Read OK, can't write to 1.2MB image

schadel
Posts: 25
Joined: Mon Jan 10, 2011 2:53 am

Re: Read OK, can't write to 1.2MB image

Post by schadel » Fri Feb 11, 2011 9:02 am

Can you make a small schematic of your buffer stage ?
Sorry, I don't understand what you mean. I added resistors to +5 and GND, no buffers.
can you take some oscilloscope pictures of the Write Data line synced on the Write gate line during sector writing ?
I'll try
-b-

Jeff
Site Admin
Posts: 7020
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Read OK, can't write to 1.2MB image

Post by Jeff » Fri Feb 11, 2011 9:43 am

schadel wrote:
Can you make a small schematic of your buffer stage ?
Sorry, I don't understand what you mean. I added resistors to +5 and GND, no buffers.
can you take some oscilloscope pictures of the Write Data line synced on the Write gate line during sector writing ?
I'll try
-b-
ok i didn't understand.

So if understand well you have added a pull up and a pull down resistor ?

schadel
Posts: 25
Joined: Mon Jan 10, 2011 2:53 am

Re: Read OK, can't write to 1.2MB image

Post by schadel » Tue Feb 15, 2011 8:54 am

Jeff wrote:Another point : can you take some oscilloscope pictures of the Write Data line synced on the Write gate line during sector writing ?
I borrowed a DSO. See the attached photos. The second photo shows that "data" begins to appear on WD about 180us after WG goes low.
So if understand well you have added a pull up and a pull down resistor ?
Yes, see the drawing. A typical termination is 220 ohms to +5 (but I used 270 because of the 1K already on the emulator) and 330 ohms to ground. That equates to a termination impedance of about 129 ohms which is ideally near the impedance of the cable to minimize reflections but more importantly it is low enough to charge the cable capacitance quickly. The line driver is open-collector and requires an external pullup. The resistor network effectively connects it to +3V through 129 ohms. 3V is well above the logic "1" threshold, at least for TTL.
Connect it on a PC and try to write something on a ms dos floppy disk image.
I connected it to an HP Vectra and after I learned to set the "File Interface Mode" in Settings to "IBM PC 1.44MB" before exporting to .hfe (I see that it sets a byte in the .hfe file), it worked. I was able to write to the emulator on the Vectra.

I have been using the default Shugart File Interface Mode for the SORD and my 50-to-34 pin adapter follows the Shugart pinout. Does the File Interface Mode choice have any effect other than changing the pinout?

Thanks,
--Bruce--
Attachments
Termination.gif
Lo-Z termination
Termination.gif (13.41 KiB) Viewed 4122 times
DSCF2474_sm.jpg
top: WD and WG; bottom: WD windowed and delayed to show actual data
DSCF2474_sm.jpg (33.32 KiB) Viewed 4122 times
DSCF2471_sm.jpg
Top: WD, bottom: WG (sync)
DSCF2471_sm.jpg (31.23 KiB) Viewed 4122 times

Jeff
Site Admin
Posts: 7020
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Read OK, can't write to 1.2MB image

Post by Jeff » Tue Feb 15, 2011 9:39 pm

schadel wrote:
Jeff wrote:Another point : can you take some oscilloscope pictures of the Write Data line synced on the Write gate line during sector writing ?
I borrowed a DSO. See the attached photos. The second photo shows that "data" begins to appear on WD about 180us after WG goes low.
So if understand well you have added a pull up and a pull down resistor ?
Yes, see the drawing. A typical termination is 220 ohms to +5 (but I used 270 because of the 1K already on the emulator) and 330 ohms to ground. That equates to a termination impedance of about 129 ohms which is ideally near the impedance of the cable to minimize reflections but more importantly it is low enough to charge the cable capacitance quickly. The line driver is open-collector and requires an external pullup. The resistor network effectively connects it to +3V through 129 ohms. 3V is well above the logic "1" threshold, at least for TTL.
Connect it on a PC and try to write something on a ms dos floppy disk image.
I connected it to an HP Vectra and after I learned to set the "File Interface Mode" in Settings to "IBM PC 1.44MB" before exporting to .hfe (I see that it sets a byte in the .hfe file), it worked. I was able to write to the emulator on the Vectra.

I have been using the default Shugart File Interface Mode for the SORD and my 50-to-34 pin adapter follows the Shugart pinout. Does the File Interface Mode choice have any effect other than changing the pinout?

Thanks,
--Bruce--
The write data input are in schmitt trigger mode. The minimum high voltage is 0.8*VDD=0.8*5=4V. 3.3V high voltage is too low.
So i recommend you to remove the pull down resistor or if you prefer increase its value to obtain at least an high voltage of 4.5V.

Another question : what is the write data pulse length ? 500ns ?

schadel
Posts: 25
Joined: Mon Jan 10, 2011 2:53 am

Re: Read OK, can't write to 1.2MB image

Post by schadel » Wed Feb 16, 2011 4:20 am

The write data input are in schmitt trigger mode. The minimum high voltage is 0.8*VDD=0.8*5=4V. 3.3V high voltage is too low.
That's a very high threshold! The Schmitt Trigger receivers (74LS19) on the mainboard (for RD, etc) switch at about 1.9V on a positive-going transition.
So i recommend you to remove the pull down resistor or if you prefer increase its value to obtain at least an high voltage of 4.5V.
I will remove them.
what is the write data pulse length ? 500ns ?
It is 250 ns. But if you look at the first 'scope photo I posted, before adding resistors, you can see that without them the emulator would stretch it to ~500ns if it switches at 4V. (Maybe that's why it didn't work then?) It will be somewhere between if I keep the 270 ohm resistors to +5V.

It would be a problem to make the 50-pin cable shorter for several reasons. Perhaps I should add a receiver chip
to my adapter if I can find a pnp open-collector or CMOS one that will enable the output to go above 4.5V quickly (The 'LS19 won't do). Do you know the threshold for a low-going transition?

merci beaucoup,
--Bruce--

Jeff
Site Admin
Posts: 7020
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Read OK, can't write to 1.2MB image

Post by Jeff » Wed Feb 16, 2011 8:38 am

schadel wrote:
The write data input are in schmitt trigger mode. The minimum high voltage is 0.8*VDD=0.8*5=4V. 3.3V high voltage is too low.
That's a very high threshold! The Schmitt Trigger receivers (74LS19) on the mainboard (for RD, etc) switch at about 1.9V on a positive-going transition.
So i recommend you to remove the pull down resistor or if you prefer increase its value to obtain at least an high voltage of 4.5V.
I will remove them.
what is the write data pulse length ? 500ns ?
It is 250 ns. But if you look at the first 'scope photo I posted, before adding resistors, you can see that without them the emulator would stretch it to ~500ns if it switches at 4V. (Maybe that's why it didn't work then?) It will be somewhere between if I keep the 270 ohm resistors to +5V.

It would be a problem to make the 50-pin cable shorter for several reasons. Perhaps I should add a receiver chip
to my adapter if I can find a pnp open-collector or CMOS one that will enable the output to go above 4.5V quickly (The 'LS19 won't do). Do you know the threshold for a low-going transition?

merci beaucoup,
--Bruce--
In fact the very first picture looks good enough to me. Don't forget that this is an open collector bus. the "slow" raise time isn't the problem here.
Generally HD/500kb/s FDC generate a 500ns pulse.

Do you know the threshold for a low-going transition?
1V

schadel
Posts: 25
Joined: Mon Jan 10, 2011 2:53 am

Re: Read OK, can't write to 1.2MB image

Post by schadel » Thu Feb 17, 2011 1:03 am

So i recommend you to remove the pull down resistor
OK, I removed the pulldown resistors and left the 270 ohm pullups in place. Now I have a different error--WEOE. What does that mean? --Bruce--

Jeff
Site Admin
Posts: 7020
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Read OK, can't write to 1.2MB image

Post by Jeff » Thu Feb 17, 2011 1:03 am

I have just done some tests on an old motherboard with a good fdc:

MFM 500kbits write (300ns pulses) :
Write Precompensation set to 0 ns : OK
Write Precompensation set to 41,7 ns : OK
Write Precompensation set to 83,3 ns : OK
Write Precompensation set to 125,0 ns : OK (the default value)
Write Precompensation set to 166,0 ns : OK
Write Precompensation set to 250,0 ns : OK

So this is definitely not a write Precompensation problem...

Jeff
Site Admin
Posts: 7020
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Read OK, can't write to 1.2MB image

Post by Jeff » Thu Feb 17, 2011 1:04 am

schadel wrote:
So i recommend you to remove the pull down resistor
OK, I removed the pulldown resistors and left the 270 ohm pullups in place. Now I have a different error--WEOE. What does that mean? --Bruce--

you got this during a formating ? if yes this is normal.
this means Write Error - Overflow Error.

This is better ;-)

This means that there are an buffer overflow. Actually the emulator support only normal write sector, not a full track formatting.
Anyway if you format an image, the write track will be rejected (WEOE...), but write sector will be done.
So the full process should works if the image already contains the right disk geometry/sectors.

schadel
Posts: 25
Joined: Mon Jan 10, 2011 2:53 am

Re: Read OK, can't write to 1.2MB image

Post by schadel » Thu Feb 17, 2011 1:19 am

you got this during a formating ?
No, I was just trying to create a small file.
Actually the emulator support only normal write sector , not a full track formatting.
The whole file was only a few bytes that I typed, much less than one sector. The sector size in the image I was using is 256.

Thanks,
--Bruce--

Jeff
Site Admin
Posts: 7020
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Read OK, can't write to 1.2MB image

Post by Jeff » Thu Feb 17, 2011 1:27 am

schadel wrote:
you got this during a formating ?
No, I was just trying to create a small file.
Actually the emulator support only normal write sector , not a full track formatting.
The whole file was only a few bytes that I typed, much less than one sector. The sector size in the image I was using is 256.

Thanks,
--Bruce--
mhh so there are a problem with the write gate line...
well, i got it... its a firmware problem. (write gate not correctly monitored at 500kbits...)
i will correct this as soon as possible.

thanks for your feedback.

Jeff
Site Admin
Posts: 7020
Joined: Fri Oct 20, 2006 12:12 am
Location: Paris
Contact:

Re: Read OK, can't write to 1.2MB image

Post by Jeff » Thu Feb 17, 2011 2:11 am

Jeff wrote:
schadel wrote:
you got this during a formating ?
No, I was just trying to create a small file.
Actually the emulator support only normal write sector , not a full track formatting.
The whole file was only a few bytes that I typed, much less than one sector. The sector size in the image I was using is 256.

Thanks,
--Bruce--
mhh so there are a problem with the write gate line...
well, i got it... its a firmware problem. (write gate not correctly monitored at 500kbits...)
i will correct this as soon as possible.

thanks for your feedback.
new version. tell me what you get with this : (update process in the user manual).
http://hxc2001.free.fr/floppy_drive_emu ... 5_1_1A.UPD

schadel
Posts: 25
Joined: Mon Jan 10, 2011 2:53 am

Re: Read OK, can't write to 1.2MB image

Post by schadel » Thu Feb 17, 2011 2:28 am

mhh so there are a problem with the write gate line...
well, i got it... its a firmware problem. (write gate not correctly monitored at 500kbits...)
i will correct this as soon as possible.
new version. tell me what you get with this : (update process in the user manual).
http://hxc2001.free.fr/floppy_drive_emu ... 5_1_1A.UPD
Thanks, Jeff! I'll flash it tonight and try it tomorrow. --Bruce--

schadel
Posts: 25
Joined: Mon Jan 10, 2011 2:53 am

Re: Read OK, can't write to 1.2MB image

Post by schadel » Thu Feb 17, 2011 2:42 am

new version. tell me what you get with this : (update process in the user manual).
http://hxc2001.free.fr/floppy_drive_emu ... 5_1_1A.UPD
Sorry, I get Error 404 on that link -b-


Post Reply