IBM 4331 Mainframe 3 phase converter

Paul Koning paulkoning at
Thu May 4 11:17:09 CDT 2017

> On May 4, 2017, at 11:54 AM, Jon Elson via cctalk <cctalk at> wrote:
> ...  But, you DO need a true sine wave source, and VFDs do not produce sine waves, they put out 400 V PWM waveforms that look fine to a motor, but not good at all to electronic loads.

I wonder how true that is.  Consider that (for machines of this era) power supplies are probably transformer input, to a rectifier and then a filter.  The transformer would smooth out the VFD output pulses, and whatever is left would definitely be removed by the output filter.  The only question I can see is whether the rectifier diodes have enough reverse voltage margin to deal with whatever peaks pass through the transformer.  (The transformer itself certainly will, given normal insulation design/test practice for power transformers.)

If the power supply is a swiching regulator, the details are slightly different but the overall picture is similar.  Then you begin with a rectifier, which would have to be able to deal with the input spikes, followed by some amount of filtering.  Once past that I don't see any further issues.

An electronic circuit that looks at the incoming AC waveform directly would certainly have issues with a VFD, but I can't think of too many examples of that.  A KW-11/L is one exception that comes to mind... :-)


More information about the cctalk mailing list