crest factor problem ?

[Robert Morein]

I have a Yamaha EF3000iSE AC generator, rated at 2800 watts continuous.
This unit uses "inverter technology", which means that it uses a DC
alternator to power an AC inverter.

A Xantrex SW4024 charger-inverter is a computer controlled combination of a
charger and inverter, for AC power backup.
The intent is to use the Yamaha on the generator input of the Xantrex to
charge a battery bank.
The Yamaha circuit protection trips after a few minutes of operation.
No significant current is being drawn. The batteries are already on float.
The Xantrex has programmable current limiting for AC sources. Reducing the
allowable current draw from the generator to 10 amps has no effect on the
behavior.

I have tested the Yamaha with two 1500 watt "cube" heaters. The Yamaha
supplies 3KW into these resistive loads without problem.
I conclude that large current pulses, typical of switching power supplies,
are glitching the Yamaha protection circuitry.

This is a serious problem. I'm wondering if I could salvage the Yamaha for
this use by placing a ferroresonant device between the Yamaha and the
Xantrex.
One possible device would be the Sola sine-wave neutralized units. However,
they are expensive.
I tossed one years ago (sigh!)

Suggestions?

 

[Winfield Hill]

Robert Morein wrote...

I have a Yamaha EF3000iSE AC generator, rated at 2800 watts continuous.
This unit uses "inverter technology", which means that it uses a DC
alternator to power an AC inverter.

A Xantrex SW4024 charger-inverter is a computer controlled combination
of a charger and inverter, for AC power backup. The intent is to use the
Yamaha on the generator input of the Xantrex to charge a battery bank.
The Yamaha circuit protection trips after a few minutes of operation.

Xantrex is famous for their high-performance power-factor-corrected
(PFC) supplies, which appear to be resistive loads to the ac line. But
the SW4024 data sheet doesn't mention PFC or power-factor correction,
so perhaps it's not PFC, but instead draws huge pulses of current at
the peak of each ac cycle, charging its storage / filter capacitors.

 

[The Phantom]

Robert Morein wrote...

Xantrex is famous for their high-performance power-factor-corrected
(PFC) supplies, which appear to be resistive loads to the ac line. But
the SW4024 data sheet doesn't mention PFC or power-factor correction,
so perhaps it's not PFC, but instead draws huge pulses of current at
the peak of each ac cycle, charging its storage / filter capacitors.

I think you've misunderstood what he's talking about, Win. There
was a company named Trace Engineering that made the original SW4024
inverter. A Canadian company named Xantrex, that made line operated
power supplies, bought Trace Engineering and Heart Interface and
Statpower in 2000. So the SW4024 inverter (and charger) is now sold
under the Xantrex name. It is that device that the OP is referring
to, not the power supplies originally made by the Xantrex of pre 2000.
The SW4024 produces a stepped sine approximation to a pure sine wave
and the high dv/dt at each of the steps can indeed draw fairly large
pulses of current.

 

[The Phantom]

I have a Yamaha EF3000iSE AC generator, rated at 2800 watts continuous.
This unit uses "inverter technology", which means that it uses a DC
alternator to power an AC inverter.

A Xantrex SW4024 charger-inverter is a computer controlled combination of a
charger and inverter, for AC power backup.
The intent is to use the Yamaha on the generator input of the Xantrex to
charge a battery bank.
The Yamaha circuit protection trips after a few minutes of operation.
No significant current is being drawn.

How did you determine this? You should be sure to measure this
current with a true RMS responding meter, preferably one with a high
crest factor rating. The fact that you used the phrase "crest factor"
in the subject line leads me to believe that you know what crest
factor is. A possible problem is that most DVM's have a crest factor
limitation of 3 or 4, and you may be seeing significantly higher crest
factors than that, which would give a false reading on a DVM. One way
to get an accurate reading would be to use a thermocouple type ammeter
which can give accurate readings with very high crest factors.

The batteries are already on float.

The Xantrex has programmable current limiting for AC sources. Reducing the
allowable current draw from the generator to 10 amps has no effect on the
behavior.

I have tested the Yamaha with two 1500 watt "cube" heaters. The Yamaha
supplies 3KW into these resistive loads without problem.
I conclude that large current pulses, typical of switching power supplies,
are glitching the Yamaha protection circuitry.

Slip a current transformer around one of the conductors between the
Yamaha and the SW4024 and have a look with a scope.

This is a serious problem. I'm wondering if I could salvage the Yamaha for
this use by placing a ferroresonant device between the Yamaha and the
Xantrex.

Try an inductor of about 1 millihenry and suitable current rating.