diy solar

diy solar

One MPPT per Panel? What are the downsides - upsides DC / AC

eXodus

Solar Addict
Joined
Jul 27, 2020
Messages
1,462
I'm just getting a bunch of quotes for my house Solar System. Most of them are Enphase Microinverters, which are AC coupled.
Now my question, can you do this DC? I want to do build a DIY power-wall anyhow. Cost aside.

I have a weird roof shape with shading, so one or two MPPTs for 8KW would struggle. Will very likely have the panels on 4 different plains of the roof with (two East, one south, one west)

What about one MPPT SCC per panel or say two panels per SCC?

Downsides: - lots more wires to run. more wire loses

Upside: Better tracking from each panel, output data from each SCC, redundancy, single panel or charge controller failures do not have a big impact.

What potential issues can you see?
 
You are better off using an optimizer on each solar panel. That will maximize the output of each panel, but still allow them to be wired in series. Take a look at Tigo Energy TS4-A-O. They give you Rapid Shutdown (required by NEC 2017), optimization, and module level monitoring. You can connect the output to either a charge controller or a grid tied inverter, depending on what you're doing.
 
You are better off using an optimizer on each solar panel. That will maximize the output of each panel, but still allow them to be wired in series. Take a look at Tigo Energy TS4-A-O. They give you Rapid Shutdown (required by NEC 2017), optimization, and module level monitoring. You can connect the output to either a charge controller or a grid tied inverter, depending on what you're doing.

Thanks for the info. Ok, I didn't explain it good enough.

Can I do this all with direct current ? Multiple MPPT charger, no Inverter, no grid. A full DC system. no AC whatsoever.
 
  • Like
Reactions: Dzl
My less than experienced answer is "Yes". However, you need to have enough panel voltage going into an MPPT to exceed the battery voltage. For the Victron controller it's something like 5 volts over the battery voltage to start, after that it's a lower voltage threshold to keep charging.
 
My less than experienced answer is "Yes". However, you need to have enough panel voltage going into an MPPT to exceed the battery voltage. For the Victron controller it's something like 5 volts over the battery voltage to start, after that it's a lower voltage threshold to keep charging.

That is something I haven't though about Thanks. 60 cell panels are at 40 something volt and 72 cell panels are barely just above it. So that would not work for a 48V battery system.
Both
 
There are two ways to get independent MPPT optimization per panel. One is by micro-inverter, other is DC optimizer.

A micro-inverter is a mini-GT inverter self contained. Every panel is an independent entity on the grid. They typically have communications links that allow monitoring each panel. Theoretically best solution but there are some issues. One challenge for all GT inverters is smoothing out DC current loading on PV panels. Allowing sine squared current profile from single phase grid to pass through to ripple current current loading on panel results in less power from panel. This requires rather large capacitors to smooth out current flow from panels. For small micro-inverter it often means electrolytic caps. Electrolytic caps are one of poorest reliability components in today's electronic equipment. Putting them in a hot box on a panel subjects them to accelerated aging.

A DC optimizer is a DC to DC converter that contains its own MPPT search adjustment. The theory is every panel can have its own MPPT point and the electronics of the DC to DC converter and MPPT control is minimal. I do not have personal experience with them but the question is can they react in the best way to all the possible variables that may arise being put on series connected panels. Probably one of the biggest question marks in my mind is how a given central GT inverter's MPPT algorthym will react to the different characteristics presented to it by the DC optimizers. SolarEdge has a setup that requires their optimizers which are their sole MPPT adjustment point and their own MPPT-less central inverter. Problem with this is you are tied to single manufacturer of a somewhat unique setup. Read reviews on your own but avoid putting too much faith in reviews from a company selling them.
 
As I've shopped for solar panels, I saw a few that had the micro inverters. I don't know anything about the micro inverters and RCinFLA's post made me feel better about avoiding micro inverter panels.

It seems that an in depth video on micro inverters would be right up the alley of @Will Prowse .
 
There are two ways to get independent MPPT optimization per panel. One is by micro-inverter, other is DC optimizer.

Thanks for the info.

Ok again, I didn't explain it good enough. I do NOT want to use Inverters.
This is an academic question, I'm exploring the option of using Direct Current MPPT Charge controllers - one for each panel (or two).

No alternating current whatsoever in the system.

Something like 24 Panels with 12 Victron 20A MPPT charge controllers directly wired to a battery.
 
If you have a battery charge controller you have a switching power converter, just not to higher voltage and not having the final stage of sinewave PWM conversion a grid tie inverter does. It is just a buck DC to DC circuit usually.

The greater the series stacking of panels the higher the input voltage to controller and there is slightly worse conversion efficiency the greater the ratio of PV side voltage to battery voltage. Benefit is less wire gauge required for lower feed current at higher PV voltage. Most charge controllers manufacturers state their efficiency at the optimum input to output voltage ratio. You rarely find a chart showing actual conversion efficiency for your particular setup. In all fairness, it is complicated as it not only depends on input to output voltage ratio but also the current levels being converted.

Same issue with DC optimizers of possibly confusing the battery charge controller's MPPT algorthm. If you don't stack more then three panels and orient such the series group of three to avoid shading you can parallel multiple series string without too much concern on different MPPT points. DC optimizer will not buy you much for that arrangement. Only cost you perhaps a bit larger gauge PV feed wire.

DC optimizers are also another energy conversion with its associated efficiency loss in the conversion.

Just academically, I have not seen such an animal for but for battery charge controller, a panel optimizer that pumped panel voltage up by 3x to 6x of panel voltage, that are then tied on a panel parallel bus might not be a bad idea.

The market is driven by GT inverters. They greatly outnumber direct battery charging controllers.
 
Last edited:
As I've shopped for solar panels, I saw a few that had the micro inverters. I don't know anything about the micro inverters and RCinFLA's post made me feel better about avoiding micro inverter panels.

It seems that an in depth video on micro inverters would be right up the alley of @Will Prowse .

I really don't have any big issues with micro-inverters. They have their place. DC optimizers I am still uncertain about.
 
The greater the series stacking of panels the higher the input voltage to controller and there is slightly worse conversion efficiency the greater the ratio of PV side voltage to battery voltage. Benefit is less wire gauge required for lower feed current at higher PV voltage. Most charge controllers manufacturers state their efficiency at the optimum input to output voltage ratio. You rarely find a chart showing actual conversion efficiency for your particular setup. In all fairness, it is complicated as it not only depends on input to output voltage ratio but also the current levels being converted.

The market is driven by GT inverters. They greatly outnumber direct battery charging controllers.
Thanks! this gives me something to work of. Efficiency of controllers at various voltages.

Is a MPPT Charge controller for a 48V battery better at 60V or 120V or 180V? Good thesis to explore.
Further Charge controller have an idle draw, is this higher for one big unit or multiple small ones?

The small all are passive cooled, while the big ones have fans.

For Grid Tie inverters the market is trending to micro inverters, with 1 or 2 or 4 panels on each.

But for mobile applications, where the environmental factors change all the time, there might be value for a multiple MPPT setup. In the end you could just print various MPPT circuits on the same board and put them in the same housing.
 
Thanks for the info.

Ok again, I didn't explain it good enough. I do NOT want to use Inverters.
This is an academic question, I'm exploring the option of using Direct Current MPPT Charge controllers - one for each panel (or two).

Have a look at Genasun MPPT controllers, they are basically designed for this use case I believe. Small Victron controllers are another option.

What you are describing is not common with houses or structures, but is semi-common in the marine world.

As someone else already noted, you want to make sure the PV voltage is optimized for the MPPTs, in the case of normal MPPTs that means a positive voltage differential, in the case of Genasun I think it depends on the controller (maybe @SolarQueen can elaborate, I know she sells them).

Another options is the Electrodacus DSSR20 paired with the SBMS0. These small and cheap controllers are designed for 2 x 60 cells panels each, and many will work together in parallel.
 
Have a look at Genasun MPPT controllers, they are basically designed for this use case I believe. Small Victron controllers are another option.

What you are describing is not common with houses or structures, but is semi-common in the marine world.

As someone else already noted, you want to make sure the PV voltage is optimized for the MPPTs, in the case of normal MPPTs that means a positive voltage differential, in the case of Genasun I think it depends on the controller (maybe @SolarQueen can elaborate, I know she sells them).

Another options is the Electrodacus DSSR20 paired with the SBMS0. These small and cheap controllers are designed for 2 x 60 cells panels each, and many will work together in parallel.

Thanks! Interesting products. That wiring https://electrodacus.com/DSSR20/DSSR20.pdf is just about what I had in mind.

Permitting in the US is terrible. When you go all DC - even when it's more dangerous - you do not need any permits (yet)
It takes months of paperwork for solar system any trained chimpanzee can put together in day.

It makes solar double as expensive installed as in the remainder of the world.
 
Back
Top