Ensoniq ASR-10 and TS-10

HxC Floppy Emulator Ensoniq Keyboards/Samplers support
Western
Posts: 37
Joined: Sat Feb 05, 2011 9:17 pm

Re: Ensoniq ASR-10 and TS-10

Post by Western » Fri Apr 22, 2011 8:11 pm

Jeff wrote:
About the disk read can you try to increase the usb buffer size ?
(Settings -> USB HxC Floppy Emulator status -> USB packet size setting.)

Have you tried the SDCard Floppy emulator ?
Hi Jeff,

I managed to adjust USB packet size setting. Emulator began work better when packet size had been decreased (not increased!) to 1536 bytes and settle time decreased to 5-10 ms. Some minor issues stays, but generally it works.

ASR-10 has two HD-disk formats - normal (1600 kb - 80 tracks) and special (1640 Kb - 82 tracks). Special format is widely used for ASR-10 and TS-10/12 Ensoniqs. It allows increase twice the speed of files loading due to certain order following sectors in the track. Normally sectors in the track come one after the other - 0,1,2,3...19. Not too much time required for processing of one sector, therefore, the operating system is just waiting next sector coming after disk does full rotation.
After processing in special format all sectors with even numbers come first, after this come all odd ones - 0,2,4...18,1,3...17,19. With this procedure the operating system is able to handle loading the two sectors per one revolution of the disc. This gives fantastic acceleration of files loading!

It seems, HxC Floppy Emulator doesn't support this feature (yet?)
Does it means, that even if in original image sectors arranged in a special order, anyway, HxC Floppy Emulator shall output them consequentially?

I made HFE image of real floppy with EDA image written on it and one more HFE image of the same EDA image. They appeared to be different! From Real floppy file loading is fast. From HFE image of this floppy files load TEN times slower! Original EDA and its HFE load some slower than from real floppy.

I'll send these images to your email.

SDCard Floppy emulator I have not had time to check yet.

Thanks,
Youriy.

Jeff
Site Admin
Posts: 5446
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Ensoniq ASR-10 and TS-10

Post by Jeff » Fri Apr 22, 2011 11:50 pm

Interleave changed for the HD mode :

http://hxc2001.free.fr/floppy_drive_emu ... t_beta.zip

Don't forget to make some tests with the sd version too ;-).

Western
Posts: 37
Joined: Sat Feb 05, 2011 9:17 pm

Re: Ensoniq ASR-10 and TS-10

Post by Western » Sat Apr 23, 2011 11:19 pm

Hi Jeff,

I tested the new beta version of HxC Floppy Emulator soft, but it failed. Sorry, it was my mistake, not your . The order of sector numbers in track I gave last time was not correct. It's true for single sided disks only.
I've got more information for this theme and now, it seems, it should be correct. I've found one old document in which stated that "special" disk format is available not only for ASR-10 & TS-10/12 (HD disks) but for EPS-16+ too (DD-disks).
This is a quote of this document:
"...a feature called sector shift speeds up
your disk access by formatting the disk sectors in a special order. The
order of sectors is staggered from one track to the next, so that the drive
doesn't have to spin all the way around to start on the next track. This
feature gives a truly amazing increase in disk performance from floppies!"

I rewrote my EDA image on real floppy in special format using DOS program. After this I've got its "disk map" information using Omniflop diagnostic tools. I did this procedure several times. Due to time unstability of floppy drive each time results were slightly different. But in general the picture is clear.

c00h00x00MFM C00 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c00h01x00MFM C00 H01 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c01h00x00MFM C01 H00 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c01h01x00MFM C01 H01 R{11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10} N02
c02h00x00MFM C02 H00 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c02h01x00MFM C02 H01 R{0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C} N02
c03h00x00MFM C03 H00 R{0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09} N02
c03h01x00MFM C03 H01 R{08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07} N02
c04h00x00MFM C04 H00 R{05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04} N02
c04h01x00MFM C04 H01 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c05h00x00MFM C05 H00 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c05h01x00MFM C05 H01 R{11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10} N02
c06h00x00MFM C06 H00 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c06h01x00MFM C06 H01 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
c07h00x00MFM C07 H00 R{09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08} N02
c07h01x00MFM C07 H01 R{07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06} N02
c08h00x00MFM C08 H00 R{04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03} N02
c08h01x00MFM C08 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c09h00x00MFM C09 H00 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c09h01x00MFM C09 H01 R{12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11} N02
c0Ah00x00MFM C0A H00 R{0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E} N02
c0Ah01x00MFM C0A H01 R{0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C} N02
c0Bh00x00MFM C0B H00 R{0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09} N02
c0Bh01x00MFM C0B H01 R{07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06} N02
c0Ch00x00MFM C0C H00 R{04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03} N02
c0Ch01x00MFM C0C H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c0Dh00x00MFM C0D H00 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c0Dh01x00MFM C0D H01 R{11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10} N02
c0Eh00x00MFM C0E H00 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c0Eh01x00MFM C0E H01 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
c0Fh00x00MFM C0F H00 R{09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08} N02
c0Fh01x00MFM C0F H01 R{08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07} N02
c10h00x00MFM C10 H00 R{05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04} N02
c10h01x00MFM C10 H01 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c11h00x00MFM C11 H00 R{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13} N02
c11h01x00MFM C11 H01 R{12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11} N02
c12h00x00MFM C12 H00 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c12h01x00MFM C12 H01 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
c13h00x00MFM C13 H00 R{09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08} N02
c13h01x00MFM C13 H01 R{07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06} N02
c14h00x00MFM C14 H00 R{04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03} N02
c14h01x00MFM C14 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c15h00x00MFM C15 H00 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c15h01x00MFM C15 H01 R{12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11} N02

As we can see, "side 1" sector numbers shifted +2 from "side 0" of the same track, and "side 0" sectors shifted +3 from "side 1" of previous track. The shifts may be adjusted to get best operation.
Is it possible to to add new parameter into the HXCSDFE.CFG to force sector order mode?

Thanks.
Youriy.

P.S. SD version I'll test after USB version will be OK.

Jeff
Site Admin
Posts: 5446
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Ensoniq ASR-10 and TS-10

Post by Jeff » Sun Apr 24, 2011 12:09 am

No problem :

http://hxc2001.free.fr/floppy_drive_emu ... t_beta.zip

Skew set to 2. (this is not exactly the same scheme in your post, but something close)

Western
Posts: 37
Joined: Sat Feb 05, 2011 9:17 pm

Re: Ensoniq ASR-10 and TS-10

Post by Western » Sun Apr 24, 2011 9:14 pm

Hi Jeff,

Congratulations! It works! I tested last beta version with USB HxC Floppy Emulator on my ASR-10 and TS-10. I loaded different EDA images - load speed is amazing! I used more stable ASR OS 1.61 (instead of last 3.53) - no errors and reboots observed now. We can assume that all is OK with EDA images on ASR-10 and TS-10/12 ensoniqs.

Unfortunately I do not have EDT images for TS-10. All my images in IMG and GKH formats. I tried to load them in ROW mode using next settings:
2 sides MFM 80 tracks 20 sectors per track Sector ID start 0 RPM 300 Sector size 512 Bitrate 250000 Auto GAP3 Format value 246 Interleave 1 Skew 0 Total sectors 3200 Total size 1638400.

I had to force IBM PC 720K mode, because in AUto mode Shugart mode choosed, and if PC 1.44 mode forced - "Disk not formatted" message appears.
Using these settings USB HxC Floppy Emulator Works with IMG files (EDT&EDA similar) but VERY-VERY slow! Even much slower (ten times) than with normal EDA images.
I see, skewing doesn't works for RAW images.
What can we do more?

Thanks,
Youriy.

Jeff
Site Admin
Posts: 5446
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Ensoniq ASR-10 and TS-10

Post by Jeff » Sun Apr 24, 2011 11:53 pm

Western wrote:...I tried to load them in ROW mode using next settings:
2 sides MFM 80 tracks 20 sectors per track Sector ID start 0 RPM 300 Sector size 512 Bitrate 250000 Auto GAP3 Format value 246 Interleave 1 Skew 0 Total sectors 3200 Total size 1638400.
...
I see, skewing doesn't works for RAW images.
What can we do more?
...
Just change the Skew value to 2 and check "Side based" :)

Western
Posts: 37
Joined: Sat Feb 05, 2011 9:17 pm

Re: Ensoniq ASR-10 and TS-10

Post by Western » Wed Apr 27, 2011 9:50 pm

Jeff wrote: Just change the Skew value to 2 and check "Side based" :)
Hi Jeff,

- I've found original EDT images and now we can state that both EDA and EDT images emulate normally.
- I made as you recommended. Unfortunately, this did not help. It seems, there is something wrong with ROW-images emulation, especially with HD images. It is very-very slow - file loading is 30 times slower than from same EDA or EDT-image. DD ROW-images emulate some better, but still slower than from DD EDE-image 5 times.
I decided to do all tests first on my TS-10 as more stable (I see, ASR-10 is very "moody" for floppy loads). At first I exported HFE-image of EDA-image and found that both of them work fast - file load speed is approximately 2 tracks per second. So, there is no difference between this two modes.
Then I loaded the same EDA image in ROW image mode with parameters as stated above. Also I tried "Skew" parameter 2 and 0, "Side based" checked, but result was almost the same. TS-10 normally sees root directory and files normally (may be, opens some slower), but due to Shugart mode I couldn't load any file. So I had to force IBM 1,44 mode. But for unknown reason "Bad device ID" message appears in this case. When I chose IBM 720 mode (for 1600k image!) I could load files. But load speed was very slow - only 1 track per 15-18 sec. I exported ROW to HFE-images for all three cases - auto, IBM 1.44 and IBM 720. "Auto" and "IBM 1,44" images have no difference. "IBM 720" image differs from them with only one byte at address 00010 hex. Its value is 00 instead of 01. Full-workable "EDA" HFE image has more difference from them. I'll send these files to your e-mail.
ASR-10 file load behavior is similar to TS-10, but sometimes different error messages and even reboots observed with ROW images loaded. EDA and EDE images works OK (I already told, now I'm using as recommended more stable ASR OS 1.61 instead of last 3.53).
- I checked "Disk Maps" in different modes of USB HxC Floppy Emulator connected instead of PC FDD with Omniflop diagnostic tools. Here they are:

EDA disk map:

c00h00 - Nothing found
c00h01x00MFM C00 H01 R{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13} N02
c01h00x00MFM C01 H00 R{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13} N02
c01h01x00MFM C01 H01 R{12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11} N02
c02h00x00MFM C02 H00 R{10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F} N02
c02h01x00MFM C02 H01 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c03h00x00MFM C03 H00 R{0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C} N02
c03h01x00MFM C03 H01 R{0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A} N02
c04h00x00MFM C04 H00 R{08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07} N02
c04h01x00MFM C04 H01 R{06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05} N02
c05h00x00MFM C05 H00 R{04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03} N02
c05h01x00MFM C05 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c06h00x00MFM C06 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c06h01x00MFM C06 H01 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c07h00x00MFM C07 H00 R{10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F} N02
c07h01x00MFM C07 H01 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c08h00x00MFM C08 H00 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
---------------------------------------


EDA_HFE disk map:

c00h00 - Nothing found
c00h01x00MFM C00 H01 R{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13} N02
c01h00x00MFM C01 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c01h01x00MFM C01 H01 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c02h00x00MFM C02 H00 R{10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F} N02
c02h01x00MFM C02 H01 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c03h00x00MFM C03 H00 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
c03h01x00MFM C03 H01 R{0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09} N02
c04h00x00MFM C04 H00 R{08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07} N02
c04h01x00MFM C04 H01 R{06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05} N02
c05h00x00MFM C05 H00 R{05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04} N02
c05h01x00MFM C05 H01 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c06h00x00MFM C06 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c06h01x00MFM C06 H01 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c07h00x00MFM C07 H00 R{11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10} N02
c07h01x00MFM C07 H01 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c08h00x00MFM C08 H00 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
c08h01x00MFM C08 H01 R{0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09} N02
c09h00x00MFM C09 H00 R{08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07} N02
c09h01x00MFM C09 H01 R{06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05} N02
c0Ah00x00MFM C0A H00 R{04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03} N02
c0Ah01x00MFM C0A H01 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
----------------------------------------------


ROW IBM 720k mode (skew=0) disk map:

c00h00x02MFM C00 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c00h01x02MFM C00 H01 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c01h00x02MFM C01 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c01h01x02MFM C01 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c02h00x02MFM C02 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c02h01x02MFM C02 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c03h00x02MFM C03 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c03h01x02MFM C03 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c04h00x02MFM C04 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c04h01x02MFM C04 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c05h00x02MFM C05 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c05h01x02MFM C05 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c06h00x02MFM C06 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c06h01x02MFM C06 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c07h00x02MFM C07 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c07h01x02MFM C07 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c08h00x02MFM C08 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
--------------------------------------------

Nothing special, but you see, skew some varies - noticeable that the tracks step a bit uneven. In the previous version (without skew) The tracks step was uniform.

Thanks,
Youriy.

Jeff
Site Admin
Posts: 5446
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Ensoniq ASR-10 and TS-10

Post by Jeff » Thu Apr 28, 2011 9:28 pm

Western wrote: Hi Jeff,

- I've found original EDT images and now we can state that both EDA and EDT images emulate normally.
- I made as you recommended. Unfortunately, this did not help. It seems, there is something wrong with ROW-images emulation, especially with HD images. It is very-very slow - file loading is 30 times slower than from same EDA or EDT-image. DD ROW-images emulate some better, but still slower than from DD EDE-image 5 times.
I decided to do all tests first on my TS-10 as more stable (I see, ASR-10 is very "moody" for floppy loads). At first I exported HFE-image of EDA-image and found that both of them work fast - file load speed is approximately 2 tracks per second. So, there is no difference between this two modes.
Then I loaded the same EDA image in ROW image mode with parameters as stated above. Also I tried "Skew" parameter 2 and 0, "Side based" checked, but result was almost the same. TS-10 normally sees root directory and files normally (may be, opens some slower), but due to Shugart mode I couldn't load any file. So I had to force IBM 1,44 mode. But for unknown reason "Bad device ID" message appears in this case. When I chose IBM 720 mode (for 1600k image!) I could load files. But load speed was very slow - only 1 track per 15-18 sec. I exported ROW to HFE-images for all three cases - auto, IBM 1.44 and IBM 720. "Auto" and "IBM 1,44" images have no difference. "IBM 720" image differs from them with only one byte at address 00010 hex. Its value is 00 instead of 01. Full-workable "EDA" HFE image has more difference from them. I'll send these files to your e-mail.
ASR-10 file load behavior is similar to TS-10, but sometimes different error messages and even reboots observed with ROW images loaded. EDA and EDE images works OK (I already told, now I'm using as recommended more stable ASR OS 1.61 instead of last 3.53).
- I checked "Disk Maps" in different modes of USB HxC Floppy Emulator connected instead of PC FDD with Omniflop diagnostic tools. Here they are:

EDA disk map:

c00h00 - Nothing found
c00h01x00MFM C00 H01 R{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13} N02
c01h00x00MFM C01 H00 R{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13} N02
c01h01x00MFM C01 H01 R{12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11} N02
c02h00x00MFM C02 H00 R{10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F} N02
c02h01x00MFM C02 H01 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c03h00x00MFM C03 H00 R{0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C} N02
c03h01x00MFM C03 H01 R{0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A} N02
c04h00x00MFM C04 H00 R{08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07} N02
c04h01x00MFM C04 H01 R{06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05} N02
c05h00x00MFM C05 H00 R{04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03} N02
c05h01x00MFM C05 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c06h00x00MFM C06 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c06h01x00MFM C06 H01 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c07h00x00MFM C07 H00 R{10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F} N02
c07h01x00MFM C07 H01 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c08h00x00MFM C08 H00 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
---------------------------------------


EDA_HFE disk map:

c00h00 - Nothing found
c00h01x00MFM C00 H01 R{00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13} N02
c01h00x00MFM C01 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c01h01x00MFM C01 H01 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c02h00x00MFM C02 H00 R{10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F} N02
c02h01x00MFM C02 H01 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c03h00x00MFM C03 H00 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
c03h01x00MFM C03 H01 R{0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09} N02
c04h00x00MFM C04 H00 R{08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07} N02
c04h01x00MFM C04 H01 R{06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05} N02
c05h00x00MFM C05 H00 R{05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04} N02
c05h01x00MFM C05 H01 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c06h00x00MFM C06 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c06h01x00MFM C06 H01 R{13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12} N02
c07h00x00MFM C07 H00 R{11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10} N02
c07h01x00MFM C07 H01 R{0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D} N02
c08h00x00MFM C08 H00 R{0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09 0A 0B} N02
c08h01x00MFM C08 H01 R{0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07 08 09} N02
c09h00x00MFM C09 H00 R{08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05 06 07} N02
c09h01x00MFM C09 H01 R{06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03 04 05} N02
c0Ah00x00MFM C0A H00 R{04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02 03} N02
c0Ah01x00MFM C0A H01 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
----------------------------------------------


ROW IBM 720k mode (skew=0) disk map:

c00h00x02MFM C00 H00 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c00h01x02MFM C00 H01 R{01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00} N02
c01h00x02MFM C01 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c01h01x02MFM C01 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c02h00x02MFM C02 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c02h01x02MFM C02 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c03h00x02MFM C03 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c03h01x02MFM C03 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c04h00x02MFM C04 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c04h01x02MFM C04 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c05h00x02MFM C05 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c05h01x02MFM C05 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c06h00x02MFM C06 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c06h01x02MFM C06 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c07h00x02MFM C07 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
c07h01x02MFM C07 H01 R{02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01} N02
c08h00x02MFM C08 H00 R{03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F 10 11 12 13 00 01 02} N02
--------------------------------------------

Nothing special, but you see, skew some varies - noticeable that the tracks step a bit uneven. In the previous version (without skew) The tracks step was uniform.

Thanks,
Youriy.
I didn't noticed but if you are trying to load an HD disk you must also change the bitrate to 500000 !
Also try to increase the skew and/or the interleave.

Western
Posts: 37
Joined: Sat Feb 05, 2011 9:17 pm

Re: Ensoniq ASR-10 and TS-10

Post by Western » Fri Apr 29, 2011 9:03 pm

Hi Jeff,

Once again you were right. My mistake is I didn't change bitrate to 500 Kbit/sec for HD-images. This bitrate never exceeded 250-270 kbit with EDA & EDT HD-images, so it led me astray. Now in ROW mode HD-images work OK (with IBM 1.44M mode forced). Any case, "skew" value tuning allows to increase file loading speed. I 've found optimal values for HD & DD-images. Now it's OK. I'll test SD-version soon and give you a report.

Thanks,
Youry.

Western
Posts: 37
Joined: Sat Feb 05, 2011 9:17 pm

Re: Ensoniq ASR-10 and TS-10

Post by Western » Wed May 04, 2011 8:36 pm

Hi Jeff,

Now I'm testing SD-version of HxC Floppy Emulator. It works exellent with EDE,EDA & EDT images (from list of supported files). I found that ROW-images works good with and skew=3 and interleave=0 (last is absolutely necessary! Otherwise, the file loading speed slows to a crawl and error messages appear). File loading speed with skew=0 is some less (USB- version works similar).

Skew=3 should be good, but there are some problems with saving files on SD-version of HxC Floppy Emulator. Perhaps the reason is that "Save" operation is slower than "Load". Therefore, The operating system does not have time to prepare the data for the next track that goes too fast, so skipping occurs. Therefore, the writing time is increased and various error messages appears on TS-10.
May be, It would be better to leave the normal rotation of sectors (without skew and interleave) when saving files on SD HxC Floppy Emulator.

One more issue:
- When 1.44Mb IBM mode forced (to prevent Shugart mode) on SD HxC Floppy Emulator, "Disk change" signal NEVER goes to 0 level when choosing a new image. So FAT table on TS-10 never rereads. (USB version has no this issue). Only power re-cycling helps.

- Is there a way to independently choose the configuration settings for HD and DD images? The reason is, when 1.44M IBM mode selected in CFG, file load from DD images is very slow.

It would be really cool if for TS-10/12 and ASR-10 IBM mode is automatically selected for all HD and DD images.

I'll continue my tests,
Sorry for the delay

Regards,
Youriy.

Jeff
Site Admin
Posts: 5446
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Ensoniq ASR-10 and TS-10

Post by Jeff » Wed May 04, 2011 8:52 pm

Western wrote:Hi Jeff,

Now I'm testing SD-version of HxC Floppy Emulator. It works exellent with EDE,EDA & EDT images (from list of supported files). I found that ROW-images works good with and skew=3 and interleave=0 (last is absolutely necessary! Otherwise, the file loading speed slows to a crawl and error messages appear). File loading speed with skew=0 is some less (USB- version works similar).

Skew=3 should be good, but there are some problems with saving files on SD-version of HxC Floppy Emulator. Perhaps the reason is that "Save" operation is slower than "Load". Therefore, The operating system does not have time to prepare the data for the next track that goes too fast, so skipping occurs. Therefore, the writing time is increased and various error messages appears on TS-10.
May be, It would be better to leave the normal rotation of sectors (without skew and interleave) when saving files on SD HxC Floppy Emulator.

One more issue:
- When 1.44Mb IBM mode forced (to prevent Shugart mode) on SD HxC Floppy Emulator, "Disk change" signal NEVER goes to 0 level when choosing a new image. So FAT table on TS-10 never rereads. (USB version has no this issue). Only power re-cycling helps.

- Is there a way to independently choose the configuration settings for HD and DD images? The reason is, when 1.44M IBM mode selected in CFG, file load from DD images is very slow.

It would be really cool if for TS-10/12 and ASR-10 IBM mode is automatically selected for all HD and DD images.

I'll continue my tests,
Sorry for the delay

Regards,
Youriy.
Is the save function working correctly with EDE,EDA & EDT images ?

Western
Posts: 37
Joined: Sat Feb 05, 2011 9:17 pm

Re: Ensoniq ASR-10 and TS-10

Post by Western » Thu May 05, 2011 4:39 pm

Jeff wrote: Is the save function working correctly with EDE,EDA & EDT images ?
No, Jeff

"Save" doesn't work with all types of files.

As I stated, TS-10 operation system stores FAT in RAM that reads ONLY ONCE after "Disk change" occurs. After that, all FAT modifications are carried out both on disk and its copy in the RAM.

I see that the system tries to save file (very slowly), but at last stage (rewriting FAT) "File operation error 14" message appears. I do not know exactly what that means, but I suppose that the OS detects the difference between the re-recorded and modified copy of the FAT in RAM.
I can see, that new file appears in the directory (FAT had modified in RAM), but when try to load it "File operation error 2" appears and nothing else happens. After recycling power ("Disk Change" imitation) re-read FAT does not contain the saved file, ie actually write FAT has not happened, and only changed its copy in RAM.

I tried also to perform "Disk format" operation.
This operation goes normally through all the tracks - write-verify is OK and very fast!
And only at the last stage - FAT writing - similar issue observed "Disk error - bad data" message appears. After recycling power FAT stays unchanged.

Maybe write FAT occurs in some other way than writing the files themselves (track-at-once?) and so the parameters of skew and interleave prevent this. But this is just a guess.

May need to try the normal version (without skew and interleave) to eliminate their influence on the "Save" operation.

Because of the inconvenience activate the IBM interface mode on SD-version of HxC Floppy Emulator I can not simultaneously test all types of image-files (HD & DD) as a configuration file allows choose only one of 1.44M or 720K mode at once.

What can I do more?

Thanks,
Youriy.

Addendum.
As I tried to format already formatted disk image, may be, write operation really did not take place at tracks too, and verification was OK because of the fact that the disk has been formatted properly before.
When I tried to format the disk only the name was changed, it explains why only the FAT did not match with the previous.

Jeff
Site Admin
Posts: 5446
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Ensoniq ASR-10 and TS-10

Post by Jeff » Thu May 05, 2011 9:30 pm

Can you send me the HFE file before and after the write process ?

Western wrote:Because of the inconvenience activate the IBM interface mode on SD-version of HxC Floppy Emulator I can not simultaneously test all types of image-files (HD & DD) as a configuration file allows choose only one of 1.44M or 720K mode at once.
Of course you can ! the interface mode is specified in each HFE file and not in the CFG file.

Western
Posts: 37
Joined: Sat Feb 05, 2011 9:17 pm

Re: Ensoniq ASR-10 and TS-10

Post by Western » Sat May 07, 2011 12:37 pm

Jeff wrote:Can you send me the HFE file before and after the write process ?
I compared HFE files before and after the write process using WinHex software - NO CHANGES was found.
Although I saw sequence of WR & RA, tracks and sides changing on SD- HxC Floppy Emulator's display during "Save" and "Format" operations. For unknown reason they are "dummy".
By the way, why "Disk change" may not work in IBM interface mode?

I'll try re-format SD-card once more with FAT32 option and do tests again.

What can I do more?

Regards,
Youriy

Jeff
Site Admin
Posts: 5446
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Ensoniq ASR-10 and TS-10

Post by Jeff » Sat May 07, 2011 7:49 pm

Can you send me the HFE file ?

Did you see a "WEBS/ WETO / ..." message on the lcd during saving ?

Post Reply