even out spikes comming from solar chargecontroler?

I have I solar panel on the roof of my van. With a MPPT chargecontroler I charge my 12v semitraction battery. This battery provides my 500watt inverter which provides my fridge.

My inverter has a max voltage input of around 15v and then that over voltage protection kicks in. I have to reset the inverter manually.

This happens when there is a big cloud passing on a sunny day. When the sun comes from behind the cloud the chargecontroler has a output spike around 14.8 volts for just a brief moment.
I measured this with my arduino taking samples every 500ms.

So is there a way I could even these spikes out so the overvoltage won't kick in? I know for arduinos we use capacitors to smooth things out, would that work here too?

I can specify the panel inverter and battery if needed.

How old is the storage battery? Sounds like it is not in proper condition. A storage battery acts like a very large capacitor. Your description makes me think it has a high internal resistance.

How many hours can the battery supply current to the inverter before the voltage drops to an unusable value?

Paul

2 months old now.

Battery can supply inverter without load "forever". With load, no idea. Fridge runs a couple of minutes every hour regulated by arduino. How much it consumes, I don't know. The peak at start-up is "unknow" by manufacturer.

Note this only happens when the battery if fully charged, float charging at 14v.

Bringamosa:
2 months old now.

Battery can supply inverter without load "forever". With load, no idea. Fridge runs a couple of minutes every hour regulated by arduino. How much it consumes, I don't know. The peak at start-up is "unknow" by manufacturer.

Note this only happens when the battery if fully charged, float charging at 14v.

Well, that certainly shoots that theory down!

Suggest a fairly large capacitor across the input to the charge controller. Perhaps that would slow down the voltage increase.

Paul

Just an idea. Haven't thought it through. Seems you need something to lower that voltage temporarily - it seems a diode (that can handle your current) would be enough to keep it under the peak.
But you don't want this normally, so a MOSFET comes in as bypass (in parallel to that diode).
Then the Arduino can monitor the voltage to the battery, and when it goes over say 14.5V switch off that MOSFET, forcing the current through the diode, dropping the voltage to the inverter by some 0.5V. Should be enough to keep it from tripping the safety.
If this Arduino can react quick enough you could this way prevent the overvoltage from kicking in.

Paul_KD7HB, what would be a fairly large capacitor?

Wvmarle,also sounds like a good idea. I habe my arduino measure every 500ms. This does record spikes. So a faster sampling might be possible. Will have to look into mosfets and diodes able to handle some current. Quite some amps flowing because of the lower voltage

Indeed, the current will be a problem. A number of diodes in parallel may be a good solution, depending on the actual current involved. They’d need a heat sink (e.g. 30A * 0.5V = 15W dissipation, even if for a few seconds that’ll quickly overheat a part).

At 500 ms I’d start calling it a surge rather than a spike :slight_smile: That you see them while measuring at 0.5 second intervals means the spikes last for some time.

You probably want to react much faster than that. An interrupt comes to mind - whenever the surge begins the interrupt triggers, and all the ISR does is setting an output pin which in turn switches the MOSFET. That’s the fastest possible.

That interrupt can come from the built-in analog comparator, or by using an external comparator (OpAmp). Voltage divider on your battery line on one input, fixed voltage divider (use a trim pot to adjust it) on the other input, output to an Arduino pin. Then in the normal sketch continue to poll that voltage, and (optionally with delay) switch the MOSFET on again.

schematic.png

This is a simplified schematic of what I have in mind. I have never tried this; I think it works as intended but it can have all kinds of other results so take care when testing. Looking forward to input from others on this!

Drawn just a single diode; more can be added in parallel as needed to handle the current. The MOSFET the same; a single one is probably enough (there are power MOSFETs that can do 50-80A), but maybe you need a second one.

I would forget about 'solutions' with diodes and mosfets, and track down the real problem.
14.8volt is not uncommon in a 12volt charging setup.
Did you check the manual/datasheet of the inverter for it's low/high cutoff voltages.

Is this a professional install, or DIY.
Does the MPPT charger AND the inverter connect directly to the battery (battery the center of the 'star').
Not solar>inverter>battery.
Leo..

Input voltage, 11 to 15 volts.

So cutoff at 14.8 measured, but once every 500ms so it might reach the 15v. However every measurement resulting in cutoff was at 14.8v

It is panel to chargecontroller to battery. Battery to inverter. So both directly conneced to the battery.

Solar to inverter is just not possible with the 30+ volts coming in, not sure where my explanation was unclear to make you think that. Then again, English is not my primary language so some sentences might be formulated in a strange way.

Did you even read that data sheet? This is the first of the features on top of page 1:

2.5V to 5.5V Operation

OP needs 14V.

Bringamosa:
Solar to inverter is just not possible with the 30+ volts coming in, not sure where my explanation was unclear to make you think that.

Not takling about that.
Wiring and terminals have resistance, so if the charge controller is first connected to the inverter before going to the battery, then inverter voltage could be a few tenths of a volt higher than battery voltage.
If battery is 'in the middle' than that's ok.
So is this a DIY install?
Leo..

It is a diy install yes, so I know what is where if I want to work on it.

All wiring 9,27 mm thick that is 67,4 mm2. That is AWG 2/0 ?? Not sure about the American wire gauge
Less than 50cm long.
Wire from mppt to battery is same length as battery to inverter.
Connectors are crimped at the show where I bought the wire
Copper bussbar at the battery connectors

And there is cables running directly from the battery to 8x 12v lights and fan. But guess that won't do much for the mppt voltage output.

Seems you have connected it right.
Since lead/acid cell voltage can spike to 2.75volt (16.5volt), I think the inverter is to blame.
High voltage shutdown should have been designed for 16.5volt, not 15volt.
Is this a cheap Chinese unit, or from a local supplier that you can call.

If you are going to solve this yourself, then I suggest some sort of shunt regulator that dumps the spike into a load.
Leo..

This one is from Conrad, it is the e ast cl500 12v
So yeah it is fairly cheap, it only had to keep my fridge running. Not sure how the cut off is set. I read about the low voltage cut off being between 10 and 11 volts. This sounds like it is manually set and not beving doen carfully... pot maybe? I might be totaly wrong about this of course. But if not, maybe the max voltage cutoff is adjustable too?

Will research a bit more see if I can find out more about this model.