[RE-wrenches] Long transmission line inductance and LED light flicker

[Exeltech]

Troy,

You're correct in your statement about dimmers operating on a phase shift within their own circuitry.  However, that isn't likely the culprit here.

1000 feet of 350 MCM would have approximately 0.65 millihenry inductance per conductor.  If the conductors are widely separated in a manner that the respective magnetic fields have minimal interaction, this inductance value essentially doubles.  If the conductors are adjacent, or nearly so, the inductance value is less by virtue of the field cancellation.  Parallel conductors in very close proximity to each other will have minimal inductance because the fields are 180 degrees out of phase, and thus cancel.

That said .. and presuming a "worst case" installation with the conductors far apart from each other (not likely .. but we'll pretend that's the case), the collective impedance of both conductors at 60 Hz would be approximately 0.5 ohm.  (Inductive reactances add in series.)

The remaining calculations pertaining to reactive voltage drop and phase shift depend on additional values that were not present in your post.  However, the values above can be plugged into calculations for you to determine the electrical effect.

If you were to attempt to cancel the inductive reactance in the conductors, a capacitor of more than 50,000 uF would be required.  Clearly NOT recommended.

You didn't specify a temperature, so I arbitrarily used 40C (104F) for the wire calculations that follow.  2,000 feet of 350 MCM at 40C would have a collective resistive loss for two conductors of 0.127 ohms.  Presuming a 40 amp load, this equates to a 5.1V drop in line voltage.  You also didn't indicate whether this was 120V or 240V .. but based on your query, I presumed the former.  If so, a 5V drop (and this drop would be in addition other loss in the connections, connectors, switches, breakers, and elsewhere) may be sufficient to cause a slight but visible flicker in some LEDs on dimmers.  This would be attributed to the change in time at which the diac (trigger element in the dimmer) is activated relative to time, thus turning on the dimmer at varying points in the wave form .. hence causing a difference in intensity from one cycle to the next.

As a point of clarification: aside from the miniscule resistive loss in their conductors and conductive material, capacitors do not have any real "power draw".  However, inverters can easily be destroyed by sufficient capacitance connected across their AC output.  This is not due to the reactive current .. but instead by destabilizing the inverter's output voltage control loop.

In summary .. is your problem due to inductance?  Not likely.  Check for undiscovered points of resistive loss in your circuit.


Regards,


Dan Lepinski




--------------------------------------------
On Sun, 2/2/14, Troy Harvey <taharvey at heliocentric.org> wrote:

Subject: [RE-wrenches] Long transmission line inductance and LED light flicker
To: "RE-wrenches" <re-wrenches at lists.re-wrenches.org>
Date: Sunday, February 2, 2014, 2:39 PM
 
We've got a PV system that has long lines from the inverters to the house (1000 ft or so). While the wires are properly sized (2x350MCMs), it inherently has a lot of inductance due to the line-length. What We are noticing is that dimmable LED lights flicker anytime a new load turned on, even if that load is purely resistive and the overall current draw is small (20 amps or so). What I "think" is happening is the LED dimmer circuits get their cue from phase delays, and that the inductance of the line length causes some phase jitter everytime a load is applied - thus causing flicker. Has anyone dealt with this issue successfully (other than switching to incandescent lights)? Would a static capacitor bank at the house do the trick, or do we need some type of active PFC? And if static did you have any issues with constant power draw from the capacitors?
  
Troy Harvey