[Yoyizit]

10A running through 10' of #12 AWG copper Romex should give you a max 320 mV drop in voltage when the 10A load is switched on. This is 1.6 milliohms per foot for copper (~2.5 for aluminum of the same gauge) with 20' loop distance.
Run the numbers based on your distances and conductor material and decide if your sump pump amperage with good wiring and connections can account for the voltage dropoff you are seeing.
Pump startup amps might be 6x the steady-state value.

I'd think CFLs with their built-in electronic power supplies are way less sensitive to input voltage variation than are incandescents.

 

[micromind]

Most sump pump motors are split-phase start. They have no capacitor.

The starting current of a typical capacitor start motor is about 6 times the running current. A split phase one is about 10 times. If you could replace the split phase motor with a capacitor start one, it'd help some, but not much. If you could find a PSC motor, that would help more. If you could find a 230 volt motor, that would help quite a bit.

Dimmers are notorious for being sensitive to voltage fluctuations. Ever a volt or two is noticeable. The lower the light level, the more noticeable.

If you're comfortable working in the panel, you could put the sump pump on a breaker that's on the other leg. The lights would brighten when it starts, but likely would be less noticeable.

I can't think of any soft-start or current limiting devices for single phase motors, though it might be possible to build something using a thermister. Not very reliable though.

Rob

 

[Josh61513]

The sump pump is fed from a 100A subpanel which is fed via 100' heavy (~2 ga.) aluminum wiring. From there a 20 amp breaker feeds 2 feet of 12 ga. wiring to the sump pump recepticle.

I borrowed a friends high $$ Fluke osciliscope & clamp on ammeter and found that amperage spikes to 29.6 to 54 amps for a few cycles and then drops to just under 10 amps steady state. Voltage drops about 5V during startup and then returns to just under 1V drop during steady state. The same voltage drop is measured at the sump recepticle as at the service entrance main.

I believe this suggests that the house wiring is up to snuff and that my service wiring / transformer isn't quite up to snuff (4 ga. triplex and don't-know-how-old transformer)?? But consumers energy wants $1,200 to upgrade both.

I am aware of a similar problem and solution for large DC car audio amplifiers. They simply wire in large 1/2 to 1 farad capacitors in parallel to the battery connection but very near to the amplifier. The capacitor is able to discharge energy very quickly during brief peaks. Is there any way to do something similar with AC?

Thanks agin for the feedback so far guys.

 

[Yoyizit]

The same voltage drop is measured at the sump recepticle as at the service entrance main.

I'd say a 100 mV to 10 mV voltage change at the service entrance with a 10A load is reasonable, based on an available short circuit current of 20kA to 200kA.

You could do a soft start by automatically switching in and out low value resistors in series with the pump [with home brew circuitry], or by using a very heavy PTC thermistor.