Use good disks!

[admin]

I spent several hours today trying to figure out what I broke in the code that stopped the copier from duplicating disks. Drive to drive and drive to flux to drive just stopped working. I was using Shadow of the Beast III for testing. It was dieing on track 74, head 1 during the loading process (I have a digital track display on the Amiga) most of the time, but sometimes track 0, head 0 was even bad. Looking at the duplicated tracks with the analyzer, it looked like the data was "smeared". I used the analyzer to read and write the track and got the same results. I changed SCP boards and went back to older firmware versions, all without change. So, I went and grabbed some disks that I had used previously and copies worked fine. It turns out the 6 "new" disks that I was now using for the first time had randomly bad tracks!

In a fit of disgust I tossed the disks in the trash and continued working on code. Right before leaving my office tonight, I pulled the disks out of the trash. Tomorrow I will write a disk tester function that will write data patterns to every track and tell you if the disk is usable... and I have some good test disks to check it out!

So, if you are having problems copying disks, make sure you are using good disks!

[LordCrass]

Could this also include some tests to see what the drive is capable of? eg, write a special pattern to track 80, read it back, write different pattern to track 81, read back, etc to determine max tracks. Or is this a bit dangerous and have the capability to jam some drives at the higher tracks?

Could also use a sequence of bits at defined intervals, then read back to see how accurate the drive/disk is, tell user if weakbits are reported properly, and anything else that might cause someone issues if they aren't sure what their drive can do.

[admin]

I really don't like the idea of extended tracks on drives because I have seen several of the worm gear type stepper motors get stuck when the head jams into the end. You have to manually turn the worm gear to unstick the head. However, there will be a test to determine the max number of tracks if you tell it to. You don't need to write a special pattern to the disk when there is a track 0 sensor. You just step to track 0 and then step the head forwards 'x' number of times, and then step backwards to track 0 counting the number of tracks and see if it matches.

I am renaming "Drive Compatibility Test" to "Drive/Disk Test" and putting the disk formatter/tester in that function as well.

[admin]

Here is what a bad disk looks like with the analyzer. This is track 79, head 1 from Shadow of the Beast III. The line should be perfectly flat, but you can see the portion of the track where the magnetic media has failed. The copy won't load because of this.