On Mon, 21 Mar 2011, Jim Brown wrote:
> On 3/21/2011 7:57 PM, Christopher E. Brown wrote:
>> The green wire in the AC cable bonds the the inside of the steel chassis
>> and the circuit board, the chassis bond is on the outside. Bonding the
>> outside lug to the same gnd as the power plug corrected things for me.
> I find the description a bit confusing, BUT -- I suspect a problem with
> PAINT preventing an electrical contact between the green wire and the
> chassis. The inside of the chassis and the outside of the chassis
> SHOULD be the same point. Further, the green wire in the AC cable MUST,
> by law, go to the chassis.
> There CAN be magnetic coupling of noise from a switching power supply to
> the outside world, but that is NOT what gets fixed with the ground lug.
> In general, it is NOT a good thing to bond V- to the chassis of the
> power supply. The reason is that doing so can couple the IR drop on the
> V- lead to accessory equipment that is powered from same supply. When
> running SSB that IR drop will sound like SSB detected by an AM
> detector, which in turn will show up as distortion in the transmitted
> signal. Bonding V- to the chassis also can couple buzz from power line
> leakage currents to the audio signal.
> If I were you, I would carefully scrape the chassis at the point where
> the green wire connects to make sure it is making good contact, but I
> would remove any connection between V- and the chassis.
> 73, Jim K9YC
The green wire of the AC cord runs into the enclosure and connects to the
common point for the chassis and the circuit board ground. Hot and
neutral go to another section of the board.
This is _not_ the output V- The IOTA supplies are isolated output,
allowing for series or parallel connection of supplies by design.
The connection inside the chassis was good, from the factory on all 3 of
the IOTA supplies I have.
The bonding lug is on the outside of the same steel chassis, very low DC
resistance, but as it is on the other side of a steel enclosure high RF
The enclosure is 1/16 steel, and in very close proximity to the very large
toroidial transformer in the switcher. With the chassis DC grounded
through the internal AC grounding lead there is a great deal more low freq
RF on the outside of the chassis (note, not on the DC lines) and the AC
leads than when the outside of the chassis is also bonded to the AC
ground. This is 2 connections to the same steel chassis, both low
impedance at DC.
Seeing the note in the install sheet, trying it and noting the reduction
in noise I working theory was that due to proximity to the transformer
some amount was being induced in the outside of the chassis, largely
unable to close the loop current wise at RF. Bonding both the inside of
the chassis and the outside of the chassis dropped both the radiated and
conducted noise way down (from no more noise than a couple of "HAM"
switchers I have to can only be detected when using a pickup loop in close
to the supply)
RFI mailing list