Fridge won't start at night

[hfraser]

I have an off-grid cabin. Last summer I installed 4x250watt panels, a 40amp mppt charge controller, 4x100 amp-hr batteries, and a 1500watt/3000 watt surge inverter. All this to run an 18cu ft 120V ac Frigidaire refrigerator. And it works. When the sun shines and we're generating power. As soon as it gets dark, it trips the inverter when the compressor starts. Yesterday I charged all 4 batteries from a generator to full, and at 8:00 pm when they were done, I plugged in the fridge. Inverter tripped. Yet it ran from 10:00 am until 3:00 pm no problem. I have tried a 900 watt toaster.. no problem. We have a 60 watt cooler running 24x7. What is the fridge looking for that needs the extra juice from the panels?

I also have a shunt dc meter on the batteries and a clamp ac meter on the output from the inverter. The fridge draws 106 watts and 400 watts when the defrost heater kicks in. I know the inverter's capable of the compressor draw because it run fine all day. BTW, the fridge is brand new.

Help? I am lost.

 

[Mattb4]

You would think that if it runs during the day on the inverter is should also at night. However it might just be slightly more than the batteries can supply without the solar kick. Things you might want to check are all connections, battery wire size and if you can get a hold of a inrush amp meter to measure actual starting current for the refrigerator. Understand that because it has a motor that starting current can be as high as 5-8 times running current. That is what trips the inverter on over load.

Devices that are resistant mostly such as heaters (toasters and such) do not have much or any difference in starting and running current.

You might need a bigger inverter.

 

[sunshine_eggo]

I have an off-grid cabin. Last summer I installed 4x250watt panels, a 40amp mppt charge controller, 4x100 amp-hr batteries, and a 1500watt/3000 watt surge inverter. All this to run an 18cu ft 120V ac Frigidaire refrigerator. And it works. When the sun shines and we're generating power. As soon as it gets dark, it trips the inverter when the compressor starts. Yesterday I charged all 4 batteries from a generator to full, and at 8:00 pm when they were done, I plugged in the fridge. Inverter tripped. Yet it ran from 10:00 am until 3:00 pm no problem. I have tried a 900 watt toaster.. no problem. We have a 60 watt cooler running 24x7. What is the fridge looking for that needs the extra juice from the panels?

I also have a shunt dc meter on the batteries and a clamp ac meter on the output from the inverter. The fridge draws 106 watts and 400 watts when the defrost heater kicks in. I know the inverter's capable of the compressor draw because it run fine all day. BTW, the fridge is brand new.

And a ~600W surge when the compressor starts.

IMHO, voltage sag at the battery due to load. Bad connections and/or low charge.

When the sun shines, and you're generating power, the incoming power raises the battery voltage to counter the sag from the compressor surge.

#1 troubleshooting step: connections.

1. Have you built the battery bank engaging in best practices when paralleling them? (ref victron wiring unlimited in my sig)
2. Confirm that all your battery to inverter cables are properly placed, of sufficient gauge, crimps are good and all threaded connections are properly torqued.
3. #2 includes all the interconnects.
4. Confirm that any fuse or breaker in that circuit is also properly torqued, is not some cheap amazon crap and is not generating heat.

Are you 100% certain that you fully charged the battery? You held them at around 14.4V measured at the battery terminals for 2+ hours?

 

[hfraser]

You would think that if it runs during the day on the inverter is should also at night. However it might just be slightly more than the batteries can supply without the solar kick. Things you might want to check are all connections, battery wire size and if you can get a hold of a inrush amp meter to measure actual starting current for the refrigerator. Understand that because it has a motor that starting current can be as high as 5-8 times running current. That is what trips the inverter on over load.

Devices that are resistant mostly such as heaters (toasters and such) do not have much or any difference in starting and running current.

You might need a bigger inverter.

I would have thought 4 fully charged 100 amp-hr batteries (actually 200 amp-hr at 24V... I have them wired serial-parallel) would provide way more power than the fridge would demand, even assuming it's 10 times the running load. And I'm ok buying a bigger inverter, but it if it works ok with batteries plus solar, it's handling more power in the daytime no problem. My meters don't update fast enough to see the startup load, but the shunt meter did show 800 watts once. Still well below the inverters limit.

I'm new to all this but thought I'd done my homework. Thanks for the feedback.

 

[sunshine_eggo]

I would have thought 4 fully charged 100 amp-hr batteries (actually 200 amp-hr at 24V... I have them wired serial-parallel) would provide way more power than the fridge would demand, even assuming it's 10 times the running load. And I'm ok buying a bigger inverter, but it if it works ok with batteries plus solar, it's handling more power in the daytime no problem. My meters don't update fast enough to see the startup load, but the shunt meter did show 800 watts once. Still well below the inverters limit.

I'm new to all this but thought I'd done my homework. Thanks for the feedback.

If you did everything right, yeah, it should be enough, but the fact that you don't have problems when you have incoming power suggests you did not engage in best practices.

Did you ACTUALLY charge the battery to full with the criteria I provided?

Can you provide pictures of the battery installation?

 

[hfraser]

And a ~600W surge when the compressor starts.

IMHO, voltage sag at the battery due to load. Bad connections and/or low charge.

When the sun shines, and you're generating power, the incoming power raises the battery voltage to counter the sag from the compressor surge.

#1 troubleshooting step: connections.

1. Have you built the battery bank engaging in best practices when paralleling them? (ref victron wiring unlimited in my sig)
2. Confirm that all your battery to inverter cables are properly placed, of sufficient gauge, crimps are good and all threaded connections are properly torqued.
3. #2 includes all the interconnects.
4. Confirm that any fuse or breaker in that circuit is also properly torqued, is not some cheap amazon crap and is not generating heat.

Are you 100% certain that you fully charged the battery? You held them at around 14.4V measured at the battery terminals for 2+ hours?

And a ~600W surge when the compressor starts.

IMHO, voltage sag at the battery due to load. Bad connections and/or low charge.

When the sun shines, and you're generating power, the incoming power raises the battery voltage to counter the sag from the compressor surge.

#1 troubleshooting step: connections.

1. Have you built the battery bank engaging in best practices when paralleling them? (ref victron wiring unlimited in my sig)
2. Confirm that all your battery to inverter cables are properly placed, of sufficient gauge, crimps are good and all threaded connections are properly torqued.
3. #2 includes all the interconnects.
4. Confirm that any fuse or breaker in that circuit is also properly torqued, is not some cheap amazon crap and is not generating heat.

Are you 100% certain that you fully charged the battery? You held them at around 14.4V measured at the battery terminals for 2+ hours?

Two pairs of series connected batteries the connected in parallel. All the wires were the reommended gauge. The fuse is rated for 175 amps at 32V.


I used a 3500 watt generator and my "smart" car charger to charge the batteries. Each was disconnected and charged until the charger said they were full. All together it took about 7 hrs.

I think you have the answer with the connections. I only made 2 cables myself, and purchased a special crimping tool made for car battery cables, so I believe the cables are ok. But the entire battery back is under a metal cabinet (12" above the batteries), and my wife said it felt warm when the fridge is running. I have felt everything checking for heat while I have the pack rolled out, but never while the fridge was running. Your suggestion explains evetything I am seeing. I will check it out tomorrow morning.

 

[hfraser]

I thought I followed best practices. But agreed, it should work but it's not, so I've screwed up.somewhere.

 

[hfraser]

You would think that if it runs during the day on the inverter is should also at night. However it might just be slightly more than the batteries can supply without the solar kick. Things you might want to check are all connections, battery wire size and if you can get a hold of a inrush amp meter to measure actual starting current for the refrigerator. Understand that because it has a motor that starting current can be as high as 5-8 times running current. That is what trips the inverter on over load.

Devices that are resistant mostly such as heaters (toasters and such) do not have much or any difference in starting and running current.

You might need

And a ~600W surge when the compressor starts.

IMHO, voltage sag at the battery due to load. Bad connections and/or low charge.

When the sun shines, and you're generating power, the incoming power raises the battery voltage to counter the sag from the compressor surge.

#1 troubleshooting step: connections.

1. Have you built the battery bank engaging in best practices when paralleling them? (ref victron wiring unlimited in my sig)
2. Confirm that all your battery to inverter cables are properly placed, of sufficient gauge, crimps are good and all threaded connections are properly torqued.
3. #2 includes all the interconnects.
4. Confirm that any fuse or breaker in that circuit is also properly torqued, is not some cheap amazon crap and is not generating heat.

Are you 100% certain that you fully charged the battery? You held them at around 14.4V measured at the battery terminals for 2+ hours?

I have downloaded the "Victron Wiring Unlimited" and will go through it tonight instead of sleeping for fear I'm going to burn down my in-law's log cabin.

Seriously, thanks for the help.

 

[sunshine_eggo]

I have downloaded the "Victron Wiring Unlimited" and will go through it tonight instead of sleeping for fear I'm going to burn down my in-law's log cabin.

Seriously, thanks for the help.

I doubt it's that bad at all!

It's hard to follow, but it looks like you cross-connected the main leads so that you're pulling across the entire battery. This is good.

I'd double check the 11 nuts I can see in the picture to make sure they're good and snug (preferably torqued to spec), there is no corrosion on any terminal faces, there is no insulation/shrink trapped under/between terminals.

Additionally, I'd check the same at the other end of the two main cables and where they attach to the inverter.

Is your inverter cut-off voltage programmable? If so, did you change it from the default? Do you know what the cut-off value is for this inverter?

 

[hfraser]

I know it's tough to follow but I was careful to pull across the entire set. I used short cables to minimize losses and ended up with this. I picked up a few longer cables today to allow me to rearrange the batteries to look simpler. I do not have a torque wrench here and home is 11 hours away so I will redo all the connections and endure they are snug until next trip.

My inverter does not have a programmable cutoff. I will check to see what it is set at.

Again, thanks. I have a solid lead to follow up on.

 

[sunshine_eggo]

Just because I found it hard to follow, doesn't mean I would increase lead length! It's fine as is. Just check all the connections.

Does the inverter have a voltage readout?

If so, does it measure the same as a separate voltmeter?

Have you ever put a voltmeter on the inverter terminals and turned on the fridge when you knew it would crap out on you?

 

[Clifton]

I didn't read all the suggestions so sorry if this is a repeat. I would check the voltage at the fridge plug on startup. I would check peak amp draw on start up, if your clamp meter has a hold setting. If not you won't catch it just watch amp draw. I have a 6 foot extension cord with n and hot separated that makes testing easy and poke a pointy probe into without harming the good cord.

 

[hfraser]

Yes it does and I have not compared the two. I have a voltmeter with me and can certainly test the scenario by turning off the solar panels (I did install a cutoff switch) and compare the voltages. What am I looking for?

 

[hfraser]

I have a permanently installed clamp meter but it does not have a highwatet hold feature. I did see about 800 watt draw on the shunt meter connected to the batteries.

 

[hfraser]

I wil

I didn't read all the suggestions so sorry if this is a repeat. I would check the voltage at the fridge plug on startup. I would check peak amp draw on start up, if your clamp meter has a hold setting. If not you won't catch it just watch amp draw. I have a 6 foot extension cord with n and hot separated that makes testing easy and poke a pointy probe into without harming the good cord.

I didn't read all the suggestions so sorry if this is a repeat. I would check the voltage at the fridge plug on startup. I would check peak amp draw on start up, if your clamp meter has a hold setting. If not you won't catch it just watch amp draw. I have a 6 foot extension cord with n and hot separated that makes testing easy and poke a pointy probe into without harming the good cord.

I will check these in the morning. My meter shows amps and volts.

 

[Supervstech]

View attachment 108720
Two pairs of series connected batteries the connected in parallel. All the wires were the reommended gauge. The fuse is rated for 175 amps at 32V.


I used a 3500 watt generator and my "smart" car charger to charge the batteries. Each was disconnected and charged until the charger said they were full. All together it took about 7 hrs.

I think you have the answer with the connections. I only made 2 cables myself, and purchased a special crimping tool made for car battery cables, so I believe the cables are ok. But the entire battery back is under a metal cabinet (12" above the batteries), and my wife said it felt warm when the fridge is running. I have felt everything checking for heat while I have the pack rolled out, but never while the fridge was running. Your suggestion explains evetything I am seeing. I will check it out tomorrow morning.

You might want to try it with the scarred battery set out of the loop.
Looks like you reverse connected a battery for a bit… scary wasn’t it!

Some batteries won’t be harmed in that situation, but some will… best to fully test it.

 

[Clifton]

If you have a current feed problem. You could have 120v with a light load, increase the load and that voltage drops. If your voltage drops too much on start up, it may shut off to protect the compressor. If you measure voltage at a main panel it will be 119 to 122. Check at a running pool pump and it will not be. If an amp meter does not have a hold feature, I promise it won't catch the high number it happens in miliseconds. Before I put a slow start on our 3.5 ton a/c it looked like it drew 40 amps. On the hold setting, it was over 70 amps. I hate saying buy a new tool but I picked up a cheap

 

[hfraser]

You might want to try it with the scarred battery set out of the loop.
Looks like you reverse connected a battery for a bit… scary wasn’t it!

Some batteries won’t be harmed in that situation, but some will… best to fully test it.

Darn... you saw that. Too many hours/days trying to get this working. Yes. It was scary. I pulled the battery out and hooked it up to the charger on the generator, and it seemed ok. Hopefully I got away with a warning.

Our cabin is in northern Quebec, 25 miles into the mountains on really bad logging roads, and 11 hrs from home. This problem has been going on since last fall when I first installed this with 2x250 watt panels, at which time we then had 14 days of rain before closing up the cabin for the winter. This year I brought up the extra panels, but the problem with the fridge has been happening all summer. I scarred the battery the day before yesterday.

 

[hfraser]

If you have a current feed problem. You could have 120v with a light load, increase the load and that voltage drops. If your voltage drops too much on start up, it may shut off to protect the compressor. If you measure voltage at a main panel it will be 119 to 122. Check at a running pool pump and it will not be. If an amp meter does not have a hold feature, I promise it won't catch the high number it happens in miliseconds. Before I put a slow start on our 3.5 ton a/c it looked like it drew 40 amps. On the hold setting, it was over 70 amps. I hate saying buy a new tool but I picked up a cheap

I have another clamp meter with the hold function I will bring, but of course it's at home, 11 hours away. I will add it to my list of stuff to bring next time.

 

[RCinFLA]

Sounds like during the day the extra juice from PV is just enough to supplement battery slump during startup surge current to prevent inverter from shutting down due to too low DC input voltage.

It could be batteries' condition, but first thing is making sure you have sufficient battery cable size and good connections to minimize voltage drop from batteries to inverter during the high startup surge current. Also make sure you give refrig enough off time to equalize its refrigerant gas and cool down any startup PTC thermistor or its startup surge current from compressor will be greater. Other possibility is defrost cycling current but that is usually not synchronous with time of day, occurring every 16 to 25 hours.

Also be aware of inverter idle current. It could be running battery down at night to point where the battery voltage slump during high startup surge current trips off inverter. Cooler night temps also reduces battery peak output current capability.