RadioBanter - View Single Post - RFI: Odd consequences of variable speed AC motors.

Telamon · December 4th 07, 03:57 AM posted to rec.radio.shortwave

In article
,
wrote:

On Dec 2, 7:52 pm, Telamon
wrote:
In article
,

wrote:
Robert Blum and I discussed this last summer and I had forgotten
all about it until I got a call last night from a fellow SWL who
was experiencing some sudden onset RFI. He had been bragging
about his new Panasonic variable speed heat pump and how much it
had saved him this summer. I asked him to take a MW transistor
radio near his fancy new heat pump and to let me know the
results.

Surprise surprise. The noise was most intense at the exterior
heat pump. . We had a very odd situation at work in that the
bearings in the vari- speed motor that drives the ventilation
system, a very big motor and fan, kept going through bearings at
a frightful rate. The bearings are supposed to be good for 10K
hours. we were getting as little as 6 weeks.

In one of my searches I had found a reference to the longitudinal
currents induced in the rotating shaft. This current wants to
flow from end to end and must punch it's way through the
lubrication. This destroys the lube and eats tiny wholes in the
bearing surface. A very bad situation. I gained permission to go
into the service area, an area off limits to even engineers with
my trusting DX398 and the 19uH noise probe. There was an intense
amount of static coming from the motor. Our electrician didn't
call me an idiot, but he stiffly agreed to check the voltage from
end to end on the shaft. It was well over 100 volts. After he
pried the dust caps off, and when we turned out the lights, you
could see a million little arcs at each end in the bearing!

The solution was to use ceramic coated steel bearings. Noise
gone, And no more bearing replacements.

The only HF reception gear where I work is a WWV clokc backup,
3rd level redundancy, that I monitor at my bench. I had noticed
periods of noise strong enough to unlock the WWV derived clock,
but hadn't been able to locate the noise source.

The audio guys were very happy because their long AES digital
audio liines were experiencing serious, as in unusable, noise
issues.

I wonder how many of the new high efficiency heat pumps and air
handlers are going to have some rather odd problems.

My friend is trying to figure how to explain what is happening to
the warranty people. I don't envy him. If work was a valid
example, his bearings ought to fail pretty soon.

And before you ask, reducing the RFI from this unit was a
nightmare. It took several hundred dollars of heavy duty RFI
filters, a lot of ferrite and a lot of bypass caps with a local
grounding rod and 4AWG bond back to the home's NEC ground.

You post about the strangest stuff. I have never heard of this kind
of problem. One or more of the rotor windings must be some how
shorted to the shaft.

Ceramic bearings are not the solution as the current can go through
whatever gear train and mechanical load to which it is connected.
The motor shaft should not be electrically hot. That would be a
shock hazard. That motor is either built wrong or designed wrong.

I knew I had a good link"
http://www.motionsystemdesign.com/Issue/Article/47423/Diffusing_current_trouble.aspx

This problem is not imaginary or all that odd. An example of a new
technology that is widely adopted and applied before all the bugs are
worked out. There are solutions. But the best solutions start with a
motor designed for variable speed operation, and by installing the
inverter very close to the motor.

And yes, under some conditions even ceramic bearings aren't an
effective solution. They worked for us, but other steps were also
taken: Inverter moved to within inches of the motor. Some very heavy
duty LC and RC 'snubbers' where installed. Serious RFI filtering was
installed between the inverter and the motor. And even more serious
heavy duty RFI filtering was installed on the AC mains. The AC mains
cable size, especially the neutral, was upgraded several wire sizes.
A kelvar fiber insulated "V belt" has to be used because standard
rubber was too conductive. And the who rig has a special rail around
it to keep people from touching the pulley.

That area was off limits before and it is doubly so now. Everyone who
enters must be on a approved list, log in and out, have training and
no one can go their alone. I wonder if the energy saving are worth
the extra heartaches.

That's a very interesting problem. This quote from the link you posted
sums it up "A normal ac power supply is symmetrical, so that its
three-phase vectors sum to zero. But variable frequency drives turn that
into dc current, and then chop it into power at rates to 20 kHz. At
those high frequencies, constant, perfect phase balance is impossible ‹
and the phase vectors sum to a nonzero voltage."

There still could be an solution in the controller that could balance
the fields and eliminate the shaft currents but you might end up losing
some of the efficiency.

--
Telamon
Ventura, California