On 22.06.11 15:19, andy pugh wrote: > I think this circuit ought to work > https://picasaweb.google.com/bodgesoc/Gibbs#5621046688941452194 > Which is something I have meant to try for a while. > I have a feeling I have drawn it before, and that time I was more > confident in it, I think I have left out some critical resistors.
Andy, the other circuit undoubtedly has some benefits over this one, which would benefit from a minor tweak or two before firing up the soldering iron. To make this one work, it is necessary to drive the input between -2v and +7v, for forward conduction of the LEDs plus a little for the (undrawn) current limiting resistor. That's not as easy as using TTL or CMOS levels. (OK, I see below that they're intended to be connected the other way around. To make that work, the location and values of resistors do become critical.) Would another (less critical) circuit variant be of interest? For ease of driving, we could parallel the gates of a CMOS hex inverting buffer (e.g.4049) in two sets of three parallel inverters (for increased current), then series connect the two composite inverters, and finally string the anti-parallel connected optocoupler LEDs and a series 330 ohm resistor across input and output of the second composite inverter. Drive the input of the first inverter, and we have H-bridge drive of the LEDs, followed by your H-bridge output. The >3v dead-band now provided by the LEDs would allow quite small current limiting resistors to be used in each opto's output H-bridge leg, instead of the enormous 15k value, which limits the available output amplitude, unless the circuit is followed by an impedance buffer. (Such as a voltage follower.) Good steep flanks on the input waveform are advisable. > The idea is that it converts a 5V PWM to a completely isolated > voltage, which could be to a different reference, and of a different > magnitude. I imagine connecting it in place of the pot on a VFD for > example, or to the 100V floating pot terminals on a KBIC controller. > > It is effectively half of a H-bridge, +5 PWM turns on an opto that > discharges the capacitor, and 0v PWM turns on an opto that charges the > cap. To avoid the need for an op-amp to buffer the output voltage, it'd be good to reduce the H-bridge leg resistance, as described above, I think. > As-drawn there is a problem that unconnected PWM turns on both optos... Well, not quite, as drawn. The opto-isolator LEDs are reverse biased with no input, and so are as off as it's possible to be. If they were connected the other way around (so that they are on), they'd need one of the mentioned "critical resistors" in series, to avoid dead optos. (And careful driving.) Accurately transmitting an analogue signal across an optoisolator is a challenge, but I suppose we don't need that if we can measure motor RPM for our feedback. Hopefully these ideas are of some use. (Even if only to trigger yet more.) Erik -- "The difference between theory and practice is much smaller in theory than in practice..." ------------------------------------------------------------------------------ Simplify data backup and recovery for your virtual environment with vRanger. Installation's a snap, and flexible recovery options mean your data is safe, secure and there when you need it. Data protection magic? Nope - It's vRanger. Get your free trial download today. http://p.sf.net/sfu/quest-sfdev2dev _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
