Fwd: eddy current brake

[Bob R. McElroy]

With a 4-point crank wheel and one sensor for that wheel, you would have =
4 pulses, each 90=B0 apart.  A sensor wheel could be constructed so that =
it gives a rising edge and a falling edge four times per revolution, =
with the falling edge 45=B0 out of phase with the rising edge.  This =
provides the 8 points necessary to have a resolution of one event  every =
45=B0.

I'm responding to this post because it took me a couple seconds to =
understand what was intended.  I first thought a wheel with four small =
(a couple degrees wide) teeth was being described.  I then realized that =
Tom was describing a wheel with teeth that spanned 45=B0 each.  However, =
I'm still puzzled when he states "2 revs =3D 360".  One revolution =
equals 360=B0.  Two crankshaft revolutions, however, does equal 360=B0 =
of camshaft revolution (oh, now I get it.  Never mind.)

When you start talking about PLL, however, I'm lost.  What exactly does =
a PLL do or tell you?  Sorry if everyone else considers this question =
trivial, but I only studied basic electronics in college.  I was lucky =
enough to take a class on internal combustion engines, but that mainly =
centered on the thermo-chemistry involved with combustion, and not many =
practical exercises (like most of my engineering courses.)  On the plus =
side, I did get to use the Heywood text book that I see listed under the =
recommended reading list for EFI information.

Thanks for any help.

Bob McElroy

----------
Lessee, 600 at idle to 6000 is 10:1, which is enough for me.  With a =
4-star
crank sensor, you get a pulse every 45 degrees, since 2 revs =3D 360, =
right?
Then, if you PLL'd the freq by 45, you'd have 1 degree pulses, by 90,
.5 degree.  Am I right?  I think I understand the need for 'reasonable'
accuracy with firing, but you've already got a hardware (crank) 45 =
degree
reference.  The PLL jitter shouldn't give but only a few tenths of a =
degree
error, if that much.  Capture shouldn't be a problem, as I don't see how =
you
can rev or decelerate the engine fast enough to lose lock.

Remember, I'm just positing here -- never actually tried this.


Tom Cloud <cloud at peaches.ph.utexas.edu>