HxC and HP1650A

[jlmartins]

Following!

Please keep us updated!


Thank you!
JL

[Jeff]

Hi Jeff,

I think what I will do is start again with my cable adapter and double check everything. It has turned into a bit of a mess! If it has been proven to work then I'll persist with the project. I'll let your know how I get on...

Cheers,
Andy

To clarify the wiring here is what to do :

hp1650a_hxc_connection.png (153.58 KiB) Viewed 31050 times

One easy way to make this is to take a normal floppy ribbon and cut the wires that need to be disconnect and redirect +5v wires from the HP1650a to the +5v floppy emulator power supply input.


it would be also interesting to try this too : disconnect the disk change signal (pin 2) and check that it is at an high state at the machine side, then try to reboot.

[Jeff]

I compiled the instructions and files in this archive :

https://hxc2001.com/download/floppy_dri ... P_1650A.7z

Code: Select all

-------------------------------------------------------------------------------
-----------H----H--X----X-----CCCCC-----22222----0000-----0000-----11----------
----------H----H----X-X-----C--------------2---0----0---0----0---1-1-----------
---------HHHHHH-----X------C----------22222---0----0---0----0-----1------------
--------H----H----X--X----C----------2-------0----0---0----0-----1-------------
-------H----H---X-----X---CCCCC-----22222----0000-----0000----11111------------
-------------------------------------------------------------------------------
-- Contact: hxc2001<at>hxc2001.com --------------------- https://hxc2001.com --
-------------------------------------------------------------------------------
HxC Floppy Emulator project
(c) 2006-2024 Jean-François DEL NERO / (c) HxC2001
Forum    : https://hxc2001.com/floppy/forum
Facebook : https://www.facebook.com/groups/hxc2001
-------------------------------------------------------------------------------

-------------------------------------------------------------------------------
                     HP 1650A HxC Floppy Emulators package
-------------------------------------------------------------------------------

-------------------------------------------------------------------------------
Original floppy drive and disk technicals informations
-------------------------------------------------------------------------------

Original drive : Sony MP-F52W-30
-------------------------------------------------------------------------------

3"1/2 600 RPM Shugart type drive with an disk change reset line on pin 1.

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
>>>>>>>>> Warning: +5V and +12V Power supply in the floppy ribbon! <<<<<<<<<<<
>>>>>>>>>>>>>>>>> !Floppy ribbon Modification needed! <<<<<<<<<<<<<<<<<<<<<<<<
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

DD capacity at HD rate (500kbits/s)

Note about the "DISK CHANGE" signal (pin) :

This line is true (low) whenever a disk is removed from the selected drive.
The line will remain true (low) until both the following conditions have been
met:

a. A disk is inserted,

and

b. A STEP pulse or DISK CHANGE RESET signal (pin 1) has been received when
the drive is selected.

Drive datasheet :

https://hxc2001.com/download/datasheet/floppy/thirdparty/Sony/9-975-174-01_MP-F52W_Product_Specifications_Sep85.pdf

Disk format
-------------------------------------------------------------------------------

Tracks : 80
Sides/heads : 2
Sectors per track : 5
Sectors size : 1024 bytes

Raw capacity : 80 * 2 * 5 * 1024 = 819200 bytes.

Note : The last track sector id start at 97. (97, 98, 99, 100, 101)

-------------------------------------------------------------------------------
HxC Floppy Emulator wiring and settings :

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
>>>>>>>>> Warning: +5V and +12V Power supply in the floppy ribbon! <<<<<<<<<<<
>>>>>>>>>>>>>>>>> !Floppy ribbon Modification needed! <<<<<<<<<<<<<<<<<<<<<<<<
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>><<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

Please check the file HP165X_HxC_Floppy_Emulator_Connection.png for the correct
wiring.

A simple way to accomplish this is to take a standard floppy ribbon and
cut the wires that need to be disconnected and redirect the +5v wires from
the HP 1650 to the +5v floppy emulator power supply input.

-------------------------------------------------------------------------------

The HFE images must be at 500kbits.

Interface mode : Custom - Pin 2 "nDiskChange 1" - Pin 34 "nReady"

(Note "nDiskChange 1" means : clear disk change signal after a head step)

Set the "Drive B input as Drives Motor On" option.
Disable "2 drives emulation"

Jumper/Switch settings: Set ID0A and ID3B to ON.
(all others switch need to be turned to OFF)

-------------------------------------------------------------------------------
Gotek specific additional instructions
-------------------------------------------------------------------------------
If you use a Gotek with the HxC firmware here is some more things to do

Go to this page :
https://hxc2001.com/custom_fw/

Set :
- Non gated mode
- Advanced pin 2 configuration to "nDiskChanged (mode 1 : DSKCHG Clear -> Step only)"
- Advanced pin 34 configuration to "nReady"
- Enable Reset DC (Diskchange) input signal (pin PA14)
- Bitrate override : "500kbit/s"

Then go down to "Build order" to generate the new firmware.
Once upd file received by email, you can update your gotek with this firmware.
Update procedure :
https://hxc2001.com/docs/gotek-floppy-emulator-hxc-firmware/pages/firmware-update.html

On the Gotek put a jumper to "MO" and remove any other jumper!

The reset diskchange signal at pin/wire 1 at the machine and pin/wire 4 at
the gotek according to the "HP165Xx_HxC_Floppy_Emulator_Connection.png"
wiring instructions need to be routed to the JTCK-SWCLK/PA14 signal.
See "gotek_jumper_and_dc_reset.jpg" to see how to add the connection.
-------------------------------------------------------------------------------

-------------------------------------------------------------------------------
Note : The HXCSDFE.CFG configuration file and file images present in this
archive already have the right settings! Just copy them to the SD Card!
-------------------------------------------------------------------------------

[Jeff]

HP1652B documented success story :

viewtopic.php?p=25945#p25945

I think this can be the same with the HP1650A (same drive)

[synthprof]

Hi Jeff,

I will report back with the 1650a progress soon. I will confirm if they work correctly for this model.

Cheers,
Andy

[synthprof]

Hi Jeff,

The solution to this is still evading me. I see our correspondent with the 1652B is using a GoTek USB drive rather than one of the original SD card type drives. Would you advise I abandon this and get myself an up-to-date GoTek and try again? There seem to be some key variables which are needed which can't be adjusted on the SD version of the drive. In the mean time I have documented what is still happening with the drive I have. I have also attached some Pulseview traces of the key signal lines. I am at your mercy!

So I have built a new interface connection system which allows me to change various wiring and connections easily. I have the lines as suggested in the chart which you previously posted.

With these connections I still run into the same problems. However, I have discovered that /STEP issue where the track number counts up past 256 and then prints 'LBA Access' was being caused by the line floating, even touching the pin with your finger would cause the track number to increment. It now has a 10K pull up resistor one the line and it seeks a sensible number of tracks. However, we still have some nonsense going on:

With the V110 OS image that was posted the screen now prints this at switch on:

"DSKA0000.HFE
RA T:000/080 S:1"

The HP's POST routines run and as this happens we get:

"RA T:079/080 S:0" (000 to 079 whizzes through in under a second)

As the HP beeps to confirm failure of the disk test we get:

"RA T:078/080 S0"

Pressing 'any key' to try and boot from the disk again resets the track number to zero:

"RA T: 000/080 S:0"

...and then rattles through again at high speed to:

"RA T: 078/080 S:0"

So some progress and a more logical action, but not quite there yet.

Cheers,
Andy

[Jeff]

With these connections I still run into the same problems. However, I have discovered that /STEP issue where the track number counts up past 256 and then prints 'LBA Access' was being caused by the line floating, even touching the pin with your finger would cause the track number to increment. It now has a 10K pull up resistor one the line and it seeks a sensible number of tracks. However, we still have some nonsense going on:

This is odd. Maybe you are missing some pull up on others signals too ? Have you an oscilloscope to check the signals ? (10K is quite high, i recommend 1K)

You should at least able to boot the 1650A with the SD HxC. Try to force / set the diskchange signal to 1 / high state.

[synthprof]

Hi,

Thanks for the help so far. I had to break away from this project but had some time over Xmas to take another look. I have basically followed the comprehensive instructions to the letter, but I am still getting the same issue as before. All I can conclude is that perhaps the FDD controller in this analyser is defective, however, the drive is trying to change side and seek so perhaps not. Let me show you a video of the problem and see if you can work out what is going wrong.

https://drive.google.com/file/d/1pSA0tt ... share_link

Watch carefully now! Analysers seem hell bent on sending to heads to a track which doesn't exist. So much in fact that the HxC device thinks it must go into LBA mode to seek it! I have no idea what is going on. All of the images and wiring instructions for the drive are off the shelf. These are the very same results I got last time so I am at a loss.

The only thing I can see is that line /TRK00 is always low, which might explain some of the seeking difficulties.

Next step is the recycling centre, regretfully.

Best wishes,
Andy

[Jeff_HxC2001]

Hi,

Thanks for the help so far. I had to break away from this project but had some time over Xmas to take another look. I have basically followed the comprehensive instructions to the letter, but I am still getting the same issue as before. All I can conclude is that perhaps the FDD controller in this analyser is defective, however, the drive is trying to change side and seek so perhaps not. Let me show you a video of the problem and see if you can work out what is going wrong.

https://drive.google.com/file/d/1pSA0tt ... share_link

Watch carefully now! Analysers seem hell bent on sending to heads to a track which doesn't exist. So much in fact that the HxC device thinks it must go into LBA mode to seek it! I have no idea what is going on. All of the images and wiring instructions for the drive are off the shelf. These are the very same results I got last time so I am at a loss.

The only thing I can see is that line /TRK00 is always low, which might explain some of the seeking difficulties.

Next step is the recycling centre, regretfully.

Best wishes,
Andy

Thanks for the video. It seems that there is maybe an issue with the "DIR" signal. Can you check if this signal change its state ?

If it doesn't : Maybe there is an issue with a buffer in the machine.

If it does: Maybe there is an issue with the emulator regarding the DIR signal. You can test it without the machine with some jumper wires by strapping the drive selection line to the VSS and making some contacts between the VSS and the STEP signal with the DIR connected to the VSS and after this disconnected. You should be able to move the head in both directions.

Clipboard_12-31-2025_02.jpg (87.45 KiB) Viewed 9780 times

This may also be an issue with a damaged floppy ribbon.

[synthprof]

Hi Jeff,

Thanks for the additional information. I have checked the original HxC drive as you suggested and would appear to be working correctly. I also have acquired two additional HxC SD card emulators, one like the one shown in your diagram with the DIP ID selection switches and another of the older type with the large 16 x 2 LCD with blue/white text.Both devices, as well as the original are running f/w 1.8.2.40. All seem still be the same case as before, but there are some developments.

I managed to borrow a second 1650A from a colleague with a working diskette drive. I can confirm that the drive from the working 1650A works fine in my unit and I am able to get to the first menu screen tuning the V110 OIS for the analyser. I created a boot disk for this using an off the shelf 1.0MB DS/DD diskette using the 'LIFARC' program. However, my HxC drive still does the same nonsense in both units. As the original Sony diskette drive works for both we can probably assume that there is something I am going wrong with the HxC drive config or the cable. But what? Both HxC drives have been removed from equipment which is know to work and the tests you suggested work fine.

Just to recap both 1650A report 'diskette failure' on boot when running through POST. If we observe the working Sony drive booting the analyser we see the drive spin up step the head back to track zero and then seek the last track. On screen the test is reported as passed and the head seeks the tracks and sectors required to load the OS. With the HxC the following happens:

On switch on the screen reads:

RA T:000/080 S:1.

As the unit runs through the POST the side toggles to 0 and reads:

RA T:000/080 S:0.

As the disk POST runs the head moves to track 80 and the screen reads:

RA T:080/080 S:0

There is then a beep from the piezo on the system board to alert that the test has failed and the screen reports:

RA T158/080 S:0

Any subsequent attempt to retry the boot by pressing 'any key' with scroll the track number to:

RA T268/080 S:0

and then further onto a message 'Direct Access LBA:00000000 8' which I don't think I have seen before anywhere? Certainly not during normal usage for the HxC for other application. For each next step the /DIR line goes low which I am guessing means the drive creeping ever forward to the last track and beyond.

Are we 100% sure that the notes made for the 1650A were actually submitted by someone who has got this working. Is there anyone who might like to tell me what I am doing wrong because there doesn't seem to be any reason for this not to work, IF it works elsewhere. Can we ask a person who actually has a working 1650A why these steps to do work with two different analysers and instructions followed to the letter? There different HxC all tested and the HFE files and configuration directly copied from the supplied archive.

Many thanks in advance,
Andy

[Jeff]

Regarding your report i think that the issue is probably related to the hfe image. The machine seems to check/search something on the track 80, but this one is just not present in the image from the archive.

Also i see "Note : Original disk copy protected : Can be dumped with the hxc software" in the support listing :
https://hxc2001.com/download/floppy_dri ... upport.htm

the "Direct Access LBA:00000000 8'" message just means that the machine try to access the track 255 ... It probably miss behave cause the "bad" track 80... (EDIT : The machine maybe just NEED an empty 80 track. The way the image is done the hxc will present the track 79 at tracks 80, 81 and so on...)

Since you have access to a working floppy disk, can you image it up to track 82 ?

Since you managed to create a working floppy, you can also to try write the attached empty HFE with LIFARC on a PC and one HxC drive.
Once done you will have to change the RPM/bitrate to make the image working on the machine.
Please share it here once done, i will check if all is fine.
Btw I can try this in my side too : Where can i find the boot/os image to use with LIFARC for this machine ?

[Jeff]

Please try to boot from this image.
If this doesn't work please share the image you used with lifarc.

DSKA0000.HFE.7z

(239.41 KiB) Downloaded 20 times

[synthprof]

Hi Jeff,

Apologies for my last downbeat update. Trying to get this working has been a real trial and quite unlike using the HxC which happily seems to work with most applications! I ma guessing the 1650A is a rare case. I will try what you suggested later on today. In the mean time I will forward these on to you. There are two methods of making diskette for the 1650A: LIFARC and IMD120. I have tested both with the working diskette drive and both give the same results. As normal, the only caveat is you need an x86 computer with a 'real' diskette drive and a Microsoft OS which can run 16 bit programs (Windows 95, MS-DOS, that kind of thing). I have attached both the utilities and image files in their various formats to this thread, so you can have a look in advance. I will try and update you with the blank image test above towards the end of the day.

https://drive.google.com/drive/folders/ ... share_link

Thanks once again for your patience all this.
Andy

[Jeff]

mhh i have see (again?) your mov video file, and i still have doubs about the "DIR" signal (hxc pin 18) : the tracks counter still increasing and is never decremented even when you retry after you press a key to retry, which is quite surprising, and probably not related at all to the image : finally I still think that there is something wrong around this DIR signal...

Have you an oscilloscope ? If you do, please connect the DIR and STEP signal to the oscilloscope with the emulator, Set the trigger on the falling edge of the STEP signal. choose a timebase to see the step signal rising and falling edges.
There is maybe an timing issue around the STEP (hxc pin 20) and DIR (hxc pin 18) signal relationship. The machine may update the DIR signal right before the STEP rising edge, and the emulator use the falling edge as strobe...

This can also be a signal level issue...

use the sd hxc rev f to have an easy access to these signals.

if you have a 4 channels oscilloscope, please add the drive selection (pin 10 or 12) too on the traces.

The "A" on the screen is quite odd too... I would try to force the drive signal (pin 10 or 12 depending of the switches configuration) to vss to see if the behavior is different.