Solar panel controller stopped working

[Accedda]

On my vehicle I have 3 x 100w panels wired in series i.e. 300w in total. The panels are connected to an MPPT controller, which is in turn connected to a 105ah 12v deep cycle battery. The panels supply 5,5amps of current. Using a multimeter I've recording voltage coming into the controller for the panels ranging from 40,6v in overcast conditions to 46,4v in partly cloudy conditions. The battery powers two 12v compressor fridges that draw 2,5v each.

The battery can also be charged from a mains supply via a cTek charger. The only battery charge indicator I have is a small robot style cTek indicator (red, amber, green). The battery is also charged from the vehicle's alternator.

The system was installed by a fitment centre and my knowledge of it is very limited.

During a camping trip in extremely hot conditions (40*C) that made the fridges work overtime the battery drained to 7,3v overnight - measured with a multimeter a the MPPT outlet terminals. By mid afternoon the voltage resorted to 13,5v. However, the following morning the MPPT controller had stopped working overnight. It did not work for the following four weeks. The battery's own charge indicator was red during this time despite doing some mileage with the vehicle. The cTek indicator would flash green when the vehicle was in motion, but immediately turned red when the vehicle was switched off. My assumption was that the battery (although only a year old) needed to be replaced, and that the controller was a dud and needed to be replaced.

Today, out of the blue, I noticed the controller was again operating and that the cTek indicator light remained green even after the vehicle was switch off.

What's going on? Your thoughts?
Why does the cTek indicator flash green when the battery's own indicator is red?
Why would the MPPT stop working and then suddenly start working weeks later?

[inetdog]

Very good questions.

As to why the CC stopped working and then started again, the most likely reason is an intermittent defect of some sort.
However some CCs by design will not try to charge a battery whose voltage has dropped below a critical level and most will not work if you apply power to the PV input while the voltage on the battery side is too low to reliably operate the processor inside the CC.

The battery indicator, if the ball type, is measuring the Specific Gravity (SG) of the electrolyte rather than the battery voltage.
If the battery is badly damaged or has become sulfated the SG may be permanently low but the internal resistance can still allow the terminal voltage to come up to normal while the CC is trying to charge it.

Without knowing more about what your indicators are looking at and what the actual SG and voltages are from an independent measurement I cannot do any better at reconciling the two indications.

[Sunking]

On my vehicle I have 3 x 100w panels wired in series i.e. 300w in total. The panels are connected to an MPPT controller, which is in turn connected to a 105ah 12v deep cycle battery. The panels supply 5,5amps of current. Using a multimeter I've recording voltage coming into the controller for the panels ranging from 40,6v in overcast conditions to 46,4v in partly cloudy conditions. The battery powers two 12v compressor fridges that draw 2,5v each.

This is complete rubbish.

First 300 watts of panels into a MPPT charge controller will deliver around 25 amps at full power. At 5.5 amps with the panels in series tells you do not have a MPPT controller, you have a fake. Or you have no idea how it works.

The fridges do not draw 2.5 Volts. I can only assume you meant 2.5 amps, and that would be way too low for even one refrigerator. More likely the two fridge's draw around 10 amps and that being said your batteries are grossly undersized. At 105 AH is about 10 hours run time when they should be sized for 120 hours run time. 10 does not equal 120.

During a camping trip in extremely hot conditions (40*C) that made the fridges work overtime the battery drained to 7,3v overnight - measured with a multimeter a the MPPT outlet terminals. By mid afternoon the voltage resorted to 13,5v. However, the following morning the MPPT controller had stopped working overnight.

This tells the tale and explains everything. You severely over discharged the batteries. They were dead at 11 volts and severely damaged at 7.3 volts.

That means your system is grossly undersized, and sounds like you were sold a fake MPPT controller. Bottom line is your batteries are boat anchors and possible a destroyed charge controller from abuse.

[inetdog]

What SK said.

If the 5.5A with the panels in series is measured at the panels, then it almost makes sense. If it is measured at the output of the CC you have been cheated. Just change to a parallel wiring of the panels and you will get three times the charging current!
Get a real MPPT CC and you will gain an additional 30%.

[Sunking]

What SK said.

If the 5.5A with the panels in series is measured at the panels, then it almost makes sense. If it is measured at the output of the CC you have been cheated.

That is what I was eluding too. A 100 watt 12 volt battery panels IMP is 5.5 amps. A PWM Controller INPUT CURRENT = OUTPUT CURRENT.

What really stinks is the guy did get cheated. Any pro would have known the system would not work and sold it to him anyway.

[Accedda]

Thanks people. Truly appreciate rapid responses. I did mean 2,5amp for the fridges and took if from the fridge user manuals but maybe I misinterpreted the data. The 5,5 amps from the solar panels is supposedly on the input side of the MPPT but I don't know how to verify that. I suspect the MPPT is junk because it isn't branded! And I also suspected the battery needs to be replaced. You also confirmed my suspicion that if the battery charge is too low, the MPPT won't work. I found a seemingly reputable solar panel RV specialist and will let them audit and probably upgrade the system. They do work for NATO and the UN here in Africa. Thanks again.