Victor 9000 / Sirius 1 Hard Disk Emulation

[djg]

I used mfm_util on your file and get

Ran out of data on sector index 17, try adding --begin_time 240000 to mfm_read command line

Your controller formats with a sector straddling the index pulse. This won't cause problem acting as an mfm disk emulator but will prevent creating an extracted data file and using ext2emu.

If you haven't invested too much time in it you can redo the file with

setup_emu
mfm_emu --file st412_a --drive 1 --initialize --heads 4 --cylinders 306 --begin_time 240000

then run FACTORYF etc on it. Only use --initialize on the first start of mfm_emu. Use your normal command line for the next run.

If you have autostart on you will need to systemctl stop mfm_emu before running it manually.

You can use mfm_util to check the file. Should decode without error.
mfm_util --emu st412_a --ana --ext /tmp/t

I might be able to make mfm_util decode the file if you decide its too much effort to regenerate and find a need but life is busy till end of April so won't be soon.

 

[1302L]

Hi djg,

thanks for this offer!

I will redo a copy of the file with your instructions and report back.

Best,

Martin.

 

[1302L]

Hi djg,
I followed your instructions, everything went quite smooth, here's the new file.
I would be happy if you could comment on this.
Thanks in advance!
Martin.

 

[djg]

Looks good to me. Added it to the images.

 

[1302L]

Looks good to me. Added it to the images.

Many thanks @djg !
Then I will make this file my default file.
Next step will be trying to create a larger image, i.e. 20MB, let's see if that works as well.
Best,
Martin.

 

[djg]

Released code that fixes the too small seek time prints. It was calculation issue so won't change emulator behavior.

 

[1302L]

Released code that fixes the too small seek time prints. It was calculation issue so won't change emulator behavior.

Hi,
then I suppose it's time for a new update of the board. ;-)

Looks good to me. Added it to the images.

I made this (st412_b) my default autostart emu file. The emulation starts well most of the time, sometimes it is getting stuck in the middle of booting DOS, mostly before loading the port drivers. Unfortunately I haven't made an emulation log, I will try to do this tonight. BTW: How am I doing this, when auto emu is on?

The other thing is, I am getting a lot of errors loading software fom this emulation image, errors are:
- Machine freezes and needs a hardware reset with the reset button on the back
- the 'usual' error is stated 'cannot load xy, program too big to fit on memory'
- or 'file allocation error'.
=> Any comments on this?

I will switch back to my former file and see how it goes with it, will report back.

Best,

Martin.

 

[djg]

Unfortunately I haven't made an emulation log, I will try to do this tonight. BTW: How am I doing this, when auto emu is on?

See Debugging in
http://www.pdp8online.com/mfm/revb/revb_board_build.shtml

The other thing is, I am getting a lot of errors loading software fom this emulation image, errors are:
- Machine freezes and needs a hardware reset with the reset button on the back
- the 'usual' error is stated 'cannot load xy, program too big to fit on memory'
- or 'file allocation error'.
=> Any comments on this?

Since 'cannot load xy, program too big to fit on memory' you were getting with the bad memory board I would be suspicious of hardware problem. The disk controller verifies using the sector header that the correct sector is being read and uses the CRC to verify that the data is correct. It is unlikely that any error from the emulator would cause the computer to see bad data from the disk without getting an error. It should just be a read error unless its not properly checking for errors.

Have you gotten your floppy working well enough that you can try doing the same things that give errors from the emulator from floppy?

Is is repeatable that one file always give an error? If so would be good to try copying the file to floppy and see if it give the same error running from floppy but booted from emulator or booted from floppy and run from floppy. If it does give error try comparing the file against the original copy to see if it has gotten corrupted.

Have you run chkdsk?

 

[pdevine]

Hi @1302L, I downloaded your image just to check if I could boot my machine, and it did indeed boot up correctly. I was only able to see the right floppy drive as drive B. Otherwise it worked fine.

 

[1302L]

Hi @djg ,

See Debugging

I will look it up and give it a try.

I would be suspicious of hardware problem.

Have you gotten your floppy working well enough that you can try doing the same things that give errors from the emulator from floppy?

Yes, I did no errors when executing from Floppy, most time, my impression is, the longer the floppy has to work, the more reading errors it gets.

Is is repeatable that one file always give an error?

Yes, it is, please see an example screenshot attached. The Chess game is working fine from floppy, but not on HDD, anytime I try it, I am getting the error message below.

Have you run chkdsk?

Yes, please see screenshot attached.

Best,

Martin.

 

[1302L]

Hi @1302L, I downloaded your image just to check if I could boot my machine, and it did indeed boot up correctly. I was only able to see the right floppy drive as drive B. Otherwise it worked fine.

Hi @pdevine ,

many thanks for this feedback, did you try some of the programs in the emu file as well? Regarding the conversation above and for reference, maybe you could try the CHESS game.

Cheers,

Martin.

 

[djg]

chkdsk /f will clean up the checkdisk error. Error isn't causing the problems your are seeing.


The error you show this time for chess is an I/O error. That wasn't in your previous list.

Check your emulator file with
~/corvus/mfm/mfm_util --emu st412_b --ext /tmp/t --ana
replace st412_b with the correct file name you are using with the emulator, emufile_a?
Does it report any errors?

If you try to copy whatever files are part of the chess program do you get the I/O error?

You said you were going to try the emulator file without --begin_time. Does that behave differently?

 

[1302L]

Hi djg,

sorry for my late reply, I was 'out of order' being sick.

I have checked the emulator file, since I have renamed st412_b in emufile_a right after the test, I performed the test with this:

debian@BeagleBone:~$ mfm_util --emu emufile_a --ext /tmp/t --ana
Original decode arguments: --heads 4 --cylinders 306 --begin_time 240000 --rate 10000000
mfm_util: unrecognized option '--rate'
Found matching format EC1841: good count difference 0
First logical sector 7
Number of heads 4 number of sectors 17 first sector 0
Interleave (not checked): 0 7 14 4 11 1 8 15 5 12 2 9 16 6 13 3 10

Command line to read disk:
--format EC1841 --sectors 17,0 --heads 4 --cylinders 306 --header_crc 0x0,0xa008 05,32,2 --data_crc 0x0,0xa00805,32,2 --sector_length 512 --retries 50,4 --drive 0 --begin_time 240000
Found cyl 0 to 305, head 0 to 3, sector 0 to 16
Expected 20808 sectors got 20808 good sectors, 0 bad header, 0 bad data
0 sectors marked bad or spare
0 sectors corrected with ECC. Max bits in burst corrected 0
debian@BeagleBone:~$

Trying to copy files from the chess game back to Floppy brings me a write error which I overran by typing (i)gnore, this made the chess game on the floppy unuseable as well, bringing a read error.

You said you were going to try the emulator file without --begin_time. Does that behave differently?

Yes, I did use the emulator file once at the beginning with --begin_time as everything worked fine, I went back to my normal emulation command:

mfm_emu --drive 1 --file emufile_a

Now the emulation runs on autostart, which works fine.

Beside the issue with the chess program the emulation runs fine and other software, i.e. SuperCalc or WordStar are running well.

Best,
Martin.