Re: testing SFD1001 spindle speed

On Sat, Feb 15, 2025 at 6:17 PM Gerrit Heitsch
<gerrit_at_laosinh.s.bawue.de> wrote:
>
> On 2/15/25 00:21, Mike Stein wrote:
> > I don't believe scale matters, just the number of bars per revolution;
> > centering is also not super-critical.
> >
> > I've had one stuck on a filing cabinet for decades and still use it
> > occasionally.
> >
> > You do need a lamp that flickers at 50 or 60Hz though.
>
> Use a larger neon bulb in series with a diode. Without the diode you get
> 100/120 Hz.

100/120 Hz works fine too. Actually the incandescent lamps flicker at
twice the mains frequency because the filament glow is "polarity
agnostic". For strobo patterns any higher harmonic is ok unless you
start running at twice the intended speed (which is unlikely on floppy
drives).

> Or a few LEDs hooked up to an AC source, also with a single diode as
> rectifier.

50/60 Hz flicker gives more contrast indeed, but I've always had very
good results with 100 Hz from incandescent or magnetic-ballasted
fluorescent lamps.
Commodore 100 TPI drives have a narrower spindle working range (they
recommend +/- 1ms on the rotation interval of nominal 200ms) than the
"regular" 48 tpi drives. As I said the JU-570-2 spindle controller is
very good (once you repair the capacitor leakage damages) and I've
always been able to tune to 200.0ms using the index sensor pulse and
an HP frequency/interval counter. And they achieved this stability
with a single turn trimmer too! The one I've just repaired started at
201ms so it was still in the allowed range after who knows how many
years.
By contrast, the SA390 (2040/3040/4040/2031 and Apple disk-II drives)
spindle controller is quite unstable even if they used a 10 turns
trimmer. Every couple of years I find them to have drifted a couple of
ms (no big deal really, but I tend to check my drives every few
years).

Best regards
Frank IZ8DWF