On Sat, 2012-03-03 at 09:40 +0200, Anders Wallin wrote: ... snip > IIRC what can cause a sudden death of the IRAMS is overcurrent for > some reason or another. > To make your design fool-proof you need to sense current, possibly on > all three motor phases, and do it reliably and quickly in order to > shutdown the IRAMS and/or pwm-source when something bad happens. > > For permanent magnet servos the current-sensing is just > fault-protection. If the current-sensing can be made to work > accurately the same power-stage could maybe be used for induction > motors with a vector-drive algorithm. Possibly this requires > voltage-sensing on the uvw-phases also?
This document shows what I think is a kind of sample hold arrangement in the lower right corner. http://www.jrkerr.com/pssc_irams.zip I was thinking of using an AVR analog input to sense the IRAMS internal sense resistor then use software to look for current peaks and separate the leg currents if needed, but this is much easier to plan, than to make real. > With higher voltage/current the IRAMS will require cooling. I used an > old cpu-heatsink+fan. You might want to put the psu/power-wiring for > this on the IRAMS-PCB also. I'm not seeing your point here. I have a couple of 1/2 HP VFD's here I can study and plan on copying their layout with a diode bridge and set of caps feeding the DC bus. I'll feed it from a transformer for the lower voltage study, then try mains input latter. I was a little surprised to see the outside of the module is all plastic which isn't the best heat conductor, but I suppose good enough in this case. I suppose one really needs to make sure this IRAMS operates at its highest efficiency. > Optoisolation between the power-stage and the fpga-logic would be a good idea. ... snip The datasheet indicates that the chip is already isolated. Since isolation can slow the signals I had not planed on additional isolation, but I see from the datasheet that the sense resistor is internally tied to Vss or logic ground. The thermistor is tied to Vss so it may need to be isolated. Itrip is on the other side of the sense resistor, so it needs to be considered too. This issue might not be easy. Maybe I can process these signals on the AVR and isolate the communications between the AVR and the PC. -- Kirk Wallace http://www.wallacecompany.com/machine_shop/ http://www.wallacecompany.com/E45/index.html California, USA ------------------------------------------------------------------------------ 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
