Power Tweak – Solar Diverter

The blue battery bank was falling behind and the tan battery bank was fully charged by noon each day. As a solution a solar diverter was developed and installed to switch that excess solar power from the tan bank to the blue bank. Once state-of-charge falls behind on any of the battery banks it’s hard to catch up without some additional outside help. A solar diverter switches one array of solar panels from one battery to another. A nice feature of solar controllers is multiple controllers can be run in parallel to provide increased charge power to a battery. Hardware is a typical 30A automotive relay controlled by a 12 volt port on a Rigrunner 4005i. The station’s solar arrays each peak out at 18 amps so a 30A relay allows a good safety margin. Here is the white-board sketch and capacitor calculation.

Why the capacitor?

A relay typically takes 3 ms to switch from one state to another. During that 3 ms no battery is connected to the solar charge controller. With no battery a controller will shut down and pass through the 20 volt solar power directly to the load port by default. That would destroy the attached 12 volt equipment ( This has been learned the hard way unfortunately and more than one Rigrunner 4005i has been burned up ). Will a 470 uF capacitor from the junk box provide power long enough to bridge the 3 ms gap and prevent the controller from shutting down? Morningstar’s manual says the controller consumes 22 mA. Using ohm’s law that means the resistance is 614 ohms when battery voltage is 13.5 V. Plugging the known values into the formula for a time constant (T = RC ) produces a result of 300 ms or roughly a 100 times factor of safety. Happy with the results and moving forward with the installation, this is what the ridiculously simple solar diverter looks like.

After going live for two days no problems have appeared and the blue battery bank has increased it’s state-of-charge. It’s closing in on 100% probably in a day or two. Rigrunner 4005i’s have a built-in timer for each port so that feature enabled the solar diverter to easily be on a noon-to-8pm schedule. No spikes have been noticed.

Update 7/3/2021 – Rethinking the above modification with regards to cold weather. Each controller has a temperature sensor on the battery it serves. Let’s say it’s the middle of winter and the diverter is on, diverting power from the tan battery to the blue. How does the tan controller know the temperature of the blue batteries? It doesn’t. It only knows the tan battery. What if the tan battery is above freezing but the blue is below freezing? The tan controller would charge the blue batteries and damage them. Not good. The diverter is ok for summer but not winter.

Next Power Tweak

One of the batteries is still struggling occasionally even with the plentiful midsummer sunshine. The WSPR station is the only load on the Green batteries and that small load is too much. It runs down after a cloudy day. Extras like WSPR, KiwiSDR, Pixel loop antenna, NUC pc’s, security camera, weather station, and mobile hotspot seem to be the culprits system wide. Batteries do fine if the extras aren’t loaded on. Extras impose a heavy load because they tend to run 24 hours a day. For example the WSPR station only draws one amp but over 24 hours that’s 24 amphours. That’s enough to effect the power budget. Next power tweak is to try moving the extras to their own dedicated battery. Green battery is the natural choice because it is already dedicated to serving an extra device. It’s running down with one device so it clearly needs an upgrade. Currently it has two panels. How many more does it need?

All the extras total a daily load of 87 AH which has to be generated by solar during the 5 hour window of peak sunshine (engineered for the shortest day of the year). For 87AH / 5 HR = 17 amps per hour. Panels generate 6 amps during peak which means we need 3 panels ( 3 panels X 6 amps = 18 AH). Currently the green batteries have 2 panels. One more can be easily added ( for $100 ). Will the green battery have enough capacity to hold up through a cloudy day?

It was capacity-tested at 142 AH. Derating for a range of 20% to 90% state-of-charge, usable capacity is 99 AH. At the start of a cloudy day it has been 19 hours since it was fully charged and has discharged 68AH ( 19 X 3.6 AH or 68 AH ). That leaves 99AH – 68AH = 31AH for the next 24 hours until peak sunshine returns. Clearly not enough because 31AH is way less than 87AH. Although adding more battery capacity would be the obvious solution, the cost rules that out for now. Plan B? Shedding some load with timers or with turning off certain devices. Plan C? Just let the low-voltage-disconnect turn off the extras when the battery is dead. They would automatically be reconnected when enough sunshine returned. Here’s an estimate of adding adequate battery capacity.

How much is needed? The original batteries would need to be replaced because batteries of differing ages or capacities cannot be added on to. The cost for a new string with a derated capacity of at least 87 X 2 = 174AH, would be $261 as of this writing.

https://www.aliexpress.com/item/1005002921126637.html?spm=a2g0o.productlist.0.0.55d618efoRvn5Q&algo_pvid=1ac013a8-be97-4dc1-ab77-3a62dafcd672&algo_exp_id=1ac013a8-be97-4dc1-ab77-3a62dafcd672-3

Increasing battery capacity means more solar is needed, too. Battery capacity jumps to 174 AH. Dividing by 5 peak hours of sunshine the needed amps is 35A. The controller maxes out at 30Amps so a bigger controller is needed. A compromise is to stay below 30A and that would make the number of solar panels 5. That’s 2 more than the 3 already installed for an additional cost of $200. Total cost with a compromise is $461. Adding hardware and cables comes up to a good estimate of $500. Doing it right with a new controller is probably ruled out because Morningstar does not offer a higher current controller with low temperature foldback. Choice at this time is Plan B: adding one panel, keeping same battery and controller, and shedding load when needed. Cost $100.

Power Tweak No. 3

With the third panel added, all battery banks now have the same number of solar panels. It seems that transfer relay was superfluous and was removed. Once the loads were rebalanced and distributed more evenly all batteries are charging normally. The relay was repurposed to disconnect the solar panels from the controller when it was generating rfi. See the post on solar panels and rfi at:

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s