Automatic DC supply selection?

[flippineck]

I have a solar array (4 x 80W panels in series) on the East facing wall of my house, and a matching array on the West facing wall. In the brightest summer sun I get about 5A at 80V. My controller has a maximum input rating of 100V/500W and it feeds a couple of 12V lead-acid leisure batteries.

One day I might decide to pay a professional company to come and fit panels way up high on top of the roof and a big grid tie controller but for now, a small wall-mounted DIY setup is more convenient. I'm trying to run a television sporadically, and a mini fridge.

At the moment I have the East facing panels hard wired most of the time for the morning sun, but sometimes (if I remember) I manually change the connections over to use the West facing panels in the afternoon.

Is there a simple arrangement of components I could use, to automatically select and connect to the charge controller, the bank of panels producing the most power at any given time of day? Whilst not overpowering the charge controller by connecting all 8 panels to it at once

I was envisaging the use of a relay or power transistor of some form, maybe in association with some form of comparator like an opamp, to flip between banks based on their voltage output? My thoughts are wooly and my circuit design skills are rusty (not that they were ever particularly great anyway!), if anyone could help illuminate or make any suggestions as to direction I'd be grateful to hear

 

[danadak]

A pair of high current Schottky diodes simplest, one in each output lead then cathodes tied together,
or a solution with even less loss :

https://www.analog.com/media/en/technical-documentation/data-sheets/4357fd.pdf

The above done for 48V solution, fine for your Voltage.


Regards, Dana.

 

[Delta Prime]

The above done for 48V solution, fine for your Voltage

As an addition to a simple, elegant solution. I must warn you,only because you are a fellow member of this fine community.
I'll let you in on a little known fact it is a secret do not tell anyone else .
In the united State's we have several satellites in orbit called heliophysics observatories their purpose is to detect or look for solar flares & C.M.E'S (Coronal Mass Ejections).
My concern for you is not solar flares but CME's.
If a strong CME is on its way towards Earth utility companies can redirect power loads to protect the grids.
Your controller input is rated for 100 volts, 500 watts.
Your solar panels Max output is 80 volts, 400 watts. Which leaves a cushion of 20 Volts, 100 watts.
This phenomenon of Solar flares & CME's can and do happen simultaneously.
A CME event by itself will wipe out your system.You got to beef It up.

 

[danadak]

As an addition to a simple, elegant solution. I must warn you,only because you are a fellow member of this fine community.
I'll let you in on a little known fact it is a secret do not tell anyone else .
In the united State's we have several satellites in orbit called heliophysics observatories their purpose is to detect or look for solar flares & C.M.E'S (Coronal Mass Ejections).
My concern for you is not solar flares but CME's.
If a strong CME is on its way towards Earth utility companies can redirect power loads to protect the grids.
Your controller input is rated for 100 volts, 500 watts.
Your solar panels Max output is 80 volts, 400 watts. Which leaves a cushion of 20 Volts, 100 watts.
This phenomenon of Solar flares & CME's can and do happen simultaneously.
A CME event by itself will wipe out your system.You got to beef It up.

What is solution to that ? I sense its the Panel production coupled with load dump
transient by utility causing excess V input to controller ?

By the way thanks for jumping in on this thread, safety first.


Regards, Dana.

 

[Delta Prime]

By the way thanks for jumping in on this thread, safety first

Jumping in, barge in, push aside, it's what I do.
He's off grid I'm talking "Ve" my neighbors on either side of me suffered the same fate destruction of their system they're off grid . I learned there were operating their photovoltaics and maximum with respect to the Max input of the controller . This happened twice and this was only solar flares event the second time their solution was to cover the photovoltaics. Given enough time as they did have enough time and we're home at the time it seemed totally random though cuz my neighbor across the street and their systems nothing happened to their systems. solution would comprise knowing exactly what equipment is being used by what manufacturer. But in practice isolation and as you say load power dumps, circulators etc. Anywho I think the big events happen every 11 years or so...

 

[flippineck]

I've already blown up one controller by not watching the power input carefully, don't want to do it again! Thanks for the warning

Would it be worth looking at a 100V varistor in parallel with the final input and a 5A fuse in series?

 

[John Canon]

OK, we had April Fool's Day, then we had May Fool's Day. Are we jumping the gun on June Fool's Day?
I am confused. Not really.

 

[Delta Prime]

OK, we had April Fool's Day, then we had May Fool's Day. Are we jumping the gun on June Fool's Day?
I am confused. Not really.

Most people who make open, non-specific posts like yours want an opinion and a discussion.
f ou only want confirmation and coddling, you should say so!

 

[AnalogKid]

Post #2 Shottkey diodes are a first step, but they will not give you the protection you want during change-over. There will be a point when both diodes are conducting because the output voltages of the two groups of panels is almost exactly equal. This is done on purpose when current-sharing two power supplies, part of a larger technique called droop-sharing.

One way to guarantee that the two sources never are on at the same time is with a control circuit that has a comparator with some hysteresis and two outputs, one for each power MOSFET. The two power sources are diode-ORed together to power the circuit through low power rectifiers. No need for a Arduino or other microcontroller, A/D converter, etc.

ak

 

[flippineck]

Because of the placement of the 2 banks, they never have direct line of sight to the sun at the same time. At noon precisely, the sun is at 90 degrees to both sets, and a significant time delay passes whilst the roof overhang's shadow transfers from the West to the East walls, engulfing both banks. Hopefully because of this, the point of equal voltage should not last too long, & coincide with minimal illumination. So even though both banks might contribute current simultaneously for a brief time it shouldn't be enough to blow the fuse.

Going to go ahead & try two 10A rated schottky diodes, a 5A fuse and a 100V VDR whilst reading further into the other ideas.

 

[flippineck]

Interesting patent

US6847249B1 - Highest available voltage selector circuit - Google Patents

A voltage selector circuit determines the highest of at least two available input voltages, and connects the highest voltage to an output terminal. The circuit includes a comparator having first and second FETs having their sources connected to first and second input voltages (V1 and V2), and...

patents.google.com

The inventor

Fig 1 looks like a solution, if built for a small hysteresis and no overlap of states?