Strobe Lights, Generators, and Electrical Theory

[Flyboy]

Hello, all. I haven't posted in a long time, and thought this would be a good question to ask any of the electrical gurus on here.

I heard once a couple of years back, that bad things happen when you power too many strobe lights on a generator; that even if they are drawing current within the limits of the generator, running them simultaneously can stall the engine. Obviously, too much of anything on a generator will have detrimental results, but what is it specifically about strobe lights that one needs to be particularly mindful?

Since this question popped into my head, I researched the power specs for a Martin Atomic 3k (pretty common), and it states that the peak power draw is 33A, but the typical draw (for the MAX-15 lamp, high power mode) is only 8 amps. LD Calculator Lite figures that the Atomic 3k draws 22.2 amps across 2 legs (3000W, 208V with a .65 power factor). This led me to more questions: why are there so many different calculations for power draw? And what should be used when calculating loads? And why is the power factor so low?

I've been dealing with power distribution for a long time, but capacitance, inductance and reactance are my weak points. Any insight would be greatly appreciated. Thanks!

 

[FMEng]

Strobe lights draw current in big, short bursts to recharge the stored energy after they fire and draw smaller currents during idle between firing. The peak currents are very high compared to the time integrated average. If a bunch of strobes are fired simultaneously, the peak currents can really add up but it wouldn't if the strobes are fired one or two at a time. Short, peak currents don't heat the conductors to the extent that more constant current draw does. LD calculator is probably trying to use a number that semi-accurately reflects heating of wiring components and circuit breakers/fuses, under "typical" firing rates.

Powering from a generator is a special case because they don't handle rapid, large changes in load very well. Engines and alternators have inertia, governors have reaction time, and voltage regulators have lag time, undershoot and overshoot. Generously over estimating very bursty loads is the way to ensure the generator will handle them without issues.

 

[JD]

The reason strobes have so many different power ratings is, How do you calculate a 1/4 wave dead short? Most high power strobes work not by capacitors, but by firing the tube as a dead short right across the incoming mains! Many are "Half Wave" meaning there is a high power diode in line with the tube. This way, as the AC reverses polarity, the tube drops out of conduction and goes back to being an open circuit. Since a given model of strobe will include a diode facing the same direction, an in-sync firing of many strobes would produce a very strange load for a gene. Haven heard of stall-outs, but there may be a loss of regulation as the gene tries to cope with the sudden change.

If strobes contain a capacitor, they would be very heavy. They aren't. No Capacitor, no inductor, no reactance. Just a brief dead short.
(A handy way to tune a strobe's brightness down, as well as limit load is to put it on a long extension cable.)

 

[epimetheus]

I agree with FMEng and JD. The sudden change in load is just too quick for the governor/voltage regulator to respond. I wonder if an inverter style generator would handle this type of load more effeciently? Maybe it's electronics can respond faster? I'm also curious about the power factor quoted. What is producing such a bad power factor? A generator would definitely not be happy with an overall pf that bad.

Does anybody have any experience with running a lighting rig consisting of mostly arc-source or led fixtures and little incandescent load on a generator?

 

[JD]

I agree with FMEng and JD. The sudden change in load is just too quick for the governor/voltage regulator to respond. I wonder if an inverter style generator would handle this type of load more effeciently? Maybe it's electronics can respond faster? I'm also curious about the power factor quoted. What is producing such a bad power factor? A generator would definitely not be happy with an overall pf that bad.

Does anybody have any experience with running a lighting rig consisting of mostly arc-source or led fixtures and little incandescent load on a generator?

Arc source lamps run great on a generator, provided they are power factored correctly. LED fixtures have an inrush issue when the fixtures are first powered up as the supply caps have to charge. Once powered, they are pretty mellow. Incandescent have a cold inrush issue, not that bad as long as the gene has headroom. What sometimes can be a problem is arc-sourced mixed with incandescent. If enough conventionals are turned on at the same time, the dip may extinguish the arc lamps.

As for strobes and PF. Remember, PF is how closely the voltage waveform lines up with the current draw waveform. A strobe will be an open circuit for a lot of the waveform, but then change to a dead short later in the waveform. Not too sinusoidal