Does the Kryoflux firmware support "track shifting"...

All questions about how to use KryoFlux go here.
Post Reply
mcc
Posts: 19
Joined: Sun Oct 28, 2018 11:19 am

Does the Kryoflux firmware support "track shifting"...

Post by mcc » Sat Nov 03, 2018 1:48 pm

Hi,

I just ordered my KryoFlux and I have to wait at least til 5th of November +shipping and delivery time before
I will be able to kryoflux my disks.

I want to read my 5.25" flippy ATARI800 disks.
But I am scared of the flippy mod...its hard enough to get a 5.25" floppy drive at all...I problem I need to solve.

So I am thinking of mods, which are not that intrusive and reverseable if something fails.

On idea was to create an artifical sync/index hole pulse by using a small refelxion light barrier and a bright
mark on the fly wheel/motor disk, which I found on the back of many drives by searching for images of that
drives.

But the position of the light barrier is critical it seems, since the pulse not only controls the motor speed but also the beginning of track 0.
On the other hand: Since the ATARI815 disk drive can handle flippy disks without any index pulse, I think
the track positioning and therefore the identification of sector 1 can done be software only.

Does the firmware of Kryoflux can mimicry the "index less" reading of a 5.25 floppy like it was done
by the ATARi815 drive ?

Then the only problem I have to solve is to keep the motor spinning...

Is this supported by the firmware?

Cheers!
mcc

SomeGuy
Posts: 179
Joined: Wed Feb 18, 2015 8:18 pm

Re: Does the Kryoflux firmware support "track shifting"...

Post by SomeGuy » Sat Nov 03, 2018 10:37 pm

The kryoflux does not care where the tracks actually begin. When decoding to a sector image file, it will look anywhere. You should be able to read the first side of an Atari 400/800 disk with no modifications even though the tracks are not aligned to anything.

There are actually two separate problems to not having an index hole on the second side:
1: Most 1.2mb drives will not even enter a "ready" state if they can not find an index pulse, preventing any kind of reading at all. (Many 360k drives do not have that issue).
2: The Kryoflux uses the index to measure the exact drive speed and track length. If no index pulse is detected, the Kryoflux will stick its nose up at any attempt to read.

Wiring in a second IR sensor set will work around both of these problems. Since the tracks are not index aligned to start with, it probably does not matter how exact it is, as long as the index gets triggered.

However, the Kryoflux folks will complain that this method is "insufficient" for flux preservation purposes as the index is technically in the "wrong" place. However, that only matters if you are dumping factory original software titles.

If these are only crufty data disks or user created copies, then you might consider just using a hole punch to create a second index opening in the jacket (not the cookie!).

mcc
Posts: 19
Joined: Sun Oct 28, 2018 11:19 am

Re: Does the Kryoflux firmware support "track shifting"...

Post by mcc » Sun Nov 04, 2018 4:29 am

Hi SomeGuy,

thank you VERY much for clearifying a bunch of technical things for me.
In articles on some sites I read, that the index pulse also marks sector 1
and thought that all my data get "shifted" somehow, if I won't hit the right
spot with my light barrier.
Opening the sleeve of disks, which are more than 35 years old, scares me a lot.
It may even be easier to get a new drive, if I damage one than getting the data
of those disks back.
I will try the method with an addition IR sensor first.

I think I can 3D-Print the lower part of a kind of "housing" or "case" for the bottom
part of the driver. This 3D printed part could held the IR sensor then.

I have to figure out the electrical part, though...

Cheers and thank you again for the explanations!
:)
mcc

User avatar
mr.vince
Posts: 2101
Joined: Tue Oct 05, 2010 5:48 pm

Re: Does the Kryoflux firmware support "track shifting"...

Post by mr.vince » Tue Nov 06, 2018 8:34 pm

SomeGuy wrote:
Sat Nov 03, 2018 10:37 pm
However, the Kryoflux folks will complain that this method is "insufficient" for flux preservation purposes as the index is technically in the "wrong" place. However, that only matters if you are dumping factory original software titles.
We'll, I am tempted to say you're already belly up when you take a disk that _was_ index aligned and try to copy it (which will use the index).

While everything was ok, having an offset will now give you grief as the write splice will be in the wrong area.

Everyone's free to do whatever seems reasonable, but usually there are reasons we do (not) recommend stuff. It's not because we like to argue. ;)

Post Reply