DAVID CUTHBERT wrote:
> Jim,
>
> thanks for the great info. I design motor drive filters for three
> phase motors and find that common-mode filtering can help.
>
> My preferred motor filter is: Motor driver circuit - series inductor -
> shunt C with series R - then CM inductor. Y-caps after the CM inductor
> sometimes are not good if the local GND reference is noisy. The series
> R (damping R) is close to the sq rt of L/C.
>
> For long lines such as the well motor we want to source terminate the
> line. Making the source look like 100 ohms above 1 MHz for example
> will keep the line from ringing. This can be taken car of with the
> shunt C damping resistor if the value works out. If not, the first LC
> can be followed by another LC having a 100 ohm damping resistor.
>
> And as you say, keep the loop area as small as you can.
>
> Dave WX7G
> Who is John Galt?
>
>
> On 5/3/09, *Jim Brown* <jim@audiosystemsgroup.com
> <mailto:jim@audiosystemsgroup.com>> wrote:
>
> On Sat, 02 May 2009 12:50:28 -0400, GEORGE WALLNER wrote:
>
> >From an
> >RFI point of view they are weapons of mass destruction!
>
> Correct on all counts. To understand the problem, study the
> section on variable frequency drives in the RFI tutorial on my
> website. The primary mechanism is the magnetic field generated by
> that current, and the fact that the current loop area is
> physically large. For any magnetic field, the field strength is
> proportional to the current multiplied by the loop area. The loop
> area is large because the return path for the HF current (the
> harmonics) is through bypbass capacitors to "ground," and because
> the well motor is separated from the controller and the stepdown
> transformer.
>
> There are several elements to the capacitance. First are the stray
> capacitances of the transformer and the motor. Second are bypass
> capacitors at the wrong place in the circuit. You WANT bypass
> capacitors next to the switching transistors so they create a very
> small loop area for HF current, but NOT at the motor.
>
> Two important steps to fix this. They are additive.
>
> 1) Slow down the rise time of the current pulses. There should be
> an RFI kit from the mfr to do this. Some big ferrite cores may
> help with this, but they must be BIG so that saturation doesn't
> kill their effectiveness. In this application, they go around a
> single conductor to form a differential choke, NOT a common mode
> choke.
>
> 2) Reduce the loop area. You can use twisted pair for the phase
> conductors, but that won't help with the harmonic current that's
> returning on ground.
>
> This is completely separate from what the controller may be
> putting back onto the AC line. A conventional line filter will
> kill that. It should be bonded to the controller.
>
> 73,
>
> Jim Brown K9YC
>
> Gents
>
Gents
The gas company has now installed shielded cables. As far as I know that
is all they've done so far. The shielded cables had no effect on noise
amplitude.
While RG8 etc could be useful in some cases this isn't one. The motors
and wiring are way too large for nostrums like RG8 and 2 inch cores, on
the order 25 horsepower. So, to help the gasco toward a fix I'm seeking
a hint on what next. Move the controllers adjacent to the pumps? (I
suggested that as they were about to install the shielded cable, fell on
deaf ears.) More output filtering at the controller?
The gasco guy told me "hand me $40,000 and I'll put the gas powered
motor back". Great idea but for the $40K!
73 Art
_______________________________________________
Topband mailing list
Topband@contesting.com
http://lists.contesting.com/mailman/listinfo/topband
|