My Conext SW 4024 Inverter Keeps Flashing the Lighs

[Matrix]

Ok so I plan to submit a claim to Schneider in the morning, But wanted to know before I do if this is typical of a 3400 watt split 240v Inverter.

The Schneider Conext SW 4024 Inverter should be enough to handle what i am doing.

This is a 240 split 3400 watt inverter capable of 41 amps AC peak current split over 2 legs - L1 and L2.

My 1/2 HP well pump runs on 240 split with 4 amps or less on each leg. Start up surge measures with a Clamp on Max Hold Amp meter is 14.7 amps total (about 7.3 amps per leg).

A resistive load from a 1150 watt 120v water heater element at 9.5 amps also flashes the lights.

So Why are my lights flashing (in some cases going completely off) with this inverter? These loads are well within the specs of the Inverter that should handle a 41 amp surge and 29 amps continuous over both legs. I could possibly see this with the Inductive startup load of the Well pump as my Clamp Max Hold meter may not be showing me the true surge amps. But this should not be happening with the resistive 9.5 amp water heater element as 9.5 amps is well below the 14.1 continuous amps per leg.

The system, inverter, and battery bank is no more than 4 months old ... all new.

The battery bank consists of 4 Trojan L16H AC 435ah 6v batteries in series for a 24v battery bank. The test in the video was with the batteries fully charged and tested for 100% Specific Gravity. So this is very frustrating. I have rewired everything. Tested and checked all wire connections. I have confirmed that when the inverter is in by pass mode and getting power from the gird (thus the pump and water heater element are using all the same wiring but using grid power passed thru the SW) that there is NO flashing of lights. So this is definitely only occurring on the 24v battery bank, when the SW 4024 is in Inverter mode.

Watch the Video Here

I also have created a Spread Sheet Log File done in 5 second internals from the SW during the tests that where done during the video.

Before I send this to Schneider ... Any ideas? Maybe this is just typical?

[Matrix]

Here is my XLS log file showing when the water pump and water heater element went off and on. It shows the loads at those times. Scroll down below line 33 in the spreadsheet file. This is where i started running in battery power off-grid only.

Are the DC voltage dips at the battery enough to cause a light flashing voltage dip from the inverter? Both the 1/2 HP Well pump and water heater element are connected to the Inverter as AC loads.

[Mike90250]

Lights flash ? Does that mean a slight flicker, or a near total blackout ? I get a slight flicker sometimes when the fridge starts up.

Now I want to dissect some of these numbers.

My 1/2 HP well pump runs on 240 split with 4 amps or less on each leg.
Start up surge measures with a Clamp on Max Hold Amp meter is 14.7 amps total (about 7.3 amps per leg).

I don't trust your meter. Pump Power Chart attached below 2 wire motor 34A or 3 wire = 21A
Best case, 21A @ 240V = 5040watts BEFORE PF (power factor) losses
at 24 volts DC, that's 210amps in your batteries and wires and connections. Add in inverter losses, closer to 230A
I'd be looking at all the cable crimps, battery bolts, breakers & fuses.

A resistive load from a 1150 watt 120v water heater element at 9.5 amps also flashes the lights.

9.5A x 120V = 1140watts AC or 47A @ 24VDC Run that puppy at night for 10 minutes, then use a $20 IR temperature gun to
measure each battery connection and bolt. Look for one thats a degree or 5 hotter than the rest.
What's the "leg imbalance spec" for your inverter ?

I found my bad wire, in the AC meter box, a bad 8ga connection from the electrician. I was there and heard it arcing a bit. I'd IR'd all
the battery stuff, the DC epanel, and nothing.




Well Pump Motor Specs.jpg

[Matrix]

Lights flash ? Does that mean a slight flicker, or a near total blackout ? I get a slight flicker sometimes when the fridge starts up.

Thanks Mike I will do that. If you have a minute, watch the video I linked in the initial post above. At the 34-37 second point is where the pump kicks on and the lights completely turn off and back on.

For a Temp gun, Would harbor freight be a good source for the temp gun, or go somewhere else? Might be something to that in my set up.

Also, I agree as I have seen this chart before and realize the Meter I used probably did not catch the highest spike. I cannot remember what brand it was, but it was indeed a very good one I borrowed from an electrician. But I think even older Flukes with out the "In-Rush" setting would not be fast enough to catch the spike. So I was skeptical. Pump is a 2 Wire with Ground. 240v

So yes I will check the DC connections. I would suspect a cable crimp. I am certain it is not on the AC side, I redid every single connection last month, check and double checked. And as I noted, this is only occurring when on DC not when AC is passing thru the inverter. Also happens when the fridge compressor kicks on.

What's the "leg imbalance spec" for your inverter ?

With this I have no idea. All I have is the spec sheet which I don't see where it lists this info

[Matrix]

I don't trust your meter. Pump Power Chart attached below 2 wire motor 34A or 3 wire = 21A
Best case, 21A @ 240V = 5040watts BEFORE PF (power factor) losses
at 24 volts DC, that's 210amps in your batteries and wires and connections. Add in inverter losses, closer to 230A
I'd be looking at all the cable crimps, battery bolts, breakers & fuses.

Yes that is a load. I have the 2 wire with ground. So we are at 34A x 240v = 8160
Inverter is rated for 7000 watts for 5 seconds and peak current of 41A
I am using 4/0 cable for all battery connections, but I used a hammer and anvil type crimper, not hydrolic.
The breaker for the DC is 275amps MidNite Solar.

But help me understand how this breaks down. the 34A would be the total on both legs - Not 34A on L1 and 34A on L2 - correct?
So would it be, on average, half of 34A on each leg? Say 17A on L1 and 17A on L2? So why if you then multiply 17A x 120v you only equal = 2040 per leg.

I am guessing it is squared thing in the math? But which would be the load per leg, 2040 Watts or half of 8160 Watts?

[Matrix]

Here is another interesting tid-bit. Might not mean much, but I videoed at 25 frames per second the meters while waiting for the pump to turn on ... Then went frame by frame until these analog meters hit peak at pump start ... this was the highest "bounce" in Watts these meters recorded at start up on L1 and L2 Max Watts.jpg

[Matrix]

Temp Gun Test:
Load was 53 DC amps for 15 minutes before test.
Batteries were at 1277 SG (100%) at start

BATTERY	AT BOLT	AT CRIMP
Bat 1 Main Connection +	83.9	82.4
Bat 1 -	88.7	86
Bat 2 +	86.3	87.8
Bat 2 -	86.3	88.7
Bat 3 +	84.8	83.9
Bat 3 -	89.6	91.4
Bat 4 +	89	90.5
Bat 4 Main Connection -	87.2	88.1
Neg cable in ePanel	86.9
Neg connect at shunt to inveter	86.3
Pos cable in ePanel	84.2
Pos connect to inverter	84.2
Inverter Pos busbar connect to breaker	91.4
Inverter Neg busbar connect to shunt	91.7

Colors indicate Series Cable Connections
So it looks like there might be a problem at the cable between Bat 3 - and Bat 4 + and maybe others? Or does this look pretty consistent.

Largest Temp Gap is 7.5 degrees.


Batter voltage under load at 53 amps = Total battery bank voltage was 24.9v

Bat 3 = 6.24v	Bat 4 = 6.23v
Bat 2 = 6.22v	Bat 1 = 6.23v

Voltage at ePanel Breaker / Busbar = 24.7 @ 53 amps
Voltage at inverter bolts = 24.7 @ 53 amps
Main Cable = 4/0 and 10 feet (less than 1% voltage drop between Battery bank and Inverter bolts)

[Matrix]

2nd Temp Gun Test:
Load was 47.6 DC amps for 15 minutes before test.
Tightened every Battery Crimp Connection
Batteries were at 1277 SG (100%) at start

BATTERY	AT BOLT	AT CRIMP
Bat 1 Main Connection +	83.9	82.7
Bat 1 -	86.9	86.9
Bat 2 +	85.4	86.3
Bat 2 -	85.1	86.3
Bat 3 +	84.2	84.2
Bat 3 -	87.8	89.6
Bat 4 +	87.2	88.1
Bat 4 Main Connection -	85.4	86
Neg cable in ePanel	NA
Neg connect at shunt to inveter	NA
Pos cable in ePanel	NA
Pos connect to inverter	NA
Inverter Pos busbar connect to breaker	NA
Inverter Neg busbar connect to shunt	NA

Colors indicate Series Cable Connections
After Tightening every connection at the battery ... all but one connection has improved, but more importantly, the largest gap in degrees has narrowed from 7.5 degrees to 6.9 degrees total. Will now test the lights for flashing and see what happens

[Matrix]

Well .... Still no change really. Maybe slightly better, but the lights still flash when 1.) the 240v well pump comes on and 2.) the 1150 Watt 120v resistive water heater element comes on.

Will this flashing and voltage dip hurt sensitive electronics, circuit boards, modern refrigerators and the like?

[Mike90250]

So, as you see, the hammer crimper, can barely do the job. There are still 3 really bad crimps, and only a couple good ones.

But the larger issue is the inverter cannot supply the load of starting the pump. It's just not big enough.

In a 240VAC circuit, each leg carrys the total current, L1 & L2 should be identical. A 2 wire pump will pull 34.5A through each leg, L1 supplies the amps, L2 carries it back to the source. And they change 60 times (cycles) a minute. V x A = W You have 8280watts + 10% inverter losses 9108watts plus the power factor losses which I can't calculate, but they are likely another 20%,
so I'm guessing your inverter is trying to supply 10,764 watts to start the pump.

Sooo.... I suggest you get a hydraulic crimper, or have a battery store crimp up some cables. That is safer than having a cable melt apart at the battery terminal.

And start shopping for a bigger inverter or a smaller pump.

The inverter will fail before the pump will.

With a load this large, most 24V inverters will struggle and you really should have a 48V system. Keep that in mind when you are shopping.

[Matrix]

So Mike, Would I do better to switch the pump back to 120 volts? I switched it to 240 hoping it would more equally distribute the load ... but as 2 wire this many not have done what i wanted if I read you correctly. I am still putting the same load on one leg, but at 240 it's just switching back and forth? My hope was to lower the load on each leg. I guess that would require a 3 wire setup.

As to The SW Inverter, it is rated for half its published watts on each leg (according to Schneider).
1700 continuous
2000 for thirty minutes
3500 for five seconds
on Each leg.

Also, I can easily have a battery store make up the cables, my son-in-law works there and will do it for free if I take him the parts.

Would I be better to
1. simply have them crimp my existing lugs?
Or
2. would it be better to cut them off and completely start over with new lugs.

I have enough Lugs for the job to do all 10 connections again.

[Matrix]

I conceed the well pump. And (in another post) I do have a potential local buyer for my 24v system. But should the 1150 Watt 120v resistive load of the water heater element also be dragging down an inverter rated for 1700W per leg?

[Mike90250]

Stay with 240V pump. That is a balanced load on the inverter. Converting to a 3 wire pump and external cap box would help some, but not a lot.
Well Pumps are a tough load. a 4.8Kw inverter @ 48VDC may be your only choice to cure most of the flicker.
(you are experiencing flicker, blinking/flashing is an intentional sequence, like for Missile Warning)

All the load on one leg (120V) will kill the inverter, if it even runs. You need an understanding of how North American Split Phase power works.
https://en.wikipedia.org/wiki/Split-...electric_power is a good intro, but you need electrical knowledge to comprehend it.

The water heater is a high, unbalanced load on one leg of the inverter. [ L1 and Neutral ] Obviously the inverter has trouble dealing with it as well.
as to the Fridge, my inverter (6kw) does not handle the starting glitch "well". About as good as a fridge in most houses, I do get
a bit of flicker in the LED lights, but 20 years ago, with incandescent bulbs, it may not have been noticeable.

If your cables are long enough , have new ends hydraulically crimped on, and then heat shrink seal them. With 48V, you only need half the current.

[Matrix]

Thanks ... I do understand electrical a bit, but always learning more and more, and your link helps. The system did run with the pump on 120v ... AND it did exactly as it is doing now on 240v, but on 120vac I guess that would spell a dramatically shorter life span for the inverter based on the info you have given here.

With my 4/0 cables, the good news is the main cables are actually about 2' each to long (+ and -), so it would be easy to simply ditch my series connections, and remake all 3 from the existing main cables keeping the mains the same length, but about 2' shorter each cable.

The reason my main is so long was I did not know when I made it up if I was going to go parallel or not with more batteries, and thought I might need the extra length do to connection locations. But 4 months later, i have certainly decided to stay away from parallel thanks to all the advice here on this forum.

Thanks again. I will remake all the cables and then re-test.

One other thing about cables. Some times it can be hard to cut a 4/0 and then get all the fine copper strands into the lug before crimping. Is it a bad idea to trim a small layer off all the way around the cable if it does not fit into the lug? OR does this actually cause part of the problem?

[Mike90250]

let the battery shop cut and crimp the cables.
If the cables do not fit all the strands into the lug, you have either disturbed the lay of the strands in the cable, or have the wrong lugs.