I am very aware of the pin 1 problem and why you would assume that there
would be common mode current on the green wire. I also expected that
the green wire should pass through the choke but with the 2 different
grow light ballasts this was not the most effective solution. Leaving
green wire untouched and having a good common mode choke on the hot and
neutral followed with capacitors to green wire ground was the most
effective filter for these devices. This was significantly more
effective than putting 8-10 turns of all 3 wires through a large #31
toroid. Initially we focused on 40m since that was the worst band for
I can only report what has been successful for several high power
electronic ballasts which exceed FCC part 18 limits by more than 40dB.
I will carefully look at the W4EF solution you referenced.
On 3/20/2012 12:32 AM, Jim Brown wrote:
> On 3/19/2012 9:29 PM, Larry Benko wrote:
>> Ferrites alone were not
>> adequate to eliminate the RF interference.
> Hi Larry,
> Thanks for documenting and sharing your work on this. One red flag that
> I see in at least some of the photos is that the Green Wire bypasses the
> common mode choke. It has been my experience that the Green Wire is
> often (usually) part of the common mode circuit, and must pass with the
> other conductors through the choke.
> As an example of this issue, a few years ago I fitted very good
> conventional line filters into boxes and connectorized them (very short
> leads) for use with some Honda generators for Field Day and California
> QSO Party county expeditions that are noisy on 20M-10M. The filters
> didn't work at all, because the trash was radiated on the Green Wire. I
> lengthened the line (portable 20A cordage) between the generator and the
> filter enough to wind about six turns through two #31 toroids and that
> killed the noise.
> Note that W4EF shows exactly that solution on the following page for
> noise on his Honda generator. Mike is an EE at JPL.
> BTW -- the REASON that the Green Wire is part of the problem is a "Pin
> One Problem-like" error in the connection of the Green Wire at the noise
> source. The same sort of error also couples RF trash out of TV sets
> onto coaxial wiring that we THINK is bonded to the chassis but is not.
> Notice also that both Mike and I were working on noise in the 14-28 MHz
> range, so we both settled on chokes resonant in that range. He used #43
> because that's what he had, and also because it's the best choice above
> about 20 MHz. I used #31 because that's what I had, and because it's
> almost as good in that range. We would both use more turns for noise on
> the lower HF bands.
> Another BTW -- what a Power System engineer calls common mode is current
> that is equal on phase and neutral. But from an RFI point of view, all
> three conductors are part of the antenna, and if the Green wire and
> internal bonding are not proper, we must choke all three conductors (or
> four in a 240V system with a neutral).
> 73, Jim Brown K9YC
> RFI mailing list
RFI mailing list