Fwd: eddy current brake

[Mazda Ebrahimi]

tom cloud wrote:
> 
> >Hi there,
> >
> >think the Eddy.effect is exlained enough, so lets go back where the idea came
> >from (can't remember who it was, but what it was):
> >reducing/removing backlash of the ignition rotor!
> 
> Help me understand ..... (I've listened to related discussions re
> ignition control for a while)
> 
> I hear the 4 pointed star crank pickup isn't good enough to use
> a PLL to develop nth degree resolution.
> 
> Same for the distributor pickup.
> 
> So, seems everyone wants to use either the flywheel teeth or a gear
> attached to the cam (maybe even the cam gear??).  Is that right?
> 
> I know n-nothing -- at least not practically.  Seems to me the crank
> pickup with PLL would work.  Any variation due to speed of the tooth
> approaching the pickup giving faster rise time ergo changing reference
> could be compensated for either in an analog circuit or in a confuser
> using either a lookup table or (probably) a simple trig function.
> I don't think the crank's gonna have much backlash :-)
> 
> As far as backlash, why wouldn't an averaging function take care of
> that?  After all, how fast can an engine change speed? and the delta-speed
> seems to me to be the most egregious factor (pun) in successfully
> implementing some kind of smoothing circuit (maybe even an f-v convertor ??)
> 
> Now, there's an idea I absolutely haven't explored .... Do an f-v conversion
> and filter it ??? would that work?
> 
> Tom Cloud <cloud at peaches.ph.utexas.edu>


The problem with all this is that it is an estimate.  What we care 
about, at least in the ignition system, is the exact location of the 
piston when the plug is fired.  The most accurate way to do it is by 
monitoring the crank.  I'm just gonna throw this out there (i.e. I'm not 
sure about this):  It seems to me the amount of error in a distributor 
measurement is a function of engine speed and rate of acceleration.  The 
engine speed factor would probably not be a linear function, and the 
rate of acceleration should affect the system because of additional 
strain on the timing belt / chain. Right?
It's obvious picking up the signal off the distributor works, but I'd 
almost always rather make a direct reading than to read a sensor and 
then perform calculations (or look-up) to figure something out.