GRONNNNK!
7 messages in this thread
Bela- I don't believe its a software problem at all… and even if it were,
its *NOT* an AmigaDOS problem, since that never even TOCHES the disks. Its the
Trackdisk.Device, which was written by Neal at Amiga (Los Gatos!). AmigaDOS
simply uses the Device calls to Trackdisk, to accomplish all its IO.
— Jez.
Is the Trackdisk.Device on the Kickstart disk or in the devs directory? I
still disagree with your conclusion that it's hardware… the suggestion of
putting sillycone glop around the business end of a high-density magnetical
recorder mechanism sounds like the Jaguar/MG Solution to excessive oil
consumption (use 80-weight oil and cork dust)… i.e., a kludge. If setting
the step rate is possible in software that is accessible to the user (not in
KS) then I know what my next project will be…… Meanwhile, keep your
Winchester Drives well-oiled….hehe -Larry-
Larry, there is a way to find out, if you feel adventurous. In the hardware
manual, there is a listing of the registers that control the disk.
Among the signals are a couple of interesting pins. They have to do with step
direction, and step. Maube even BASIC would be fast enough for this?
Larry-
Trackdisk.Device is in the kickstart, and not in the devs folder…. On
friday, I am getting the new trackdisk documentation so I will see if there
is a way of changing the step speed, to satisfy you people who are SO SURE
that the reason for the noise is step rate.
I agree with your assesment. Those 'cogs' that Define individual tracks are
individual steps in the stepper motor. Therefore, if you hear a 'cogging' noise
during a seek it must be originating in the stepper. (amplified by other
surfaces) All stepper motors I have ever encountered have a particular speed at
which there is no cogging, but continuous, smooth rotation. This would be very
easy to test by toggling the step line under software control, indeed, I
believe that stepping is actually performed by the processor itself, not the
disk controller hardware. The WD 17XX chips do the stepping in hardware, but I
don't believe this to be the case here. Witness the effect of hires video with
4 planes on the step rate! This video mode steals lotsa cycles from the proc.
Guess what? The drive steps slowly. You wouldn't expect that if the harware
were doing it. All this means that, yes, Virginia, you can have a silent Amiga.
(maybe?)
* Reply:
11292
640 x 400 with 4 bit planes steals alot of time from both the Processor and the
Chip cycles. I'm not sure when stepping actually takes place, but all disk I/O
takes place during Chip time, not processor time.
-Dave Haynie
* RR 11229
Well, if the Amiga controller was sorta like a WD17xx, the step pulses would
be derived from the system clock. What you seem to be saying is that this
step mechanism, whatever it is, requires access to the bus. This is true,
since the step line is memory mapped a'la CIA chip. It certainly should be
able to get at the bus once every scan line, I would think. That would allow
it to step at 15 KHz. Seems more likely that trackdisk is using a software
loop to determine step rate. Then again, what do I know?