On Saturday, March 03, 2012 01:46:58 PM Peter C. Wallace did opine: > On Sat, 3 Mar 2012, gene heskett wrote: > > Date: Sat, 3 Mar 2012 12:20:02 -0500 > > From: gene heskett <ghesk...@wdtv.com> > > Reply-To: "Enhanced Machine Controller (EMC)" > > > > <emc-users@lists.sourceforge.net> > > > > To: emc-users@lists.sourceforge.net > > Subject: Re: [Emc-users] IRAMS Plan > > > > On Saturday, March 03, 2012 12:03:01 PM Przemek Klosowski did opine: > >>>> Anders, can You suggest, how to implement that (preferably a simple > >>>> and easy way, so that even child or golden retriever :) would > >>>> understand)? The scheme in my IRAMS datasheet suggests using Shunt > >>>> resistors for that, but I suspect that something else should also > >>>> be needed, but I have no idea, how these things actually work. > >> > >> In principle you could measure the current by attaching current > >> transformers or Hall sensors to the motor leads, avoiding the > >> high-side and high common mode voltage issues. The downside might be > >> speed/frequency response. What do people think about these > >> alternatives to resistive shunts? > > > > Excellent idea. Current transformers can have frequency response > > problems depending on the magnetic materials used. but are pretty > > accurate. Hall effect stuff is probably faster but from my > > experience, not as linear. > > > > For this duty, either should be more than fast enough. One of the > > side effects of the current transformer will be a slowing of the > > Di-Dt under fault conditions, which may give enough time for > > shutdowns to be done. To me its advantageous when a $1 fuse blows > > instead of the $500 circuit. :) > > > > The current transformers output voltage will be very high under fault > > conditions, so figure on some sort of over-voltage protection in the > > circuitry its feeding so it doesn't get blown by a momentary carbon > > track as a June bug gets to meet its maker. > > > > Oh, and Kirchoffs applies to the delta connected circuit. Barring > > ground faults, sensing any 2 is sufficient. But around machinery, > > and in the real world, ground faults DO happen and I would design > > accordingly. IOW, sense all three phases. > > > > Cheers, Gene > > Note that current transformers will not work on PMSM (permanet magnet > rotor motors) as these can have DC drive current when applying constant > torque when stalled.
I had forgotten that (can I blame it on oldtimers?) detail Peter, many thanks for bailing me out. I was thinking only in terms of the AC input power, which can put a pretty decent energy sink in the path, and slowing the response since a fault would have to dump the sink before it would show up in the input currents. That sink dumping can easily turn semiconductor stuff into lid-less smoke generators. :( However, and back to "load" sensing, I was under the impression that stalled torque in these was probably limited by the duty cycle of a pwm drive, and that should still give a signal. What that signal means could be subject to interpretation of course since the AC isn't going to represent even a fraction of the DC flowing when the DC is largely the free-wheeling current flow when the pwm is in the off state. In that case, would it not make more sense to pass the low side currents thru a fullwave bridge with a very small R as its load so that an led would light up and initiate the shutdown when it comes on? I have an ammeter I made that works on that idea so I can see how hard my spindle motor is working since its control is so stiff, the rpm change from no load to blowing a fuse is perhaps 1%. > Normally you need to measure 2 phase currents. Common way to do this is > with Hall effect devices in the motor leads (this works because the > Hall effect devices are isolated so the 360V PWM voltage does not > affect them) > > The other way to measure phase currents is with 3 current shunt > resistors in the three bottom IRAM 1/2 bridge legs (on the modeules > that bring the individual emitter pins out). Its more difficult to get > good current reading here as the current is only valid when the bottom > leg of the shunt resistors 1/2 bridge is on (and the shunt voltage must > be measured when switching transients have died down). This leads to > interesting code to select the best 2 shunt resistors to sample > depending on the drive phase angle and polarity. > > IF you are not using shunt resistors for current control (FOC) you can > use the IRAM modules with a built-in single shunt resistor (and > assosciated comparator and shut down logic) in the bottom leg of the > IRAM module for over current protection. > > For a simple voltage mode device this may be enough protection. > > As gene mentioned, this will not protect against U,V,W to ground shorts > that normal requires fast sensing in U,V,W motor wires or both high and > low side VBUS current sensing And my idea of a FW bridge with a very small R for output load driving an led would still give plenty of isolation, and that could detect a ground fault in a few nanoseconds if on the high side lines. But power SI stuff is also notoriously slow, so that few nanoseconds could well be several microseconds, so unless the bridge is carefully chosen, the fast response may not be as doable as calculated. You are I suspect in a far better position to set that up and measure it than I. > Note that the modules are NOT isolated, all the gate drive and fault > signals are relative to the negative VBUS terminal so you need to > provide isolation. (I think 3750V isolation from SELV signals is a > legal requiremant for 240V line operated equipment) > > Rather than try and isolate all the analog (VBUS,phase currents), gate > drive, fault, and temperaure sense stuff (something like 12 signals in > our case) , on our 3 phase drives the processor ground runs at VBUS- > and the isolation is just in the 2 serial comm lines. > > > Peter Wallace > Mesa Electronics > > > ------------------------------------------------------------------------ > ------ Virtualization & Cloud Management Using Capacity Planning > Cloud computing makes use of virtualization - but cloud computing > also focuses on allowing computing to be delivered as a service. > http://www.accelacomm.com/jaw/sfnl/114/51521223/ > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users Cheers, Gene -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) My web page: <http://coyoteden.dyndns-free.com:85/gene> Yes, but will I see the EASTER BUNNY in skintight leather at an IRON MAIDEN concert? ------------------------------------------------------------------------------ Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
