Here are some more pictures... (top red thing is the read head) http://electronicsam.com/images/KandT/conversion/accpinset1.jpg http://electronicsam.com/images/KandT/conversion/accpinset.jpg
This is how I understand it as of today.... ;) There are 4 coils - they are hooked up in a center tap config (see schem) - 2 sets of 2 coils. An excitation signal (250khz square wave) is sent to the outside connections of the 2 center tapped coils. The center taps are summed together and turned into a square wave. That square wave is shifted compared to the exciter signal depending on the position relative to the .1 pin. Now the way I think the controller did it was this - it had a 250khz clock - they used this to count the shift between the exciter signal and the summed square wave back from the center taps. this would give you 250khz/250hz - 1000 divisions within each pin. thanks sam On 4/11/2010 08:03 PM, Jon Elson wrote: > sam sokolik wrote: > >> there are actually 4 coils. Each head has 2 shielded cables coming from >> the head - each cable has 4 conductors + shield. At the controller - >> the 2 coils on each cable are hooked together to form a center tapped >> setup. (agian - if I have it right - they excite the 2 outside >> connections of the 2 center tapped hookups - then the center taps get >> summed together and shaped. this from trying to read the desciption on >> the schematic I scanned - plus you can see the coil hookups) :) >> >> > I don't know, looking at the jpg of the schematic, it doesn't really > look like the windings will work the way you want for the AD chip. It > really doesn't look like there is an excitation winding and a pair of > sense windings. With 114 Ohms per coil, the drive requirement can't be > terribly high, so that may not be a problem. If the AD chip can be made > to work, the resolution will be 4096 counts per period of the teeth on > the long scale. That probably is OK, as I think these teeth are about > 10 per inch. Ah, yes, I see it IS a GE Accupin scale, I had already > guessed it might be from your description. The way one of these schemes > worked is they drove sine-wave signals in quadrature to the two sin/cos > windings, and then looked at the time of the zero crossing on the other > winding. That told the position of the windings relative to each > other. This one almost sounds like it works the same way, but the > description says square wave. So, maybe they are using some analog > scheme to also sense the voltage of the output as well as the phase. > > Anyway, it looks like this may be fairly hard to make work. > > Jon > > ------------------------------------------------------------------------------ > Download Intel® Parallel Studio Eval > Try the new software tools for yourself. Speed compiling, find bugs > proactively, and fine-tune applications for parallel performance. > See why Intel Parallel Studio got high marks during beta. > http://p.sf.net/sfu/intel-sw-dev > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users > > ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
