Dual Schneider Dual XW Pro 6848 power imbalance

So does anyone else have a Dual XW Pro system? If so is it normal for the slave to only produce a small fraction of the power of the master inverter?

On my system when in Inverter mode I might have something like this:

Master: 0.65 KW
Slave: 0.15 KW

Every thing works fine. I am just wonder if this out of balance is normal or if it is something that should be looked into.

Is it the same as loads increase? Low load imbalances are fairly common.

The slave on my dual XW6848+ system is usually in idle mode until the load on the master is over 3-4 kW. ( I don't watch it closely.)

Shimmy said:

Is it the same as loads increase? Low load imbalances are fairly common.

Click to expand...

Thanks for the info. I tended to think this was not critical. My house typically does not use a lot of electricity. I haver a gas furnace, stove and water heater. So other than AC in the summer when power is plentiful my typical usage is low.

I do have an electric Drier that I could choose not to run in a power outage. On rare occasions I use my big wall oven. I also have a small combination microwave / convection oven. It can use around 2000 watts.

Mjs said:

The slave on my dual XW6848+ system is usually in idle mode until the load on the master is over 3-4 kW. ( I don't watch it closely.)

Click to expand...

Thanks for the info. I have noticed that too. Usually the slave is either in idle or not putting out way less than the master.

I tended to think this was not critical. My house typically does not use a lot of electricity. I have a gas furnace, stove and water heater. So other than AC in the summer when power is plentiful my typical usage is low. I do have an electric Drier that I could choose not to run in a power outage. On rare occasions I use my big wall oven. I also have a small combination microwave / convection oven. It can use around 2000 watts.

My only BIG issue at this point is that I can't get my IQ8s to produce any power when off grid. With 15 KWH I have about one day of winter consumption and less than that if the AC is running. I'm still hopeful Enphase will fix this. Apparently IQ8s for whatever reason have a problem working at AC coupled PV. I am planning to get a small generator and possibly an auxiliary charger. My inverters are wired for a generator and I have a PV / Gen interlock contactor. So I could just use that.

I have heard that dirty power from portable generators can cause issues. When hooked to AC2 (gen input), this input essentially becomes a pass through and feeds the generator power into your home. Also, only the excess above your loads can then charge the battery. I can control the charge rate, but have to set this low enough that the generator can still supply the loads. If a big load like a water heater or AC is cycling on and off, it will limit how much you can charge and still supply that load when it comes on. With an auxiliary charger, you can set the charge rate to a certain level to match the generator and it will stay there. Also that dirty generator power never makes it into your house.

Did you (or the tech that installed) set the output voltages on the inverters? There is a specific range the output voltages must match by. That is part of the setup requirements and requires the Conext Configuration Tool. It cannot be done on the Gateway.

Mjs said:

Did you (or the tech that installed) set the output voltages on the inverters? There is a specific range the output voltages must match by. That is part of the setup requirements and requires the Conext Configuration Tool. It cannot be done on the Gateway.

Click to expand...

This is a good point. I am not sure if I individually checked the output voltage. It has been a while since I went through the setup procedure. My recollection is that I followed the procedure in the XW Pro Installation guide that had information on wiring both single and multiple units . I looked their commissioning steps and it does not call out measuring the output voltages separately. I does say to perform a system calibration if the system "Oscillates". They don't define what they mean by "Oscillate". There is an "Multi-unit" configuration guide that does mention measuring the output voltages of each inverter to make sure they are within 0.5 volts.

It looks like testing for this is as simple as just turning both output breakers off and then individually checking the input voltage. If it is out of spec, I would need to either get the tool or find someone that has it to make the adjustment.

Someone on this forum mentioned that Schneider tech support loaned them the tool to get their output voltage up to 124 to make some sensitive piece of equipment happy. I don't know that this is something they normally do or if this guy had some special arrangement.

I will check that out this weekend. I have gotten pretty good at resetting my alarm clock. Amazingly my wall ovens all seem to have enough of a backup battery that they keep time through an outage. So my twenty five year old alarm clock is about the only thing I have to reset after turning power off.

When I installed the second Xw Pro, before calibration, they would oscillate
(load would switch from one inverter to the next every few seconds)
I bought the configuration tool and did the calibration procedure listed on pages 59-63
in the XW-Pro-Multi-unit-Design-Guide-990-91373.pdf

Oscillation stopped, system now working as it should.
Occasionally the load favors one inverter more than the other for "Schneider reasons",
but normally the two inverters share the load evenly. (never exactly even, but close)

Although this thread deals with good recommendations for balancing AC output voltage on multiple XW Pro's, I have a somewhat different problem, where the DC battery voltage readings are mismatched from each other by 0.6V. One being 0.3V below, the other 0.3V above a decent digital multi-meter read.
(Edited to clarify the difference.)

This seems to create odd behavior at the end of the charger cycle in 'Absorption' / 'Absorption Exit Pending' states where DC Power on one XW Pro charger is positive (charging) while at the same time the other XW Pro charger is negative (discharging).

From the research I have done in the Modbus cross reference, many of the XW Pro internal "voltage sensors" have associated calibration scaling registers (all of these being "read only" from the Modbus access). I believe one of these registers is what is getting adjusted for the AC output voltage matching procedure discussed above.

Since I don't have access to a Conext Configuration Tool and since they are quite pricey, does anyone know if it also allows calibration of other XW-Pro voltage sensors ... particularly the charger DC Battery volts?

As expensive as the XW Pro's are, I find it very disappointing that they are not tested and adjusted to a higher precision when manufactured. :/

JustJoe said:

As expensive as the XW Pro's are, I find it very disappointing that they are not tested and adjusted to a higher precision when manufactured. :/

Click to expand...

Calibration is a bit of a black art, and it is generally part of start-up. I remember a conversation with a tech where I was complaining the readings don't match... he said "what do you want it to read?" I took offense, thinking it should read the correct value... but the reality was the values just need to work together.

There will always be some calibration difference between units. The imbalance shows up more for light loads. You will also have some calibration difference causing charging bulk, absorb voltage, and float voltage differences. If you have lead-acid battery temp compensation sensors attached there will also be calibration differences with them.

You cannot run a generator input with PV GT inverters AC coupling. You should have a breaker to PV GT inverters you can open if you want to run with a generator input.

Generator is not likely to have accurate enough frequency for PV GT inverters to lock on.

With any AC input source applied, XW inverters lock to AC input freq/phase/voltage, the inverters are slave to following AC input freq/phase/voltage.

Since XW inverters are slaved to generator input, the XW's cannot do freq shifting to regulate PV GT inverters output power.

Shimmy said:

Calibration is a bit of a black art, and it is generally part of start-up. I remember a conversation with a tech where I was complaining the readings don't match... he said "what do you want it to read?" I took offense, thinking it should read the correct value... but the reality was the values just need to work together.

Click to expand...

Thanks for the quick reply. I agree that for the AC output voltage, they just need to match.
But for the DC battery voltage, it's much more critical they be accurate. They state in
XW Pro: Li-ion Battery Solution Guide
p.3 under Supported Architectures

The following diagrams provide a typical system architecture of the Conext Gateway, XWPro, and a Li-ion battery with BMS. Currently BMS communication is supported for systems with a single XW Pro inverter and optional AC coupled PV inverters. Systems with multipleXW Pro inverters and DC coupled systems with Conext MPPT solar charge controller not currently available with BMS communication (future release). Systems with multiple XWPro inverters and/or DC coupling with Conext MPPT solar charge controller may beinstalled without BMS communications subject to the battery vendor approval of the usecase. A Conext Battery Monitor is recommended for these applications to monitor the battery SOC

Click to expand...

So we entered the static voltage values that were listed for the SimpliPhi PHI 3.8 kWh batteries for things like Bulk & Absorption. That makes decent precision rather critical to keep from shortening battery life.

JustJoe said:

But for the DC battery voltage, it's much more critical they be accurate. They state in
XW Pro: Li-ion Battery Solution Guide
p.3 under Supported Architectures

Click to expand...

Wow... isn't closed-loop communications needed for UL9540, or just future versions?

Good luck; wish I could offer more than "contact Schneider support."

Shimmy said:

Wow... isn't closed-loop communications needed for UL9540, or just future versions?

Good luck; wish I could offer more than "contact Schneider support."

Click to expand...

Well, I wasn't involved in the selection/design of this system until all of the "parts" (except of course the solar panels) were already paid for and sitting in my neighbors storage shed. SimpliPhi PHI 3.8 kWh are kinda of a strange duck. These are 48V version. They have a rather robust internal BMS to protect from catastrophic failure in each battery. But they don't talk to the outside. For extra fun, the PV inverters are AC coupled SMA SunnyBoys.
I actually helped them to get it through all the approvals and it's working for over one year now. It goes into "Power Shaving" each day at 16:00 (start of high rate) and runs their house on batteries that way until the next day at 07:00. The Charge window opens between 09:30 to 15:30. During that time, I wrote a program to run on an OpenWRT router and Modbus adjust the max charge watts to use what was left from solar after whole house consumption, selling the rest to the grid.
What I am doing now is fine tuning for longevity. I will probably need to contact Schneider, but I was trying to learn what a Conext Configuration Tool is capable of which could put me in a better situation if I have to deal with them.

JustJoe said:

I was trying to learn what a Conext Configuration Tool is capable of which could put me in a better situation if I have to deal with them.

Click to expand...

@GXMnow @400bird Any chance you have details on the configuration tool and calibration?

I don't have the configuration tool, I've only got a single XW pro.
I'm pretty sure @pvdude does, maybe he can check if DC voltage calibration is possible.

I have the configuration tool. You cannot adjust the battery voltage calibration with the configuration tool. The configuration tool is for installers to set ac voltage of multi unit systems, 3 phase, etc.

There is a calibration tool for field testing/calibration that is not for sale. You would have to find a technician with the tool to visit your site.

I'm off grid with 2 xw6848+ and 4 mppt 60/150 CC connected to jakiper batteries. I don't have any experience with selling power. My system is open loop and so I have to manually set the charge points in the CCs and XW+s. The XW+s show a dc voltage difference of about 0.4 volts. Doesn't seem to affect anything with the jakiper battery. When I had the previous AGM bank I would get a low voltage warning from the one XW+ that reads low.

The XWPros are supposed to do SOC control through the schneider battery monitor or the battery bms if closed loop. Do you have the schneider battery monitor connected? You sound like you have a good grasp of the XW modbus maps and charging parameters. (More than me.) Can you get the system voltage info from the battery monitor modbus and use that to control the charge program?

ETA, new insight firmware is supposed to have ac voltage calibration included so the configuration tool may no longer be needed for multiple unit configuration.

Wow, thanks to everyone for the very useful info! Especially @Mjs confirming that the regular Conext Configuration Tool won't adjust the DC voltage reading calibration. That's the exact info, good to have, when reaching the first line of Schneider support where they might not always "hear" the details of the problem. Could very well have saved my neighbor $600.

When I first did the commissioning, back in April of 2022, just as now, the AC1 outputs were within 0.2VAC, so didn't give the need for the tool much thought. Also seems logical newer Conext Gateway/Insight firmware could handle the AC1 measurement match calibration, because as Shimmy pointed out, accuracy to a standard is not as important.

The "engineer" from the online "solar" supplier that did the original spec document did not include a battery monitor. To be honest, after seeing that the Bulk & Absorption cycles complete every day as long as there is enough sunshine, there may well not be a need for one.

I did think about trying to use Modbus to sort of emulate a BMS. But the way I interface to it from the OpenWRT router is Modbus over (WiFi) TCP through the Conext Gateway. Even with a bit of a lag in that network path, setting the Watt power limit every few seconds does not cause a problem. Buuut I think with as tight of a closed loop feedback that the MPPT switching charge controllers need, that voltage info feedback needs to either be right on the same Xanbus/Modbus segment as with a BMS, or be internal to the charge controller.