Topband: Elimination of Treadmill RFI on 160 meters

[Brad Rehm]

Don,

You make a good point about the difference between our definition of CM
noise and the definition implied in Part 15.  If the external torroid made
a difference for you, then it's important and worth having.

We can't be too optimistic about the manufacturer's interest in fixing this
kind of problem.  He won't be willing to install a $50 or even $25 filter
to satisfy a few hams who have EMC complaints.  At the same time, because
the products you and I have dealt with responded so well when we added
filters, we have to wonder if and how the Part 15 compliance testing was
done.

In the EMC lab where I worked, we tested them under load (with someone
walking on the treadmill or with a fixed load applied), but some labs might
be tempted to test only the uP controller.  It's hard to imagine how this
could be justified, because the belt is usually driven with "square"
pulses.  RF noise from these can be hard to suppress.

I didn't think to check the label for the machine I worked on, but I wonder
if you've looked for an FCC compliance mark on your treadmill.  If it has
one, it might be interesting to go to the FCC URL and look at the report
that was filed for it.  And BTW, if the treadmill was manufactured in the
far east, the mark could be bogus or filtering could have been deleted
after the initial testing was done.  This kind of thing isn't all that
unusual.

73,
Brad  KV5V

On Tue, Jan 27, 2015 at 4:05 PM, Don Kirk <wd8dsb at gmail.com> wrote:

> Hi Brad,
>
> Great to hear you had similar results.
>
> The 20VR1 and the 20DRGG5 filter appear to be of very similar design (two
> stage filters, and component values not vastly different).  The 20VR1 looks
> to be double the price of the 20DRGG5, but it might offset some of the
> additional cost if it allows elimination of the additional Toroid core
> choke.  Since all cases are different I would have to test the 20VR1 in my
> application to see if it would indeed allow elimination of the Toroid core
> choke, but I doubt I would be able to eliminate the Toroid core choke for
> the following reason that Jim K9YC mentioned on another reflector : "What
> the power industry calls "common mode" is NOT what we call common mode.
> They are talking about the voltage between neutral and the green wire. We
> are talking about current that is flowing in the same direction on three
> conductors."  In my case the additional Toroid core choke had a very
> noticeable impact, and I now can't tell when the treadmill is on when using
> the combination of both filters.
>
> P.S. I sent an e-mail to the manufacturer of our Treadmill, and they never
> responded.  I am now going to call them on the phone in an attempt to get
> in touch with their design engineers, and likely will also file a complaint
> with the FCC regarding treadmills in general.  There are a couple more
> treadmills near my home that generate very strong RFI, but fortunately they
> normally use their Treadmills during the day, and I operate mostly at night.
>
> 73,
> Don (wd8dsb)
>
>
>
>
> On Tue, Jan 27, 2015 at 1:26 PM, Brad Rehm <bradrehm at gmail.com> wrote:
>
>> Don,
>>
>> FWIW, today, W5UJE and I dealt with a treadmill problem similar to yours
>> by installing a commercial line filter between the line and the treadmill
>> electronics.  A Corcom/TE Connectivity 20VR1 filter reduced the 40m and 75m
>> noise at his receiver from S9+10db to less than S2 (his noise floor this
>> morning).  The 20 Amp rating was needed because the manufacturer's
>> published current requirement for the treadmill was 15 Amps.
>>
>> In measurements we'd made earlier, we found that this was a common-mode
>> problem and that the noise spurs were about 20 kHz apart when the treadmill
>> was operating under a moderate load.  The spectrum scope on his radio
>> showed 20 kHz-spaced broadband noise up through 29 MHz, peaking between 1.8
>> and 15 MHz.  The line filter we chose offers 10-20 dB of supperssion below
>> 50 kHz and 60-80 dB of suppression between 300 kHz and 29 MHz.  No
>> additional filtering with torroids or capacitors was needed.
>>
>> In other words, our results were similar to yours, and one wonders how
>> manufacturers can say these things meet Part 15 requirements for conducted
>> emissions.
>>
>> Brad  KV5V
>>
>> On Mon, Jan 26, 2015 at 7:22 PM, Don Kirk <wd8dsb at gmail.com> wrote:
>>
>>> Treadmill combination filter update.
>>>
>>> Today I replaced the 10 amp commercial filter with the 20 amp version of
>>> the filter, and the results are similar.  Below are a few measurements on
>>> 160 meters and 80 meters showing how effective the combination of the two
>>> filters are (14 turn toroid choke, and commercial filter model 20DRGG5
>>> made
>>> by Delta) in reducing my treadmill RFI.
>>>
>>> *1.8068 Mhz*
>>> No Filters : 19db over S9
>>> With Filters : S7 which is my noise floor
>>>
>>> *1.8291 Mhz*
>>> No Filters : 15db over S9
>>> With Filters : S7 which is my noise floor
>>>
>>> *3.5250 Mhz*
>>> No Filters : 28db over S9
>>> With Filters : S6 which is my noise floor
>>>
>>>
>>> *3.5475 Mhz*
>>> No Filters : 25db over S9
>>> With Filters : S6 which is my noise floor
>>>
>>> The 3.5250 Mhz readings indicate the filter is knocking the signal down
>>> at
>>> least 46db (and probably more).
>>>
>>> 73,
>>> Don (wd8dsb)
>>>
>>>
>