ballasts & long conduit runs

Has anyone *ever* seen noise induced into unshielded cat5 in the real world?

Al Dykes wrote in part:

Not into Cat5, but I have seen interference radiated from some Cat3 running 100baseTX on NTSC TV channels [snow lines].

As for the OP, I doubt ballasts will cause the slightest trouble. At least half of all Cat5 ever made is run in plenum areas infested by ballasts. 60 Hz won't do anything.

-- Robert

I recall a test made by someone early in the CAT3 days where they wrapped loops of CAT3 around all manor of electrical noise inducing machinery while running a BERT test. They never showed any negative effect.

Due to great care in symmetry and balance, CATx cable and equipment should be getting more and more immune to induced noise.

I do have a lot of cat 3 to use up, and it is only for voice anyway. Replies were appreciated!

snipped-for-privacy@mhtc.net wrote in part:

Well, that is a little different. This is c._dcom_.c, we assumed data [digital] signals. Analog and digital have very different noise tolerance. Analog accumulates noise, digital ignores noise until it breaks. 60 Hz is within hearing range. Cat 3 is good wire (well twisted), but I would try to keep it 6-12" away from point radiators (ballasts and lamp leads) to let the 1" twists work and balance out hum. No worries about the longer AC line. The twists will balance it out.

-- Robert

The point of the ballast is to take the 60Hz and create much higher somewhat low power frequencies to excite the gas inside the tubes. And cheap ones can really generate some EMI noise.

It's a good idea to avoid them but not go nuts with 6' separations and such.

David

If the "do", I bet its noise induced into the power line leaking into the system

Referring to the new electronic or "solid state" ballasts. The older ones simply were current limiters.

Good point. The older [heavy] ballasts didn't go above

I'm not an electrician. Has anyone seen switchers on industrial 4' or 8' fixtures?

-- Robert

(snip)

In both cases the goal is current limiting, but the way it is done is different. At 60Hz it takes a large inductor to appropriately limit the current. It would take a gigantic capacitor if you wanted to do it that way. At 20kHz or so it can be done with a nice small capacitor, and any associated vibrations are not audible.

Also, the usual PC power supply runs at around 20kHz to use a nice small ferrite core transformer instead of the 30 pounds (guess) of iron that would be needed at 60Hz. The availability of transistors that can switch 300 volts at a few amps at 20kHz is the breakthrough that led to such power supplies. (the breakdown voltage will be somewhat higher than 300.)

Properly balanced UTP cable won't pick up much differential mode signal, though.

-- glen

I have heard of weird interactions between such and PC displays; I don't know the details...

Robert Redelmeier wrote: (snip)

Well, there is still some effect as the arc restarts on each half cycle. Some will be filtered out by the inductor, though.

I thought so, but I haven't looked lately. You can tell by weight. Also, 60Hz (magnetic) ballasts cause a strobe effect at 120Hz, which sometimes is not desired. (Especially around circular saws with induction motors.) Electronic ballasts have a much higher frequency, so you don't notice the effect.

Here are some I found from Google:

-- glen

It is pretty much the same way that PC power supplies work:

Voltage double to 300VDC (or bridge rectifier for 240VAC input). Switching transistor at 20kHz into a ferrite core transformer.

I would expect any such interaction to be just as likely with a PC power supply, though of course any device can fail and not work as designed.

-- glen