8 Jun
2014
8 Jun
'14
2:50 p.m.
On 06/08/2014 01:04 PM, Eric Smith wrote:
I'll see if I have anything.
--Chuck
On Jun 8, 2014 11:03 AM, "Chuck Guzis"
<cclist at sydex.com> wrote:
Remember that we're dealing with 8" drives--big AC motor with a
flywheel/pulley on the hub that you could injure someone with. Really
simple to design for. Early 5.25" drives, on the other hand were a
nightmore in terms of ISV and speed control in general. It wasn't until
the direct-drive crystal-controlled 5.25" drives that these problems
pretty much went away.
And yet, the digital WD9216 (or its kin) pretty much displaced all of
the analog data separators. The big problem with an analog PLL is that
the thing has to be adjusted--manually, after assembly. That's the
antithesis of "stuff a board, quick test and ship". Contrast, as an
example, the floppy controllers on the 5150 and 5160.
Take a look at some of the WD179x app notes,
particularly the variations
on data separators. There are several decent ones
using counters (digital
rather than analog) that are actually quite good. I'll wager that you'll
probably find your circuit there.
I'm familiar with those. I was involved with testing those and several
others in 1980-81. While it's possible to make a good DPLL out of TTL,
those aren't. When adjusted correctly, the analog PLL in the app note has
nearly an order of magnitude lower error rate than the other designs shown.
We used the analog PLL until we could get the FDC9216, and later, the 2793
with integrated PLL. The error rate of the 2793 was somewhat higher than
the separate analog PLL, but not enough to forego the cost savings. Even when you add the cost of the additional suppport,
such as a BRG?
Well, you stated that the CTC was used. The SIO, IIRC, did not have an
internal BRG, so you were in for that one, regardless. Since you get 4
timers in a CTC, there were other uses for the remaining timers.
What's the
part number for the PWA on the 8" drive that you have?
77615454