Russell McMahon wrote: > What I am saying is that if you can tolerate the ringing voltage it > has least effect on operation if you leave the diodes open (or with a > very light load). Sure. But just to be safe, if you use a Zener rated for slightly more than Vcc, returned to Vcc, paralleled by a small snubber cap of maybe 10 nF, then the autotransformer action we both described will not cause this to conduct, though true "ringing" or circuit shutoff *will* be limited. Mark Willis wrote: > I was thinking Sqrt(2) gain in torque, roughly, if you energize 2 > adjacent windings? I suspect this will be defeated because the equilibrium position involves only partial alignment of pole pieces. Could be wrong I suppose. > No comprendo how you get 4 wires out of a 6 wire unit? Unless you > drive it with bipolar drivers & use the diodes to bypass half the > coil, is that the trick here? A five wire unit has a wire connected to the centre tap of both "axis" windings. A six-wire has the centre tap of *each* winding brought out separately. On odd occasion you may get an 8-wire unit with four separate windings. Now... If you connect both CTs of a 6-wire together, and to Vcc, you have used it as a 5-wire units haven't you? You have the option however of using *only* the "ends" and using it as a 4-wire (2-coil) unit. I'm not at all sure which should be called "bipolar" and which "unipolar", so I won't use the terms. *Next* trick. If you have a suitably-behaved H-bridge, you can *still* use a 5-wire as a 4-wire! As long as the H-bridge drives either one direction, the other, or tri-states to produce no net drive, then the centre tap of each coil will sit at ¸Vcc and it does not matter that they are connected. Mmmm! > Whimper Seems people are always arguing this one, if via a > resistor or Zener, how do you do the 6-wires to 4-wires trick? Whatever you drive with an H-bridge, the reverse diodes across each driver perform this task automatically. Just takes a think to sort it out! In fact, the reverse diodes across the drivers *do* limit flyback as long as the two coils are 100% coupled. > Why does "Jones on Stepping Motors" etc. have diodes across the > coils? Or to Vcc and Ground, depending on coil type. I don't wish to be too PICky, but I suspect Jones only *ever* uses bridges for which his diodes are perfectly correct. His comment about using slow switching diodes reveals another common mis-perception. Slow diodes switch *on* quite fast, which is what is required in this application. Their problem is not switching *off* when reverse-biassed, but here the current has generally decayed to zero by the time they are subjected to reverse bias. > Lasers, I wish I had a spare, could seriously use 48 or 64 LED's at > 1/300th inch spacing (That, with a 10-aperture Nipkow disk, would give > me a nice HUD I could afford easily...) Out of an *Un-*Laser printer you mean? I've never seen their internals, but AFAIK the scanners are of course only in *real* laser printers (thus the difference in cost), and all facets of the octagon mirror are aligned, but I have a few in the shed here so they are apparently available. Not as small as you want though... (You presumably mean helmet display?) > Going to have to balance torque out, or just assume that things get > weird when you move while computing... I'm not 100% sure you *can* balance the torque fully... Anyway, I'm all ears for any absurd claims in the above to be refuted. -- Cheers, Paul B.