diy solar

diy solar

XW 6848 Pro running at 70hz!

ce2078

New Member
Joined
Jul 11, 2022
Messages
4
Title says it. I just installed an XW6848 Pro in a solar / backup system. Unit was purchased new in June. When in grid support it works fine, but when it goes to battery backup (invert) the cycles will jump to 70hz. I can hear the change in a fan nearby and the inverter itself, so its not just a false reading. When the grid comes back on it sync's properly and reconnects. This is day one of this inverter being installed. Firmware was updated to latest available on 7/3/22. I have contacted support and waiting to hear back, but wondering if there is something I have wrong in the settings. I have an SCP and Conext available. I have gone through the entire advanced config on the Conext and didn't see anything I thought would affect this. This inverter is 1 of 4 we are installing at this location. This unit will be a stand alone unit (only shares the battery bank).

The battery bank is a custom modified Tesla Lithium-ion, 16 cells in series x 6 parallel, 80kw, input voltage can range from 48 to 64. One thought was the inverter might be trying to do frequency shift because the battery voltage is high? I have all the charge set points at 64V. My older XW6848+ worked fine on this pack.
 

Attachments

  • 70hz.jpg
    70hz.jpg
    134.9 KB · Views: 8
Nothing on your Hz issue.

16S for NCA chemistry is very atypical. 14S is "standard" 15S is tolerable/workable if one limits peak voltage.

Recommend you limit cell voltage range from 3.5 to 3.92V. This will get you working in the 20-80% SoC range and dramatically increase your cycle life. Tesla NCA cycle life is not particularly good when you run full range SoC.

What BMS?
 
Active Frequency Shifting for AC Coupling and Ride-Thru Frequency tolerances have been expanded with the newer UL1741SA but certainly not as high as 70Hz. Seems you have an interesting issue, let us know what the resolution is. I'm using an XW+ 6848 AC Coupled, so far so good.
If you have a hard time getting tech support from Schneider, Eric Bentsen is one the lead applications engineer, been with Schneider for quite a while. Sorry I can't seem to find his mobile number at the moment.
 
@GXMnow knows these inverters pretty well. Maybe he can help. I had a similar issue recently when my schneider 6848 was running 64hz after switching from generator to battery. In my case, just turning the inverter off then back on fixed my issue.
 
Nothing on your Hz issue.

16S for NCA chemistry is very atypical. 14S is "standard" 15S is tolerable/workable if one limits peak voltage.

Recommend you limit cell voltage range from 3.5 to 3.92V. This will get you working in the 20-80% SoC range and dramatically increase your cycle life. Tesla NCA cycle life is not particularly good when you run full range SoC.

What BMS?
I've used this pack for 3 years so far on the older xw6848+. Expanding the system with 3 more xw's. 64v is 4v/cell. Cutout is 48v, which is 3v/cell. I know it's a little low, but it hardly goes below 55v or 3.4v/cell. It's a grid tie solar setup to get me through 3 hours of on-peak time (used to be 5 hours). So I can pre-charge on cloudy days or drop to the grid if needed.

BMS is a custom PIC setup to monitor the voltages, which displays the voltages and will cut power if something gets to high/low. Not perfect I know and I've got plans to make it actively balance.
 
As for the frequency issue, tech support emailed me back and had me do another firmware redeploy followed by a factory reset of all settings. So far so good on short tests, will know more later today.

One thing I noticed is this XW Pro whines at a higher pitch than the XW6848+ it replaced, struggles more and sometimes fails to start the 1 window AC hooked to it. Could the running at 70hz for about 30 minutes have caused any damage?
 
Looks like the frequency issue is fixed. Been stable at 60hz since Tuesday.

Now I just need to find out how to get my 3 other inverters to load share equally (6848+, 6048, 4548). They are seperate from the Pro and are for the high power 220v loads. It seems their voltages are out of calibration with respect to each other and one draws power while another is putting power back into the battery. An inefficient loop. Looks like I might need to find someone in the Phoenix area with Config AI?
 
"load share equally (6848+, 6048, 4548)"

I've always been told different models won't and can't be tied together on their output lines.

Now if they have separate loads, it should work.

I have two inverters (6848 and 4548 on the same gateway) only because the gateway model has two distinct XanBus connections.
Althou the gateway does have issues with data on the local webpage on the 4548. Not seeing voltages that are there on screen.
 
Only time I've seen anything other than 60 Hz off-grid is when the 6848 is trying to 'shed' an AC coupled inverter shoving more energy back into the 6848 than it can handle with battery recharging. Never more than 65 Hz thou as that dumps the offending grid-tie units.
 
Was there an event that triggered this or did it seem to happen at random?
It was just doing it. Appareently, it was a known issue from Scheider when you paralleled two 6848's. That didn't stop the company I hired to design my build from building a "wired system" for me to DIY. (Solar Unbound). There was a new firmware that fixed it.
 
"load share equally (6848+, 6048, 4548)"

I've always been told different models won't and can't be tied together on their output lines.

Now if they have separate loads, it should work.

I have two inverters (6848 and 4548 on the same gateway) only because the gateway model has two distinct XanBus connections.
Althou the gateway does have issues with data on the local webpage on the 4548. Not seeing voltages that are there on screen.
Could you give a little more detail on running the Gateway with two xanbus connections. I've been considering getting a Gateway - and running another SW4024 alongside my original unit (not sync-tied, but used for discrete loads and as a backup "generator" in pass/thru mode). Already maxed out my original SW4024 (6kw DC coupled and 3kw AC coupled.) It's been working beautifully, but a split system could give me more flexibility - and dual monitoring on the same interface would be handy. I am completely off-grid - for 30 plus years. Thanks.
 
From my experience ... the gateway (out of production) has 2 XanBus chains. The Insight Home dose not. That's all I know.
I have 6848 and mppt on one and a 4548 on the other bus. Modbus works to all devices. I don't like the gen assist on the 6848. So my 4548 accepts genset input and will charge the common battery bank while 6848 always powers loads and manages ac coupled inverters. I do have the grid here but two hurricanes fall of 2022 had me out for three 24 hour periods.
 
From my experience ... the gateway (out of production) has 2 XanBus chains. The Insight Home dose not. That's all I know.
I have 6848 and mppt on one and a 4548 on the other bus. Modbus works to all devices. I don't like the gen assist on the 6848. So my 4548 accepts genset input and will charge the common battery bank while 6848 always powers loads and manages ac coupled inverters. I do have the grid here but two hurricanes fall of 2022 had me out for three 24 hour periods.
I can still get one on eBay. How do the reporting/control functions work across the two networks? Problems? Screenshots? Thx again.
 
Title says it. I just installed an XW6848 Pro in a solar / backup system. Unit was purchased new in June. When in grid support it works fine, but when it goes to battery backup (invert) the cycles will jump to 70hz. I can hear the change in a fan nearby and the inverter itself, so its not just a false reading. When the grid comes back on it sync's properly and reconnects. This is day one of this inverter being installed. Firmware was updated to latest available on 7/3/22. I have contacted support and waiting to hear back, but wondering if there is something I have wrong in the settings. I have an SCP and Conext available. I have gone through the entire advanced config on the Conext and didn't see anything I thought would affect this. This inverter is 1 of 4 we are installing at this location. This unit will be a stand alone unit (only shares the battery bank).

The battery bank is a custom modified Tesla Lithium-ion, 16 cells in series x 6 parallel, 80kw, input voltage can range from 48 to 64. One thought was the inverter might be trying to do frequency shift because the battery voltage is high? I have all the charge set points at 64V. My older XW6848+ worked fine on this pack.
Update a firmware to 1.11 bn49 is all
 
I am at bn49 on the 6848.
I can still get one on eBay. How do the reporting/control functions work across the two networks? Problems? Screenshots? Thx again.
I find as time goes on and settings get changed, these inverters internals get out of sync with what the local webpage say the settings are.
I lost webpage readings on batt voltage and load voltage, nothing i did would bring those readings back, thou the functionality was still 240v coming out. In the end a 'factory reset of the user settings' made it whole again. Go figure.
 
I still wish S/E would allow me to calibrate that battery voltage errors. The mppt60 batt says 0.6 volts different that the 6848. Under no currents, so that can not be the excuse. Hard to correlate thresholds when the two S/E devices are that far off.
 
Back
Top