How do large power supplies work? (VFD's etc.)

I've looked at the designs of switching power supplies (buck/boost/buck-boost) and I basically understand how they work. But is this the same thing used in large power supplies such as those in the VFD of a tower crane (etc.)? Like the larger of the two hoist winches offered for the Liebherr 710 HCL 25/50 tower crane, which uses a 220 kW (294 H.P.) AC motor. How does the power supply maintain DC output while the motor is bogged out to ~18% of full speed?** Do large VFD's and whatnot still use a conventional buck (boost) switching converter to get the required "rail voltage"?

Note that 480 Vrms coming in gets you approx. 680 VDC through the rectifier and reservoir capacitors.

**Here's part of the performance data if anyone wants to know:

0 - 3450 kg*** at 0 - 238 m/min (0 - 6900 kg at 0 - 119 m/min)

0 - 25000 kg at 0 - 43 m/min (0 - 50000 kg at 0 - 22 m/min)

***238 m/min only goes up to 1800 kg on 1-part line if the reduced speed curve is used.

• Google AC drive. You can scale it up to as high power as you would like given enough rating on your transistors, inductors and capacitors. – winny Aug 14 '17 at 14:24
• Why would a tower crane have a Vacuum Florescent Display (VFD)? – Olin Lathrop Aug 14 '17 at 14:42
• Variable Frequency Drive – El Ectric Aug 14 '17 at 14:44
• You don't control a rail voltage. Motor voltage is a function of speed and that one depends on the mechanical speed. And that one is a sum function of the torque. So you have to control the current to gain control over the torque a motor applies to the mechanics. Then control the speed by adjusting the current set-value. Then control the position by adjustion the speed set-value. – Janka Aug 14 '17 at 14:57
• It doesn't matter whether it is an AC motor or a DC motor. The fundamental connections are U ~ v and M ~ I. Three-phase AC motors draw contant power as a DC motor does. It's the same mechanical-electrical connection. Single phase AC motors are a bit more complicated since they cannot draw contant power. But your crane most likely doesn't use the latter apart from motor cooling fans drives. – Janka Aug 14 '17 at 15:15

You don't want to control a "rail voltage". You want to control current. Why? Because the fundamental connections between electrical and mechanical world are

$$P_\mathrm{el} = I \cdot U = M \cdot \omega = P_\mathrm{mech}$$

$$U \sim \omega$$

$$M \sim I$$

It's as great as it is simple.

For your crane example this means, depending on the load on the hook, your motor needs a certain torque – a certain current – to keep that load at an equilibrium of constant speed.

That's why you want to measure and control the current. The current depends directly on the load applied. Speed doesn't, voltage doesn't.

• So it doesn't need a complicated switching regulator to supply a 480VDC rail that supplies MOSFETs or IGBT's? Then how Liebherr can achieve "millimetric precision" is still a mystery to me. nrcan.gc.ca/sites/www.nrcan.gc.ca/files/energy/images/… – El Ectric Aug 14 '17 at 15:51
• It needs an even more complicated switching regulator which controls motor current. Not voltage. And a speed regulator on top of that. And a position regulator on top of that. – Janka Aug 14 '17 at 15:53
• OK. I'm starting to get it. – El Ectric Aug 14 '17 at 15:54
• On some of the big VFDs that I've seen, the 480V input just goes to a SCR bridge rectifier and then a capacitor bank. – W5VO Aug 14 '17 at 16:52
• Oh. Just like a normal switching power supply where all the low frequency ripple is "chopped up" and becomes high frequency. The six transistors chop up the unregulated DC rail voltage so it doesn't really matter. And I'm assuming "SCR bridge rectifier" means an ACTIVE rectifier. – El Ectric Aug 27 '17 at 2:05