My outdated brush DC motor controller wisdom is that you have a current loop controlled by a speed loop. The current loop is set up to cancel out the effect of the R(esistive) L(inductive) effects of the rotor inductance. and the speedloop is a staightforward PI loop. This makes for a very stable and easy tunable motor drive. In fact I have never set up a pure speed loop (i.e. voltage control) drive in all these years. I know they have been popular in the US for smaller setups (fractional horsepower) So I would say that ultimately also a 3phase drive will have a current loop to negate the RL factor of the winding. Interesting to see how the lag is accounted for since the speed is done by the field frequency. Must be by torque control i.e current control. Alternatively just jam it with power like in a microstepping setup.
Greetings. Jan On Wed, Apr 28, 2010 at 6:41 PM, Peter C. Wallace <p...@mesanet.com> wrote: > On Wed, 28 Apr 2010, Jon Elson wrote: > > > Date: Wed, 28 Apr 2010 11:16:14 -0500 > > From: Jon Elson <el...@pico-systems.com> > > Reply-To: "Enhanced Machine Controller (EMC)" > > <emc-users@lists.sourceforge.net> > > To: "Enhanced Machine Controller (EMC)" <emc-users@lists.sourceforge.net > > > > Subject: Re: [Emc-users] Microcontroller motor drive [Was: Resolver to > > Quadrature Convertor] > > > > Andy Pugh wrote: > >> I just found this application note from Atmel > >> http://www.atmel.com/dyn/resources/prod_documents/doc2592.pdf > >> > > The problem with all of these designs is they have no current loop > > control. I could be wrong, but I think you really need closed-loop > > current (torque) control for this to work well. If you do this in > > voltage-mode control (just adjusting the PWM duty cycle without sensing > > motor current) then I think you will find that the loop becomes very > > hard to tune near zero speed, where the requirement for PWM duty cycle > > will be VERY low. This is how my drives work, but since they are only > > doing trapezoidal drive, you don't have to impose the sinusoidal drive > > waveforms on top of that already small pulse width. > > > > Jon > > > > > We've (and our customers) have run motors in sinusoidal voltage mode with > our > 7I39 with good results for many years, so it is possible. Its tougher with > IGBTs than MOSFETS since the IGBT deadtime is a so much greater proportion > of > the total on time. This requires good deadzone compensation so the actual > motor voltage is more linear near zero (The pulse width is ~50% at ~0 > voltage, its just the overlaps that get small) > > Voltage mode actually has some advantages over current controlled mode: > > Better inherent high frequency damping (driver has low impedance) > > Quieter (Current control loops invariably hiss and whistle) > > > balanced with some disadvantages: > > No torque limit/control other than overall current limit > > Lower bandwidth (inductance feed forward can help here) > > > Peter Wallace > > > ------------------------------------------------------------------------------ > _______________________________________________ > Emc-users mailing list > Emc-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/emc-users > ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
