Trying to seek the disk past cylinder 202 - what is the behavior?

[Carl Claunch]

I am in the final tests of an added capability for 1130 systems which involves working with the internal disk drive. I didn't fully test the disk drive by running diagnostics during the restoration of this 1130 system because I planned on using it with the new capability I developed. However,  the disk diagnostic never attempts to seek beyond 202 thus would never have detected this situation anyway. 

If I issue an XIO Seek with a cylinder count that would result in the arm trying to move past cylinder 202, the drive itself locks up with a continual buzzing, the voice coil trying to move the arm past the mechanical stop at 202 and it won't reset even when the Reset button of the 1130 is pressed. I have to switch off the drive to reset it. 

Nothing in any of the 1130 documentation specifically addresses this case. My working assumption is that the disk drive should finish the seek with the arm remaining at 202 and not lock up. If so I have to figure out what is wrong and correct it. However, it may be that the lock up behavior is what would happen on any internal or 2310 drive.

I realize I am straining the memories of everyone here, but if anyone knows definitively how the drive should behave, it might save me a lot of fruitless debugging. 

Carl

[John Doty]

My memory is that there were limit switches preventing seeks outside the range.

--
IBM1130 discord channel: https://discord.gg/nrFEMt48mE
---
You received this message because you are subscribed to the Google Groups "IBM1130" group.
To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/ibm1130/b238cc2c-755b-4b37-a30a-2f4705d71459n%40googlegroups.com.

John Doty

j...@noqsi.com

[Robert Lerche]

I recall hearing that knocking sound (definitely "something is wrong") though I don't remember needing to reset a drive. I think there was a diagnostic program that caused it deliberately. These are all vague memories but the sound I can still hear.

--
IBM1130 discord channel: https://discord.gg/nrFEMt48mE
---
You received this message because you are subscribed to the Google Groups "IBM1130" group.

To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+unsubscribe@googlegroups.com.

[Carl Claunch]

Hi John

There is a limit switch at cylinder zero which not only reports the Home bit in the DSW but blocks any seek command from activating the drive in reverse. This stops it from attempting to move backwards past 0.

There are also two mechanical stops which block the arm from moving inward past 202 or outward past 0. 

There is no limit switch for 202, thus the drive attempts the seek. 

At this point, I think the mechanical stop for 202 does not have enough gap from the location of 202 until the arm is blocked. I will be doing an adjustment of it later today and will test again.

Interestingly, even if the mechanical stop works and the seek ends with the arm not moving past 202, it can get into an incorrect state. The drive physically moves either 1 or 2 tracks per movement. The disk controller circuitry in the 1130 converts the cylinder count from the seek IOCC into a number of steps, with the first being 1 track if the count in the IOCC is odd otherwise it is all 2 track steps. The drive has two different detents - teeth that lock into notches in a positioning rack to hold the arm at one of the 203 cylinders 0 through 202. Even positions use the right (even) detent, while odd steps use the left (odd) detent to stop. Thus, any time the drive takes a 1 cylinder step, it toggles which detent it is using. 

This means that if the drive is legitimately at cylinder 202 and gets a 1 track step request, it will switch from the even to the odd detent and thus can't actually remain at 202. It won't move beyond because of the mechanical lock, but it might slip back to 201 or lock into the notch for 203. The drive should remain at 202 regardless of the attempts to move forward, but it could be at 202 or not depending on which detent was active and the last step size. 

Carl

[Carl Claunch]

Hi Robert

If a program was already at a very high cylinder number and tried to move forward another large number of cylinders, it would be hitting the mechanical stop for each 1 or 2 cylinder step until the entire seek ended. 

That would be reasonable behavior. The buzz that I hear that continues forever even when you press Imm Stop and Reset on the console is a pathology, one I don't expect a properly working drive to exhibit. 

The IBM disk diagnostic program checks and does not try to seek past 202, at least based on the code for the version I have. 

Carl

To unsubscribe from this group and stop receiving emails from it, send an email to ibm1130+u...@googlegroups.com.