Replacing Tandon TM-100-2A in HP 9130 disk drive for HP 86A

HxC Floppy emulator support for all others computers...
Post Reply
Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Replacing Tandon TM-100-2A in HP 9130 disk drive for HP 86A

Post by Martin Hepperle »

Hi,

just to add to the success stories: I was able to replace the Tandon 5-1/4" disk drive in a HP 9130 disk drive unit with a HxC slim emulator.

The only problem that I have is that I must use the preformatted LIF image creator in the HxC software to create an empty but already formatted disk image. If I try to format an image using the HP 86As INITIALIZE command formatting stops after an arbitrary(?) number of tracks (usually 7, 14, or 21 ... there seems to be some pattern). This is not fatal as I can use the preformatted images.

I used an adapter board for 5-1/4" edge connector to the 3-1/2" connector on the HxC board as shown at the end of this document
http://www.mh-aerotools.de/hp/hxc/HxC%2 ... Drives.zip

Martin

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

Just to add to the information above:
there is another issue with this setup replacing the Tandon TM-100 on the HP-86:
- I can create programs and STORE and LOAD them on the SD card without problems.
- I can GSTORE and GLOAD graphics screens
- I can create a large (say 1000 records of 256 bytes) file using the CREATE command of the BASIC system. I can then write and read to this file without problems from BASIC.
- However, when I try to copy a file of similar size (1000 records) from a second (real) disk drive the copy operation fails after 3 tracks. This symptom is similar to the INITIALIZE (formatting) problem: it starts to run and stops after a number of tracks. Maybe a timing problem.

Martin

Jeff
Site Admin
Posts: 8092
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Jeff »

Martin Hepperle wrote:Just to add to the information above:
there is another issue with this setup replacing the Tandon TM-100 on the HP-86:
- I can create programs and STORE and LOAD them on the SD card without problems.
- I can GSTORE and GLOAD graphics screens
- I can create a large (say 1000 records of 256 bytes) file using the CREATE command of the BASIC system. I can then write and read to this file without problems from BASIC.
- However, when I try to copy a file of similar size (1000 records) from a second (real) disk drive the copy operation fails after 3 tracks. This symptom is similar to the INITIALIZE (formatting) problem: it starts to run and stops after a number of tracks. Maybe a timing problem.

Martin
Try to use a faster sdcard (Class 10 Sandisk Extreme).

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

Okay I will try with new Class 10 cards.

Another thing I noticed:
I was sometimes able to copy the large file when I set pregap to a value of 100. I am not sure what pregap or gap3 actually means.

However, there seems to be a problem with the Software.

I am using HxC FloppyEmulator 2.1.4.5 Software

I tried to redefine the HP LIF file format for my own purposes.

However no matter what I specify in the "RAW File format definition" dialog
some of the adjustments never make it to the file.

Looking into an exported XML file I see that the values for
<gap3>255</gap3>
<pregap>0</pregap>
always have the values 255 resp. 0.
The values given in the dialog box seem to be ignored.

The same seems to be the case for <format>
The value is always <format>IBM_MFM</format> no matter what I select in the dialog.
There seems to be something broken in the handling of input values in this dialog.

What "seems" to be working is:
- export the image to an XML file
- change the value for pregap with a text editor
- load the modified XML file into the software
- export the file as HFE

Martin

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

Hi again,

So I bought a new Sandisk 16 MB Ultra UHS-I Class 10 card which is rated at up to 80 MB/s.
Formatted with FAT32 in a Windows 7 command window (I saw that the card is already preformatted, but just to be sure).

Then I try to initialize (format) the .hfe image in place of the Tandon TM-100-2A drive.
I can hear 28 track change clicks. Then I get a disc error.

If I export the resulting partially formatted .hfe file and look look into the RAW file I see that it is filled with 0xDB bytes up to offset 0x3AFFF.

0x3AFFF = 241663d
241663 bytes / 256 bytes / sector = 943 sectors
943 sectors / 16 sectors / track = 58 tracks
58 tracks / 2 heads = 29 cylinders
This corresponds to the number of clicks I hear.

I had the same reliability problem when I tried to copy a large file (265 kB) from a 3.5" SONY disk to the HxC emulator.

Then I did another experiment: I used a second 5.25" Tandon disk drive as a source and copied files from this drive. In this case the copying went through without errors, but the resulting image cannot be read. The 5.25 inch source drive is noticeable slower than the 3.5 inch drive, this may explain why the copy operation worked. However I can only access the first sectors (where the directory is located) without errors. I can also delete files in the directory, change the volume label etc.
All these things seem to work without problems in the first sectors of the image.

If I try to load a file from a higher track number it fails.

If I load the .hfe image into the Floppy Emulator Software and export it as a RAW file I can see that all the data looks good. For example I can see that a certain BASIC program resides on track 32.

When I try to load it the emulator seeks to sector 0 (to read the directory and the starting address), then to track 32 (which is correct) and then fails with "ERROR 130 : DISC".

Another thing I noticed is that the HxC uses 1K pullup resistors on some of its input lines. If I look at the Tandon drive schematics I see that they use terminating resistors of only 150 Ohm on the last drive in the chain. In my case there is no chain - the Tandon drives are connected straight to the computer.

Was anyone successful to replace a Tandon TM-100 drive on e.g. a TRS-80? How did the cabling look like and were there any special tricks involved?

Still confused, Martin

Jeff
Site Admin
Posts: 8092
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Jeff »

Martin Hepperle wrote:Hi again,

So I bought a new Sandisk 16 MB Ultra UHS-I Class 10 card which is rated at up to 80 MB/s.
Stop. I said Sandisk Extreme... Not Ultra...

Retry with this firmware :
https://hxc2001.com/download/floppy_driv ... rmware.zip

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

The firmware update did not make a difference (I had already installed a beta with the same version number in order to attach a display to the HxC Slim).

So: I have now ordered one of the EXTREME cards (it was not clear to me that read and write speeds can be SO MUCH different and on Lotharek's web shop I saw that he sells "Sandisk 8GB Ultra SDHC Class 10" cards so I thought this would be good enough). I have collected some technical data on the SanDisk cards in the attached document.

We will see whether this makes a difference - the current Ultra Cards (silver label) seem to have a write speed in the order of 10-15 MB/s (when copying a large file in Windows). I hope the Extreme cards will be faster in writing.

I will come back in 1-2 days when I have the new card.
Attachments
SanDisk Cards.zip
(188.33 KiB) Downloaded 399 times

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

Hi Jeff,

in the meantime I obtained a SanDisk Extreme Card (16 GB, 90 MB/S read speed, U3 write speed, Class 10, UHS-I). The keyword seems to be U3 for recording speed.

Using this card I now obtain the following results:

Formatting a Disk Image
  • This now proceeds over all cylinders (0...34, heads 0,1) and writes the "empty" byte 0xDB into all sectors - this is okay. So the faster write speed of the SD card fixed the issue with stopping after some cylinders.
  • Next, the head is moving back to sector 0 to write the Volume and Directory information. At this point the formatting fails with "ERROR 130 : DISC".
It looks like the HxC would be too slow or out of sync when seeking back to track 0.

Using a Preformatted Disk Image
  • I then wrote a small BASIC program reading complete 256 byte sectors in a sequence from logical sector 0 to 1199 over the complete disk.
    This needs no Volume or directoy information. It worked without problem, so the image is completely readable.
  • Then the program was modified to force the HxC emulator to seek for each sector.
    A reading of sector 0 was inserted just before reading the desired sector. Thus the emulator had to seek to sector 0, read it, and then seek to the desired sector and then read this sector.
  • This procedure worked up to about cylinder 27 was reached. Then it failed.
Obviously the back and forth seek time was too high so that the HP 86 issued “ERROR 130 : DISC”.

I also hooked a 8 channel logic analyzer to the signals.
Looking at the signals does not show me anything special (but this means nothing).
Only the reading attempt of the last sector on track 27 is very long, compared to the previous reads.

Some timings
  • Index pulse duration 4 ms, interval 0.195s which corresponds to a speed of 308 1/min, which is very slightly out of the 300 ± 1.5% band
  • Read data bits duration 2 μs
  • Head step pulse duration 4 μs (o.k > 200 ns)
Note that these times may be inaccurate due to the limited resolution of my cheap analyzer.

Martin
Attachments
screenshot.png
screenshot.png (27.16 KiB) Viewed 9551 times

Jeff
Site Admin
Posts: 8092
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Jeff »

Please share the HFE files : the formatted one and the pre-formatted with the write attempt.

(Is the double step needed by your system ?)

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

Jeff,

the attached file contains three .HFE files copied from my SD card and a readme.txt which explains the differences.

The original disk drive is a classical Tandon TM-100-2A - the same was also used in the IBM-PC as a 360 kB disk drive. It does not use any of the "modern" features like double-step as on the later 1.2MB drives for the IBM-AT.

Martin
Attachments
HP-86.zip
(178.89 KiB) Downloaded 460 times

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

Jeff wrote:(Is the double step needed by your system ?)
This "double step" (seek to a low sector and then seek to another higher numbered sector) is occurring whenever a file is to be accessed: first the directory entry is read (usually at sector 3 on cylinder 0 and following) and then the head must seek to the first sector of the file, which may be e.g. at cylinder 30 or so. I think this is a normal operating scheme.

Martin

Jeff
Site Admin
Posts: 8092
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Jeff »

File to try into your email box.

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

Jeff wrote:File to try into your email box.
Thank you (your e-mail ended up in my SPAM folder because it did not contain any text, so I did not see it at first).
I tried the DSKA0000.hfe you provided.

Unfortunately it did not improve the situation.

CAT - show directory
  • T:000/35 S:0 head moves to cylinder 0
  • directory is read from sectors 3 up and displayed. O.K., as before

LOAD "Autost" - try to load a file
  • T:016/35 S:1 head moves forward, but only to cylinder 16
    - should move to cylinder 32 where "Autost" begins
    - track change sounds like a continuous beep (fast pulses)
  • ERROR 130 : DISC
LOAD "Autost" - retry
  • T:000/35 S:1 head moves back to cylinder 0
    - track change sounds like single beeps (slower pulses)
  • ERROR 130 : DISC
If I retry my test program (for N=0 to 1199: {read sector 0; read sector N}):
DSKA0000.hfe: errors at sector 576, track 17 (seeks very slow, bip-bip-...)
DSKA0001.hfe: errors at sector 588, track 18 (sometimes +/- a few sectors)
DSKA0002.hfe: errors at sector 867, track 27 (seeks very fast, short beep)

The first HFE file is the one you provided, the other two are the same that I had sent in the previous ZIP file.

So the new DSKA0000.hfe file seems to have a problem when it shall seek further than approximately 16-17 cylinders. I have not yet tried to seek e.g. from cylinder 10 to 27+ to see whether the seek distance also makes the same problem when jumping between higher cylinder numbers.
The file DSKA0002.hfe seems to be better, but not good enough: is can seek up to 27 tracks.

So my interpretation is that the seek takes too long.
On the other hand the real floppy disk is rather slow, so I am not sure what is going on.

Martin

Martin Hepperle
Posts: 24
Joined: Fri Sep 09, 2016 11:20 am
Location: Germany

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Martin Hepperle »

I also changed my test program to jump back not to sector zero but to a higher sector number
to see whether this shifted the possible cylinder range to higher values.

The results using DISKA0001.hfe are
starting at sector 000 = from cylinder 00 to 18 then error, delta = +/- 18
starting at sector 300 = from cylinder 09 to 18 then error, delta = +/- 9
starting at sector 400 = from cylinder 12 to 24 then error, delta = +/- 12
starting at sector 500 = from cylinder 15 to 25 then error, delta = +/- 10

The number of cylinders that can be skipped without error is not constant and varies between 9 and 18 cylinders.
When I seek from a higher starting cylinder errors occur a bit later, but the range of cylinders which can be skipped does never increase. Sometimes it is much lower than 18.

Martin

Jeff
Site Admin
Posts: 8092
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Replacing Tandon TM-100-2A in HP 9130 disk drive for HP

Post by Jeff »

So forget the last HFE. This mean than you don't the double step with this machine.

I just send 2 new hfe to try.

Post Reply