf7, Sony 400K drive, using full 20 conductor cable, A8 I/O ROM, stock CPU speed:

• Switch on USB power. If a disk was already inserted, the drive runs briefly then stops (which is normal).
• Press Lisa Control Power on, arrives at Startup From menu (normal)
• Apple-3 to boot from ESProFile, boots BLU 0.90, all normal
• Insert floppy, runs briefly then stops (normal)
• Select Floppy - Format - Yes erase
• Drive starts, head steps, speed changes, seems normal, completes normally (sometimes).
• Assuming formatting completed: Select Floppy - Read
• Drive starts, progress counts normally, completes normally.
• Select Floppy - Write - Yes overwrite
• Drive starts, a few sectors appear to write properly, then a floppy drive error is reported
• The drive continues to rotate (it should stop)
• Select Fail to return to the BLU menu, drive continues to rotate
• Select Floppy - Eject
• Floppy is ejected, but spindle motor continues to rotate.
• Insert disk and eject again, spindle continues to rotate.
• Press Lisa Control Reset, sometimes self-test shows I/O Board error 57 (probably because FDC is non-responsive)
• Boot into BLU again
• can still insert floppy and eject and motor continues to run.
• Selecting Floppy - Read and first sector generates error
• Press Lisa Control Power off, Powers down but drive continues to rotate
• Lisa Control Power on, Lisa powers up, but floppy drive remains unresponsive aside from ejecting on command.

Switch off and on main power, and drive is again responsive until the hiccup is triggered again.

On my setup this is reliable, so I think unlikely that others would not have seen such floppy failures, making me think it has something to do with this particular unit.

Given the inscrutability (or at least un-probe-ability) of an FPGA, I wonder if there is a hardware test suite for the LisaFPGA to help confirm or disprove this suspicion?

Wow, that's really odd! Does this only happen on a real drive, or on the emulator too? Can anyone else replicate this? I tried and I can't replicate it myself.

[sigma7]

Does this only happen on a real drive, or on the emulator too?

The ESFloppy seems to work fine... can Format, Read, Write multiple times in a row without a hiccup.

I'll probe the cable signals, perhaps the 74HCT245 drivers are damaged (or marginal/inadequate compared the the 74LS244 and 8T97 of a real I/O Board).

[sigma7]

probe the cable signals, perhaps the 74HCT245 drivers are damaged...

It looks like the voltages are good and the edges square (with a 400K Sony drive attached).

Sometimes when I retry after the format fails, it recals to track 0, then steps through the first set of tracks at the first speed normally, then after changing speeds the head carriage clicks at the appropriate rate for track stepping but it does not move.

Very strange that this is a problem with a single unit.

How many other users have used a real 400K Sony drive without problems?

[ried]

No issues with a 400K drive on my unit.

[Lisa2]

How many other users have used a real 400K Sony drive without problems?

I know that Todd had a 400K drive working with his board.  I will try and test this with my board this weekend and let you know the results.
Rick

[sigma7]

Sometimes when I retry after the format fails, it recals to track 0, then steps through the first set of tracks at the first speed normally, then after changing speeds the head carriage clicks at the appropriate rate for track stepping but it does not move.

My 400K Sony problem solved, I think: stepper lead screw was fouled; seems to work now after cleaning & lubrication.

Going back to working on Twiggies...

[AlexTheCat123]

My 400K Sony problem solved, I think: stepper lead screw was fouled; seems to work now after cleaning & lubrication.

Going back to working on Twiggies...

Glad to hear it! I was scared that I had screwed something up with my board design...

[slewis1962]

I tried hooking up my bench supply and was successful in getting the FPGA board to function. I booted into LOS 1.0 and the Twiggy drive would not function. I inserted the disk and it didn't attempt to clamp down or anything. As you can see from the following photo it was only drawing 1.026A @ 5v. When the disk was inserted the current draw didn't change. I tried with my other bench power supply which can supply 10A @ 5v and it was pretty much spot on (1.025A @ 5v) and didn't function either. All my cabling looks to be ok.


I decided I better get out my real Lisa and test the drive to make sure it was still functioning. Here is the result.

So it looks like my drive still works with no problems. Not sure what else to try. The power supply didn't go into current limiting mode or anything. 10A should be plenty. If I designed an interface board with the 12V coming into the board instead of going through the LisaFPGA am I likely to get a better result or would that be a waste of time and resources? Is it possible there is some problem with my board? I installed the latest firmware several days ago before trying any of this.

*Edit* As you can see on the shelf above the Lisa is my other bench supply.
Also, I used the ESProfile on my Lisa and it works just fine.