Diode help required (Creating solar and 12v charging cable)

[TheOtherShiny]

I need the same (matching) size/type diode on both inputs? And on just the positive side, like I have in my original design?

 

[rjenkinsgb]

They do not have to be identical, but high current and low voltage drop (or heatsinkable) is best.

I use 35A encapsulated bridge rectifiers for some high current stuff, they are very cheap for the ratings and one would work for what you need, though with a bit more loss than schottky diodes.

eg.

https://www.mouser.co.uk/ProductDetail/Micro-Commercial-Components-MCC/GBPC3510-BP?qs=y6ZabgHbY%252Bxu4fAzjTlszQ%3D%3D

Use the two AC inputs for the two sources and the + output as the combined positive out. Ignore the negative out.

Bolt it on to some convenient metalwork for cooling.

 

[Tony Stewart]

Just one from each source to block the lower voltage from getting current..

Ring = Cathode tied to V=DC in+

Two diodes are better than 4 in a full bridge with half the drop. and heat loss.

 

[rjenkinsgb]

Two diodes are better than 4 in a full bridge with half the drop. and heat loss.

The bridge is a convenient way of getting heatsinkable high power diodes cheaply.

Ignore the negative out

In other words, using the two diodes that have the cathodes linked to + out and ignoring the other two. Only a single diode drop in each circuit.

 

[Tony Stewart]

Good point RJ. I added a diagram for his confirmation and Iagree this is an easier device to thermally heatsink (like an old CPU heatsink) and insulate and they are cheap.

Just to clarify for others, you can't expect MPT max power transfer with such a wide range of input voltages as I initially said, but at reduced powers, it works.

This is due to the max junction temperatures being current limited.

 

[Nigel Goodwin]

Good point RJ. I added a diagram for his confirmation and Iagree this is an easier device to thermally heatsink (like an old CPU heatsink) and insulate and they are cheap.

But it just looks SO wrong

Damn clever idea though.

 

[TheOtherShiny]

I'm pleased that you guys are enjoying solving this problem!

If you leave the 12V input disconnected but with everything else connected - and no sun - what voltage do you get at the Ecoflow input?

Without actual loads connected, a digital multimeter may be reading from minuscule leakage currents that would have no effect with all parts connected.

If you get little or no voltage at the Ecoflow input, then try connecting something as a load at the 12V connector - a resistor, 10K or so, or a a low power filament lamp, and see what voltage you get across it with the solar panels producing a good output?

I checked the PV voltage at 22h (dark) this evening the reading was 10.4V. I tried a 10K resistor at the 12V connector and it reduced slightly to 10.08V. I can try again at midday tomorrow, but I'm not really sure what I'm trying to measure.

Would adding an additional fuse at the 12V connector act as a secondary protection against potential damage? There is already a 10A fuse in the main fuse block of the van (but it's at the front and the connector is at the back). And only 2 or 2.5mm2 cable used.

 

[Tony Stewart]

If the 45V PV is optimal, that is usually 82% of the NO LOAD or open circuit voltage [Voc].

As sun declines , the max power efficiency point of ~ 82% drops to 72% of 49V?

So anything around 10V is like 20% of Voc and thus will be very little power and cannot even tolerate a 10K resistor being added as the current is so low. Thus the useful solar charge voltage will be from 35V to 46V. Any less or more might be just a trickle of current.

Don't forget the heatsink. and case must be insulated from voltage.

 

[Diver300]

But it just looks SO wrong

Damn clever idea though.

I've used a bridge rectifier before for just two of the diodes for the same reasons of cheapness compared to two diodes and ease of heat-sinking and mounting.