Gen charging. Loosing 2kw.

porch · December 2019

I seem to be loosing 2kw of power charging my batteries and I can't figure out where.

Inverter system is a dual Outback FX3648 with x240. It's the off the shelf assembled unit.

Generator is a 5,500w 240v unit from the 1950s. I am unsure of brand, but it's a big air cooled cast iron 1,800 RPM unit with no electronics next to "condensers". It runs great and has no issues even overloaded.

When I am charging my batteries using the generator running to the AC IN on the inverter, I seem to be loosing about 2kw.

Exact voltages/amps very, but this is from what I was measuring as of last night using my clamp amp probe.

ACIN LEG 1 = 18amps @ 110 v

ACIN LEG 2 = 20amps @ 110 v

Total watts = 4180

DC out to battery = 33 amps at 55 volts

Total watts = 2235

Total house loads are just under 200 watts.

I know the charge process is not 100% efficient, but it seems I am missing about 2,000 watts some place. What should check?

Estragon · December 2019

Do you have a "mate" display for the OB inverters? If so, it should give in/output numbers that might confirm/refute the clamp meter #s.

Is the clamp meter definitely a DC version? Many are AC only.

porch · December 2019

The Mate confirms the clamp meters readings. The Clamp meter us the UNI-T one that does DC amps.

Estragon · December 2019

I dunno. 33adc + a few for loads would come close to the 38aac input, but for the 2x voltage. Maybe just coincidental.

My genny is wired for 120v into a pair of parallel stacked (master/slave) FXs. An x240 provides 240v for a water pump.

I assume in your setup, the inverters are series stacked, and the x240 steps down to 120v to balance the leg loads on the FX outputs?

porch · December 2019

At the moment, I have the max input amps set to 20 for each FX. The inverters can draw more and the generator outputs more, but 20 amps per leg puts the total wattage at 4,800 watts from the generator giving it a bit easer life.

The FX are series stacked, or whatever Outback calls their custom configuration. Each inverter provides 120v, but the x240 provides the 240 volts with only one inverter on. So the master inverter is always on, and the slave (OB slave) kicks in when the loads go up.

I thought the x240 was causing my issue, but I disconnected it and reset the inverter settings, and no difference.

MarkP · December 2019

The AC power in is Volts x Amps x Power Factor. It is not unusual for battery chargers to operate with a low PF of .5 or .6 although I would expect an on board charger in an Outback to perform better.

porch · December 2019

That's what I was thinking. I am expecting better with the on board charger. The power loss is not nearly as bad when doing normal inverting.

Estragon · December 2019

I was also thinking PF, but I thought the FX charger was power factor corrected. I don't know if the stacking could affect it. Seems to me the in-out numbers are much closer with my parallel stack.

It doesn't really matter in terms of this discussion, but measuring current on L1/L2 on 120/240 split phase, I don't think you can add the phase currents. It would be 240v though, so VxA comes out about the same.

Theoretically, it seems to me the readings on L1/L2 should be identical. In practise, it's hard to read both accurately and nearly simultaneously. Still, with the 120v loads balanced by the x240 downsteam from the inverters the difference (~2a@240v) makes me wonder.

I have a max input set for my FXs as well (>30yr old 4kw diesel). There's also a max charger output setting for each port/FX IIRC. If there's more AC loads, the FX cuts back charging. Is it possible the timing of readings and apparent discrepency were affected by this?

wellbuilt · December 2019

I have a fxr3648 and I run at 2100watts in and I see 2000watts going to charging , I have a inverter generator and the power walks down to 600 watts in 500 wats out for a full charge.

Once I hit 93% I have less power going to the battery’s as they get full. I’m not sure how a regular genarator reads on the mate .

porch · December 2019

Reading the L1/L2 is the thing that is tripping me up too. Technically, the FX pair are two 120v loads. I would think the x240 would balance the charging load, but it does not seem to as much. There can be a 5 amp difference between L1 and L2.

I am sort of wondering if something is messed up with the x240. But if it was turning 2kw into heat, it would be very hot, and it's not.

The only thing that cycles in the house is the refer, and it's off most of the time. House loads are spread across any old breaker, so some could be more on one leg more then the other. It would be less the 200 watts.

I am questioning how I should even be measuring this with the FX pair and the x240. The generator has lost all labelling from sitting out in the rain most of it's life, so I am not even sure of it's rating. But it has a 5500 stamped in it's cover, so I am assuming 5500 watts and based on the cable size, I believe it's at least that. But I can't discount that something is screwed up in the wiring inside it, but then again, not a lot to go wrong with it.

Next to clearing all settings, disconnecting one inverter and x240 then testing one inverter at a time, any troubleshooting suggestions?

Thanks WellBuilt for the confirmation that I am turning propane into heat in unneeded ways. I would be thrilled to see those numbers.

mike95490 · December 2019

I suspect it's something to do with power factor losses.

I doubt there is anything wrong with the genset, but if it's really old, maybe the output correction caps are old and getting dried out. If you can get to them and read the labels, maybe a re-cap is needed .

wellbuilt · December 2019

Maybe you could try a different generator to rule out that as the problem?

jonr · December 2019

Might be a good excuse to buy an energy meter that can actually measure AC watts. RMS volts x RMS amps often results in non-sense, not watts - with AC or DC.

Estragon · December 2019

If things are wired the way I'm thinking, the pair of FX inputs are wired to the 240v legs of the genny, with the x240 on the FX outputs stepping down and balancing 120v loads. There should be no difference in leg currents from there.

If the Fx chargers are both set with the same limits, I think they should be a balanced 240v load to the genny, with zero current on the neutral. Maybe try measuring current on the neutral to see if it really is unballanced, or it's just a measurement or timing thing?

PF sometimes confuses me. It helps me to remember that although generators and inverters are sold by watt ratings, that's a bit deceptive. They're really V(olt)A(amp) rated, which is only the same as watts for a purely resistive load. The genny could put out 4800va (240v x 20a), but if the voltage and current are out of phase, wattage is less. How much less depends on the PF of the loads.

You may not be burning (much) more propane. My understanding is (eg) a load of 2400w and PF of 0.5 would be 4800va. At 240v, that would be 20a. With a 2400w resistive load and PF of 1, it would be 10a. The difference would be a bit higher wire loss at 20a vs 10a, but not a difference of 2400w. The genny puts out ~2400w in either case.

Put another way, in order to make 1 watt, you need 1v and 1a at the same instant in time. If the volts peak earlier and the amps peak later, you get less than 1w (both produced and consumed), but still measure 1volt x 1amp.

porch · December 2019

Thanks for all the great suggestions.

I ordered 2 power meters from Amazon and wired them in. This is what it looks like under full load.

The power factor is really bad on one leg. I swapped the legs around on the generator side and the problem followed the leg. I swapped the legs around on the inverters, and the problem didn't follow. So the problem is that one leg. And it gets worse then 0.5. As the inverter is ramping up, I see it hit 0.3 for a few seconds.

The generator has two 28MFD condensers (one for each leg) so I bet one is dead. They look just like distributor condensers and likely are. Parts on order.

Oh, and the load on the neutral wire is 25 amps with the below load. I think this messed up power factor is screwing up my readings.

jonr · December 2019

While you are fixing it, I'd also adjust your rpm to get the right Hz.

porch · December 2019

I did adjust it to see if it affected the power factor and it didn't.

The generator has no active electronics and just a manual RPM governor, so the RPM drops a bit under full load. No load it's at 62Hz.

porch · January 2020

I found the issue. It was not the condensers. I bought a Ampprobe clamp on meter that does power factor and traced the issue all over the place.

Turns out the problem went away when I disconnected the X240 transformer. Under full load, I now see a power factor of .92 on both legs of the generator.

Right now I am drawing 4500 watts from the generator.

The batteries are charging at 3700 watts.

So I my losses went from 2,000 watts with the X240 to 800 watts without the X240.

Or in percent, the efficiency went from 52% to 82%.

Estragon · January 2020

Interesting. In post #5 you disconnected the x240 but it didn't appear to help. Anyway, good that it's better now!

porch · January 2020

You are right, I did disconnect the X240. At least I thought I did. It was a few years ago when I last got into the weeds to figure this out. I am sure I disconnected it then, but maybe not.

I also think the x240 gives an high PF reading if the generator legs are unbalanced? I did swap the legs on the generator, and the issue stayed with one generator leg, but maybe that is because that leg is slightly weaker, causing the x240 to work harder and cause PF problems on that one leg.

I did scope out the waveform from both legs of the generator and didn't see a big difference between the two. Or really any difference. I was expecting a half shorted rotor or something, but no.

I can't explain it. But I am seeing a large increase in charging amps going to the batteries now.

Gen charging. Loosing 2kw.

Comments

Categories