I need a way to cut the connection between my solar array and the solar charge controller.
Solar array has a max voltage of 346V, and a short circuit current of 27A.
The solar charge controller limits the max current to 22A.

The situation is that I've moved to a farm where a mountain blocks the sun from the house for 10 weeks during winter. Hence I mounted to solar array 75m from the house. Because of the distance, I had to aim for higher voltage and lower current in order to reduce losses. There are not many solar charge controllers in this market, so I bought a Suyeego 480V / 120A controller. This controller has a self consumption in low light conditions of 5W, while the alternative controllers consumes 50+W. This makes the difference if you get any charging on cloudy days or not.

The Suyeego controller however does have one issue I need to solve.

The logic in the controller is build up of a voltage sensor on the solar input. When input voltage is above a fixed level, a relay enables and the solar input is connected to the charging logic. When the solar voltage drops below a set level, the relay disables and disconnect the solar input. In theory it sound good. What happens in practice, during low light conditions, is that the solar gets connected while the solar input is too low to contribute with any charging. Once connected the voltage drops and soon the logic realize the voltage is too low and disconnects the relay. The result is that the relay cycles on/off every 5 seconds. On a regular day this goes on for one hour during sunrise, and another hour during sunset. And on a really cloudy winter day this cycling may go on all day during 6-7 hours of daylight. With such frequent relay cycling, the relay will wear out fast.

The correct way to solve this would be to reprogram the controller to change the trigger levels of the build in relay. But I can't figure out how to do that. Hence I'm looking for an alternative solution. What has come to my mind is to put a relay in front of the solar input to the charge controller. I want a voltage sensor and logic in front that enables this relay when solar voltage is above 298V, and disables the relay if voltage drops below 58V. As this relay should switch only when current through the cable is near 0A (much less than 0.1A), the switching load of the relay, and there by arcing, should not be an issue. If it can hold off 350V in OFF-position, and pass 25A in ON-possition without overheating, it should be able to do the job. This is because when the relay connects, the charger will spend a few seconds before any load is applied. And when the relay disconnects, it is because there is no current flowing.

BUT if the control circuit would fail during charging on a sunny day, the relay would jump from ON to OFF with a load of 22A and a voltage of 346V DC, and the arcing will be significant and possible catch fire.

What can be done to reduce this risk of fire in case of control circuit failure?

The cables between solar panels and charger are twisted in order to reduce inductance. But still the circuit must be considered inductive.

Will a capacitor over the relay terminals do the job? It will charge when relay turns off. But wouldn't this introduce an issue with capacitor quickly discharge when relay connects, so it introduces arcing issue when relay connects?

Another solution that came to my mind was to invert the relay function - using a NC relay, and the control circuit that activates the relay breaks the solar input rather than connecting the input. With such approach, a control circuit failure during high power charging will not affect the relay at all. But it will end up powering the relay all night, and possible also all day during the darkest days. Those relays I've seen that could possible be used for this consumes 3w. Is there an option to reduce this power consumption?

I guess what I'm looking for is either:
1) A relay that could handle to be switched off during max solar load (22A, 346Vdc)
2) A snubber circuit that could make a smaller spec'ed relay work
3) A NC relay with low power consumption
or
4) A much better idea for how to work around this issue

As I think the voltage sensor circuit must be powered from battery side of the charge controller, the solar negative cannot be disconnected from the charge controller. Solar disconnect must be done on the positive wire. This complicates the potential usage of a NMOS to replace the relay. (Unless someone has a great idea for how to sense voltage that is 340V below the sensor circuit ground level.)
An other option could be to power the sensor-/control circuit from the solar input directly, but step down converters on such high voltages are rare. (They are said to become unstable.)

 

The correct way to solve this would be to reprogram the controller to change the trigger levels of the build in relay. But I can't figure out how to do that.

did you ask the manufacturer for assistance? it does not hurt to try - i work with wide range of automation products and i tried such approach in the past - several times it worked ... i got modified firmware and update instructions. in one case I asked for completely new feature (a bit bigger software mod) and for that i had to pay of course.

 

if the internal relay is accessible, then a modification can be done to the coil circuit only. this way existing relay that is already rated for the robot is used, and change is only on low voltage side.

 

An other option could be to power the sensor-/control circuit from the solar input directly, but step down converters on such high voltages are rare. (They are said to become unstable.)

Most low wattage 'universal' 85 - 240v AC/DC switch mode power supplies will happily work from 346v DC input.

 

I have sent email to the manufacturer. No reply yet (and I don't expect any. Thats the downside of buying from china.)
Doing mods to the internal relay control has also been on my list of ides. I was thinking breaking the path to the relay control on the pcb, and route it through another relay that I'm in control of. However I think the controller will get confused in some way when it sees solar voltage, gives signal to enable the relay, and nothing happens.
So my next thought in that matter was to break the path of the voltage sensor for the solar input with a relay so that the controller reads 0V solar whenever I want it to not trigger the relay. This sounds doable, but I haven't figured where this sensor connects. I expect to see a voltage divider somewhere, but so far I have not found it. (I'm a noob at reading circuit boards.)

Using a 240VAC -> usb charger might be a good idea. So far I've only tried two such chargers, and they both use a transformer to lower the AC before converting to DC. These do not work with DC input. Is there any sign I could look for in the store to find a model that does not use a transformer?

 

Many cheap USB chargers use a capacitor 'dropper' rather than a transformer - there will still be a transformer in there but that's part of the DC/DC conversion - you
can tell because it'll be tiny, no more than a couple of cm size cube, far too small to be an AC mains transformer.

You can't go wrong with a Meanwell PSU, they're an industry staple - here's a snippet of the catalog for their RS series of low wattage, low cost AC/DC to DC SMPS - I've highlighted the input spec for the smallest in the range, the 15Watt units. You can find these on Amazon or similar for around 7 or 8USD.

 

That looks like the device I should use to power the logic directly from the solar input. That way the power save function would be acceptable. When solar goes out, the relay does not consume any power.

The issues around the relay and the load still reminds though. Does it make sense to use a NC relay? Only reason being that if the control circuit fails (relay loses control signal) during high amp charging, the relay will remain in connected state,
Would it make more sense to use a snubber circuit?

 

Another option for when the solar array is too high is a series resistor, with a relay to bypass the resistor when the voltage is not excessively high. Power the relay from the solar array to open when it is powered. Certainly it is not efficient at all, BUT it is wasting power that you are not able to use. The resistor will burn lots of power and get quite hot, so it needs to be safely away from everything.