Problem with welder on generator

[kingnero]

I am doing a study of a welding machine that is to be used "in the field".
The generator is a 3 phase+N (3x400V) 12 kVA - 8.8 kW Honda Europower,
which turns out to be a bit underpowered when starting with welding - you can hear the engine stutter.

The welding machine is a Miller 350 XMT invertor, used for MIG welding at 28V - 160-200 Amps.
P=sqrt(3) x 28 x 180 = 8.7 kW, is this correct for an invertor? the efficiency not taken into consideration.

I am mostly interested in the welding part, however if the generator isn't good enough than that is also my problem. I know the basics about electricity, but I have no experience with gensets.
Could anybody help me out here?
I will try next week hooking up a large light on the gen, so that it won't need to start from stationary when welding (reducing the "lag", I hope).

Should I buy a 3-phase "energy analyzer" (or how it might be called in english), that gives direct reading of the three line voltages and currents, power factor, frequency, ... to check for the inrush and continuous current,
or will that only indicate what I think (genset is underdimensioned), or are there other tips and tricks that might be interesting for this particular case ?

 

[ccjersey]

http://www.millerwelds.com/products/multiprocess/comparisons/xmt_350mpa.html

They quote an efficiency rating there. I don't think you need the square root of 3 in your calculation.

Using a simple proportion to estimate the input power requirements to weld at 200 amps instead of 350 amps yields ~4.7kVA.

It's possible that adding a little base load to the generator will help it support the welder when striking the initial arc. It is also likely that the fuel/air mixture could be adjusted more precisely to allow it to take that surge without stuttering.

If that Honda is also an inverter generator, one that runs only as fast as required to meet the demand of the load on the inverter side, you are likely in the same situation as the usual engine driven welder with an idle down feature.

When you strike the initial welding arc, the engine must accelerate from a low idle speed to full speed while under load. The trick is to strike an arc somewhere to get the engine to rev up and then either continue into the weld or restrike the arc in the weld before the engine idled down again.

I'm not sure how fast the Honda inverter generators will idle down after a load is removed, but you may be able to "trick" it into speeding up to make starting the weld easier.

 

[wayne440]

I suggest using the money you would spend on the "energy analyzer" toward a larger generator set, or an actual engine driven welder.

 

[kingnero]

Indeed I do not need the sqrt(3), as when welding, it is simple DC power. I was mistaken with the 3ph invertor.

28V x 200 amps is still about 5.6 kW, so primary in the same ballpark, so the generator should be able to take that.

I have not been complete with my information: the welding is performed by a robot, and all the equipment has been bought my the person that worked here before me: generator, invertor and robot. There are about 50+ sets of those, so replacing this (all new) equipment is not an option.

"pre-striking" the arc cannot be done by the robot, however that is the goal I am trying to accomplish with the large light hooked up.

I will try to talk with a honda technician, for adjusting the mixture or the idling speed.

As for the "energy analyzer", that was the answer I expected...