Re: VFDs, Noise, and RS-485

On Sat, 30 Jul 2005 12:55:56 +1200, Terry Given <my_name@xxxxxxxx>
wrote:

>The Real Andy wrote:
>> On Fri, 29 Jul 2005 17:43:58 +1200, Terry Given <my_name@xxxxxxxx>
>> wrote:
>>
>>
>>>The Real Andy wrote:
>>>
>>>>On Thu, 28 Jul 2005 12:48:22 +1200, Terry Given <my_name@xxxxxxxx>
>>>>wrote:
>>
>> <snip>
>>
>>>>I have read here before about this business of putting different
>>>>signals into the drive. Why is this? I think last time someone
>>>>mentioned (sphero perhaps) injecting a 3rd harmonic or something of
>>>>the like.
>>>>
>>>>Andy
>>>
>>>the idea is to continue increasing the RMS output voltage when the peak
>>>is fixed - clipping aka overmodulation.
>>
>>
>> Ahh, interesting. Do you know of any papers on this? One would expect
>> that 'over modulating' would have on adverse affect on the performance
>> of the motor, ie heating. I guess this is not the case though if it is
>> common practice.
>
>hundreds, if not thousands.
>
>>
>>
>>>many digital algorithms like Space Vector Modulation (SVM) automagically
>>>do this. Many digital algorithms dont, and the 3rd harmonic (and various
>>>other stuff) gets tacked on as an afterthought.
>>>
>>>In the case of naturally sampled PWM (3ph LF sinusoidal voltage feeding
>>>3 comparators along with an HF triangular carrier wave), the harmonics
>>>get added in to the voltage reference. Back before micros, that was
>>>really the only way to do it - I have seen a paper describing a direct
>>>field oriented controlledr done entirely in analogue, circa 1974 IIRC.
>>>
>>>
>>>Cheers
>>>Terry
>>
>>
>> I am building my first VFD for my father who is a retired engineer. So
>> far so good. I am running motors, albeit, unloaded atm. Using the new
>> IR iMotion IGBT module and a pic18f4431. I often wondered about the
>> third harmonic theory. I guess for driving a 3phase lathe motor from
>> single phase its probably not worth the effort. :)
>
>Space Vector Modulation is by far the easiest to do digitally, and
>automagically takes care of overmodulation. Retter's Matrix and Space
>Phasor Theory of Electrical Machines is a darned good reference, and a
>lot cheaper than anything by Peter Vas.

I did do some reading on this concept. There is a good note on
Microchips website.

>
>The basic idea is to convert the 3-phase rotating vectors into an
>equivalent single rotating vector (can do because of fixed 120 degree
>relationships). This vector is broken into x- and y- components (called
>a,b). The observer then rotates at the same frequency, generating a
>stationary (wrt obs) pair of coordinates d,q. Align to one axis, the
>other is by def. zero. I'll see if I cant dig up a decent SVM paper...
>

If you can find that paper that would be great.

>Cheers
>Terry

Thanks:

Andy
.

Relevant Pages

  • Re: VFDs, Noise, and RS-485
    ... >The Real Andy wrote: ... >the idea is to continue increasing the RMS output voltage when the peak ... >3 comparators along with an HF triangular carrier wave), the harmonics ... I guess for driving a 3phase lathe motor from ...
    (sci.electronics.design)
  • Re: VFDs, Noise, and RS-485
    ... >>>the idea is to continue increasing the RMS output voltage when the peak ... >>>many digital algorithms like Space Vector Modulation automagically ... I guess for driving a 3phase lathe motor from ... vailid email address, after replacing the obvious in brackets. ...
    (sci.electronics.design)
Quantcast