LV-6548s switching to bypass on large loads

[Spelunker]

I'm a bit confused by the behavior of my split-phase configured pair of LV-6548s. They are a couple years old and have only be put into service a few months ago, adding solar panels and battery capacity as I go. These units are used as off-grid in my home and have A/C grid available when solar or battery are not enough. First thing I've noticed is these units seem to be heat adverse since... I experienced 90-100 F degree temperatures in my garage this summer and they appear to de-rate themselves significantly as internal temps approach 70 C (as reported by Solar Assistant).
My latest odd behavior is related to large loads coming online.
Running in SBU mode, if solar is negligible at the time and battery is nearly 100%, a 4500 watt load switched on causes the inverters to immediately switch to grid bypass even though initially there was only a load of a few hundred watts. Batteries should have no problem delivering the power (two 14.5 kwh banks, each set for 120 amp discharge).
Anyone have any ideas what might be looked at? When solar is available, it seems to have no problem powering the load, even with battery assist. Do these inverters still deliver 6500 watts AC even if powered from battery only?

Thanks for anyone's help!
-TM

 

[sunshine_eggo]

90-100°F means they're hot bits notably warmer than in controlled conditions. Victron de-rates based on temperature, which starts around 45°C. It's hard on them.

Isn't there a setting to fall back to utility based on voltage?

A big load may cause voltage sag and trigger utility. Have you confirmed all connections are fully torqued, of high quality, and sufficient gauge? Have you compared voltage measured at the inverter battery cable terminals vs. the battery terminals when this occurs?

 

[42OhmsPA]

Check setting 23 for overload bypass.

 

[Spelunker]

90-100°F means they're hot bits notably warmer than in controlled conditions. Victron de-rates based on temperature, which starts around 45°C. It's hard on them.

Isn't there a setting to fall back to utility based on voltage?

A big load may cause voltage sag and trigger utility. Have you confirmed all connections are fully torqued, of high quality, and sufficient gauge? Have you compared voltage measured at the inverter battery cable terminals vs. the battery terminals when this occurs?

Yes, now that the weather has cooled down somewhat, they seem happier. With the amount of heat these things give off, it didn't make sense having them in an air conditioned space. If I have some excess watts, maybe a small efficient wall unit next summer if needed. (seems so counter productive)
Option 12 is voltage setting for reverting to utility on low voltage. (48 V currently) I'm not seeing anything approaching that voltage on the Solar Assistant graphs, however I realize it might be a very brief event that doesn't get caught with the relatively slow sampling.

All connections have been carefully checked and quality copper has been used throughout but there still could be something hiding. Voltage sag on the battery's cabling might be something I can check. Since the large offending load is a resistive element (conventional hot water heater) I guess I didn't worry too much about massive inrush current. Another load that is behaving more adequately is my 5 ton A/C, but it doesn't draw quite that much. (it also has a soft start kit installed that seems effective)

I'll see if I can find some voltage drop on the battery cabling. I sure wish there was some additional detailed specs on these LV6548s that would outline it's behavior under certain conditions. i.e. temperature de-ratings, other conditions for it switching sources, etc. Maybe most folks use them off-grid without backup utility.

Thanks!
- TM

 

[Spelunker]

Check setting 23 for overload bypass.

Okay, I just confirmed. Overload bypass is disabled.
Thanks for the suggestion!
- TM

 

[42OhmsPA]

Okay, I just confirmed. Overload bypass is disabled.
Thanks for the suggestion!
- TM

Hmmm.
Paging @Zwy for LV6548 expertise.

 

[Zwy]

Yes, now that the weather has cooled down somewhat, they seem happier. With the amount of heat these things give off, it didn't make sense having them in an air conditioned space. If I have some excess watts, maybe a small efficient wall unit next summer if needed. (seems so counter productive)
Option 12 is voltage setting for reverting to utility on low voltage. (48 V currently) I'm not seeing anything approaching that voltage on the Solar Assistant graphs, however I realize it might be a very brief event that doesn't get caught with the relatively slow sampling.

All connections have been carefully checked and quality copper has been used throughout but there still could be something hiding. Voltage sag on the battery's cabling might be something I can check. Since the large offending load is a resistive element (conventional hot water heater) I guess I didn't worry too much about massive inrush current. Another load that is behaving more adequately is my 5 ton A/C, but it doesn't draw quite that much. (it also has a soft start kit installed that seems effective)

I'll see if I can find some voltage drop on the battery cabling. I sure wish there was some additional detailed specs on these LV6548s that would outline it's behavior under certain conditions. i.e. temperature de-ratings, other conditions for it switching sources, etc. Maybe most folks use them off-grid without backup utility.

Thanks!
- TM

Have you cleaned the filters? And did you follow the space requirements for the units?

These will derate when they get hot. You may need to provide additional air movement with a fan in warmer ambient temps. If it gets hot enough, you may have to air condition the space. If you mounted directly to a flat wall, it might help to space the units away from the wall so air can move behind the units. I have unistrut fastened to a concrete wall and the LV6548's fastened to the unistrut so there is an air space behind.

Voltage sag could be possible. Cable size should be 2/0 to each unit and 4/0 from battery to a central busbar that splits power evenly to each inverter (cables from busbar to inverters same length).

Look at settings 12, 13, 29, and 41. Each one of these can cause a switch to bypass.

 

[Quattrohead]

These units seem to be more than capable electrically but on the verge of their limits thermally. They run warm even when doing nothing and they would definitely heat your house in the winter running full bore.
My rebadged SP variants are beasts at delivering power but they have absolutely no finesse whatsoever LOL

 

[Spelunker]

Have you cleaned the filters? And did you follow the space requirements for the units?

These will derate when they get hot. You may need to provide additional air movement with a fan in warmer ambient temps. If it gets hot enough, you may have to air condition the space. If you mounted directly to a flat wall, it might help to space the units away from the wall so air can move behind the units. I have unistrut fastened to a concrete wall and the LV6548's fastened to the unistrut so there is an air space behind.

Voltage sag could be possible. Cable size should be 2/0 to each unit and 4/0 from battery to a central busbar that splits power evenly to each inverter (cables from busbar to inverters same length).

Look at settings 12, 13, 29, and 41. Each one of these can cause a switch to bypass.

Checked filters... actually pretty clean. Only the slightest hint of dust on them.
I did follow spacing requirements when mounting, however they are hung directly to the cement board behind them so there's only a slim gap behind.

Checked for voltage sag when powering the water heater. Initial load running was about 700 watts, inverter temp was 40-45C degrees indicated by Solar Assistant. Battery voltage at my main buss before the inverters was 52.4V. When I switched on the water heater voltage would drop to 51.8V and the units would immediately switch to bypass. (Solar Assistant measured 51.4 at the inverter) At the time, virtually no solar power was assisting. The water heater draws about 4500-4800 watts. When the solar input came up to a couple thousand watts and I tried again, the inverter would handle the load. I also looked with the o-scope and didn't see anything indicating a very brief surge.

So, currently I'm wondering why I can power larger loads (over 5-7k watts) when solar is providing some margin of that power. But when solar is minimal or not existent, the battery appears to not handle it even with plenty of capacity. I suppose internally these DC functions are different, one is a boost conversion for the battery while solar requires a buck conversion.

If temperature is a factor, why does a solar source behave adequately while battery does not?

Does powering large loads from only a battery source work normally for others with this unit? (when internal temps are ~40C or so)

My devices are MPP Solar units from Watts247 and I never read of anything quirky like this. (but yes the fan noise is loud and annoying sometimes)

Thanks for the help! -TM

 

[Spelunker]

These units seem to be more than capable electrically but on the verge of their limits thermally. They run warm even when doing nothing and they would definitely heat your house in the winter running full bore.
My rebadged SP variants are beasts at delivering power but they have absolutely no finesse whatsoever LOL

Please see my reply above to Zwy.
Thanks!
- TM

 

[Zwy]

These units seem to be more than capable electrically but on the verge of their limits thermally. They run warm even when doing nothing and they would definitely heat your house in the winter running full bore.
My rebadged SP variants are beasts at delivering power but they have absolutely no finesse whatsoever LOL

I wish I could get more heat from the units in winter, it isn't that warm in my basement.

 

[Zwy]

Checked filters... actually pretty clean. Only the slightest hint of dust on them.
I did follow spacing requirements when mounting, however they are hung directly to the cement board behind them so there's only a slim gap behind.

South wall? Any insulation?

Checked for voltage sag when powering the water heater. Initial load running was about 700 watts, inverter temp was 40-45C degrees indicated by Solar Assistant. Battery voltage at my main buss before the inverters was 52.4V. When I switched on the water heater voltage would drop to 51.8V and the units would immediately switch to bypass. (Solar Assistant measured 51.4 at the inverter) At the time, virtually no solar power was assisting. The water heater draws about 4500-4800 watts. When the solar input came up to a couple thousand watts and I tried again, the inverter would handle the load. I also looked with the o-scope and didn't see anything indicating a very brief surge.

What size battery cables and what is the length? Water heater should be 240V so 4800W would be 2400W per leg unless you have a 120V model.

What are you using for batteries and how large is the bank? As you have SA, are you using any functions under the Power tab?

You should just list all your inverter settings as we would need to know what charge settings are being used.

Most likely the problem is cable size or resistance in a connection. What are you using for a busbar? Some busbars are not copper but brass which leads to resistance. Even though 45C is 113F, I'm not concerned as much about inverter temp. The reason why is when the water heater was switched on it immediately went to bypass. This indicates a voltage drop somewhere.

So, currently I'm wondering why I can power larger loads (over 5-7k watts) when solar is providing some margin of that power. But when solar is minimal or not existent, the battery appears to not handle it even with plenty of capacity. I suppose internally these DC functions are different, one is a boost conversion for the battery while solar requires a buck conversion.

SCC puts power on DC bus from my understanding. Just the distance is shorter, less resistance.

If temperature is a factor, why does a solar source behave adequately while battery does not?

I would not focus on temp as much as battery voltage.

Does powering large loads from only a battery source work normally for others with this unit? (when internal temps are ~40C or so)

I don't use the MPPT's in my LV6548's as the VOC was not high enough. Basically my inverters use whatever DC bus power is available and it would be just like running on a battery.

The drop to 51.8V is a tell, use a multimeter at the inverter battery terminals as SA only updates every 10 seconds. You need a multimeter with min/max function. Again, list all of your settings as I'm very curious what you are using for charge settings plus we would like to see all the settings.

My devices are MPP Solar units from Watts247 and I never read of anything quirky like this. (but yes the fan noise is loud and annoying sometimes)

I purchased mine from Ian, the units power everything I throw at them. We have an electric dryer which is way more watts than your water heater. The Official Loads Tester like to run many things at once, I've seen it where she easily pulls 6000W per leg.

Thanks for the help! -TM

 

[Spelunker]

South wall? Any insulation?



What size battery cables and what is the length? Water heater should be 240V so 4800W would be 2400W per leg unless you have a 120V model.

What are you using for batteries and how large is the bank? As you have SA, are you using any functions under the Power tab?

You should just list all your inverter settings as we would need to know what charge settings are being used.

Most likely the problem is cable size or resistance in a connection. What are you using for a busbar? Some busbars are not copper but brass which leads to resistance. Even though 45C is 113F, I'm not concerned as much about inverter temp. The reason why is when the water heater was switched on it immediately went to bypass. This indicates a voltage drop somewhere.




SCC puts power on DC bus from my understanding. Just the distance is shorter, less resistance.



I would not focus on temp as much as battery voltage.



I don't use the MPPT's in my LV6548's as the VOC was not high enough. Basically my inverters use whatever DC bus power is available and it would be just like running on a battery.

The drop to 51.8V is a tell, use a multimeter at the inverter battery terminals as SA only updates every 10 seconds. You need a multimeter with min/max function. Again, list all of your settings as I'm very curious what you are using for charge settings plus we would like to see all the settings.



I purchased mine from Ian, the units power everything I throw at them. We have an electric dryer which is way more watts than your water heater. The Official Loads Tester like to run many things at once, I've seen it where she easily pulls 6000W per leg.

Wow... thanks for all the information!! I've done a bit more poking around and I think I've identified the issue... ME!

You were correct with focusing in on the battery/cables/etc., since the behavior appeared to be lack of current somehow triggering the switch to grid even though the battery capacity is definitely ample.
I decided again to do more load testing this morning since everything was generally cool and heat shouldn't be a problem. I took the solar panels and the utility grid offline so as to force all loads to be powered from the battery. Starting with around 500 watts I added the hot water heater, a small space heater, and some lights while I watched the power draw and voltage on the battery(s). Then wham!, the inverters went offline and the alarms were blaring. Both inverters indicated: F13 - Battery discharge over current.
I thought to myself, with the load only a little over 5000 watts (2500 each) what was going on?!

Both my batteries BMSs were set to 130 amps which was conservative for 304 ah cells. There couldn't be an issue there since having two, the inverters have 260 amps at their disposal.
That's when it hit me... wasn't there a setting in the inverters for max current draw? Sure enough... Program 41 - Maximum discharging current.
It was set on 50 amps. Exactly like I left it when I was testing things and had only a single battery hooked up. Wanted to play it safe since this was my first attempt at a DIY battery.
Doh! That was the issue! I set it to 120 on both inverters and did a successful test of over 8000 watts!

Thanks for all the help! I might be back though... On Saturday the #1 inverter blew the 63 amp fuse (again) when it was doing its switching back and forth during my testing. Now that the battery(s) are not the problem, hopefully the fuse will calm down... I've blown 3 so far... not sure why other than surge currents during switching.

- TM