Trying to track down the problem

My problems with the solar power system on the bus likely come from one or two sources. In order to track this down I put my watt meter on the load from the charge controller. Here’s what I found...

• The current usage from the door unlocker is between 0.01 and 0.02 amps which is in line with the stated 160ma.
• The current usage from my bigger LED bulbs is between 0.09 and 0.15 amps.
• The current drain on the 2.1A setting of my USB charging socket is about 0.9 of an amp
• The current drain on the 1A setting of my USB charging socket is about 0.88 of an amp
• The current drain of my smaller LED bulbs is about 0.03 of an amp
• The current drain of my shower pump is about 2.8 amps
• The current drain of my CPU circulation fan is about 0.25 amps
• The current drain of my twin ventilation fans is 3.8 amps

I think we are all probably agreed that is not really very much power at all. I am, however surprised by a couple of things.

• The CPU fans and larger LED bulbs draw slightly more current than I had expected.
• The twin ventilation fans and the small LED draw less than I had expected.
• The shower pump draws less than I had expected

Overnight the watt meter should have recorded far more power usage than it did, leading me to suspect the system ran out of power and shut off. This is just a theory - though I was awake at 5:00 I did not go to check.

My total current draw - assuming everything is turned on...

• 4 big LED lights 0.6A
• 2 small LED lights 0.06A
• 4 twin USB chargers 7.12A
• Both ventilation fans 3.8A
• Both circulation fans 0.5A
• The shower pump 2.8A
• The door unlocker (not counting the unlocking relay/motor) 0.02A

A grand total of 14.9A plus the unknown door unlocker. If I lay under the bus to activate it, in the 3 seconds it takes to work, my watt meter would not adequately record the amperage. Not only that but I physically could not roll out from underneath, rush up the stairs and down the aisle in 3 seconds.

Given that the battery voltage drops very quickly but hovers around 13.1v when seemingly fully charged, it’s a fair guess that there’s something very wrong with the Harbor Freight battery. It is supposed to be 35AH. There have been many anecdotal accounts however of it being far less than that with some estimates going as low as 15AH.

The charge controller is not exonerated. The charge controller is a diabolical nightmare. Looking around though there are quite a few other options. The MPPT controllers interested me but I was perplexed by the size and the massive heatsinks needed. Indeed one fellow commented that he had an EPEver charge controller that caught fire after a couple of months usage. That was a little alarming and as no details were forthcoming I don’t know whether he did something silly, whether it was a faulty unit or a faulty design. 

According to what I’ve seen from another reviewer, the MPPT controllers work as PWM until they have more than 10W of solar power as an MPPT conversion would be inefficient. Clearly then there’s a bit of power loss in an MPPT system that probably wouldn’t be noticed much if one was not normally running on 35W. I have two extra 30W panels but they’re not designated for regular every day use.  

Checking the system voltage with an independent voltmeter then checking it again with another independent voltmeter showed that the onboard voltmeters are inaccurate. These are the button voltmeters from China. Mind, the hand held voltmeters (one from Harbor Freight and one from Walmart) are hardly paragons of virtue both also hailing from China. The onboard voltmeters measured 13v while the handheld measured about 11.9 to 12v. That’s pretty much an empty battery.

The interesting thing is that the charge controller has three dots for the battery level and two are lit at 11.9-12v which to my mind should just have one LED illuminated when the battery is that low. Sadly, today and tomorrow are scheduled to be fairly overcast. I have my extra solar panels out, waiting to see how much power I can pull in. The panels have been reading 19 volts most of the day during the dry patches. I might have to do a part two of this entry tomorrow. I don’t want to pull the battery and put it on a car charger just yet. That’s a plan for another day.

I wish now that I’d bought two of these watt meters. They’re very handy when it comes to debugging a solar system. I’ve established my solar panels all generate pretty much what they should be and that the charge controller reduces what’s coming it, Putting the one I have as shown here between the battery and the devices really showed me how much power I’ve used and how much power each device uses.

My cunning plan now is to wait until the batteries are full then put both fans on and one light. The flashing of the light will tell me even if the fan noise desn’t when the battery is out of power according to the charge controller. I just time from when I turn the fans and light on until the fans and light go off. I check the current consumption a few times during this and know exactly how much power has been consumed between the point the charge controller declares the battery full versus empty.

In theory I should get 60% out of a 35AH battery or 21AH. That means the fans should run together with the light for exactly 5 hours and 7 minutes. I guarantee this won’t be the case however. I’m not even sure the charge controller will be able to fill the battery.

Transferring the watt meter to the solar input from the twin 30W panels was interesting. That proved very little current was going into the charge controller. Putting a 2.5W bulb onto a portable 5W panels produced plenty light from the bulb - just holding the panel up to the cloudy sky. It looks very much as though the charge controller is the bottleneck in the system.

There could still be two outcomes and not just one. The battery could indeed be bad and the charge controller could be bad. Indeed, I’m thinking favorably given the worthlessness of using MPPT on such a small system of just getting a linear controller. They’re pretty cheap and as we already know, I have an absolute maximum current burn of 15A. The extraction fans spike over 5A because they blew. 5A fuse and I had to replace it with a 7.5A fuse.

In other news, I found my missing pack of LED bulbs. These are the 1.5w bulbs that I prefer in the bathroom and back closet. Whether I’ve mentioned it or not I did also manage to complete the wiring for one of my cigarette lighter plugs. It was a significant challenge but it is done. The second plug seems to be missing its collar. I know its here somewhere and I might still find it.

At the moment, the finger seems to be pointing at the charge controller as being the problem child. More on this as I proceed!