evad wrote:Gazza,
Looking at the index pulse with a scope during normal read write operation does mimic a tandon well, However looking at the index pulse during the format ( 305 Rpm so we can see the 40 track operation ) shows a different story. As I said before the pulse stream is a bit erratic and there are very quick neg pulses in between the larger index pulses, however after a few revs the index pulse is held down for over 10x the normal pulse width. Using a slower SD card I believe yields a longer width during the time it is held down. Hence maybe this is how the emulator throttles the WD controller.
Exactly. During the write process, the virtual disk rotation is paused. So if the host machine have a too strict write/disk rotation timeout value, this maybe a problem. That's why a fast SD is recommended in these particular cases.
During any read process, you shouldn't see these kind of "throttles". Anyway, try to use the very last firmware version :
evad wrote:Gazza,
Looking at the index pulse with a scope during normal read write operation does mimic a tandon well, However looking at the index pulse during the format ( 305 Rpm so we can see the 40 track operation ) shows a different story. As I said before the pulse stream is a bit erratic and there are very quick neg pulses in between the larger index pulses, however after a few revs the index pulse is held down for over 10x the normal pulse width. Using a slower SD card I believe yields a longer width during the time it is held down. Hence maybe this is how the emulator throttles the WD controller.
Exactly. During the write process, the virtual disk rotation is paused. So if the host machine have a too strict write/disk rotation timeout value, this maybe a problem. That's why a fast SD is recommended in these particular cases.
During any read process, you shouldn't see these kind of "throttles". Anyway, try to use the very last firmware version :
Jeff,
Thanks for the response.
Already running v1.8.2.40 as I just received the unit 2 weeks ago. ( Version shown on boot flashes by very fast )
I am using Sandisk 8gb. Ultra SDHC Class 10. They are 40MB/s instead of 30MB/s as shown on the site http://lotharek.pl as tested cards.
I have ordered SanDisk Extreme SDHC UHS-I/U3 16GB Memory rated at 90MB/s read and 60MB/s write.
Hopefully these will work
Jeff wrote:
Exactly. During the write process, the virtual disk rotation is paused. So if the host machine have a too strict write/disk rotation timeout value, this maybe a problem. That's why a fast SD is recommended in these particular cases.
During any read process, you shouldn't see these kind of "throttles". Anyway, try to use the very last firmware version :
Jeff,
Thanks for the response.
Already running v1.8.2.40 as I just received the unit 2 weeks ago. ( Version shown on boot flashes by very fast )
I am using Sandisk 8gb. Ultra SDHC Class 10. They are 40MB/s instead of 30MB/s as shown on the site http://lotharek.pl as tested cards.
I have ordered SanDisk Extreme SDHC UHS-I/U3 16GB Memory rated at 90MB/s read and 60MB/s write.
Hopefully these will work
evad wrote:Jeff,
Since in the document " SDCard HxC Floppy Emulator HFE File format - Rev.1.1 - 06/20/2012 "
typedef struct picfileformatheader_
unsigned short floppyRPM; // Rotation per minute (Not used by the emulator)
Where can I find this document? I checked out the libhfe portion of the sources and could find nothing in terms of documentation. If it's there I'll feel embarrassed.
Thanks - not sure how I missed that. I'm interested in generating an HFE to emulate hard-sector diskettes, but I'm not seeing a way to define the sector index pulses. I recall you mentioning that this capability was added recently. Does the doc need an update?
Also, if the bitstream itself carries the full encoding, why are the 'trackencoding' and 'floppyintererfacemode' bytes required? Are they used by the emulator? If so, what for?