SMA String VS Solar Edge with Power Optimizers

[ButchDeal]

I just do not like the whole serviceability issue with a distributed architecture. What no one seems to think about is that sooner or later, someone will have to go out to replace these optimizers one by one as they croak. Why marry a super reliable solar panel with a chunk of electronics that is less so - in a hard to service location? I was the first guy in Arizona to go whole hog into the SolarEdge mindset and learned first hand (the early versions had a horrible failure rate) that I do not want to be the guy responsible for servicing all these optimizers.

Actually the optimizers are very reliable. The early versions had lots of electolytic caps, the current ones have none of those in them.

[bcroe]

Actually the optimizers are very reliable. The early versions had lots of electrolytic caps, the current ones have none of those in them.

Any switcher puts an awful strain on electrolytic caps; designing them out should really help. I don't know any
details about optimizers. But if they are a variation of the well known buck converter, they should be FAR simpler
than any inverter. Multiple phase bucks and high frequencies would help. I would avoid them here, with little
shade (where is my chain saw), as they will be another source of radio interference. Bruce Roe

[ButchDeal]

Any switcher puts an awful strain on electrolytic caps; designing them out should really help. I don't know any
details about optimizers. But if they are a variation of the well known buck converter, they should be FAR simpler
than any inverter. Multiple phase bucks and high frequencies would help. I would avoid them here, with little
shade (where is my chain saw), as they will be another source of radio interference. Bruce Roe

This is older but shows some of the internal components:

https://www.solaredge.com/sites/default/files/technical_comparison_microinverters_power_optimizers-na.pdf

[bcroe]

This is older but shows some of the internal components:
https://www.solaredge.com/sites/defa...imizers-na.pdf

Can't tell for sure, but the quite high efficiency and limited # of parts implies a non isolated buck or
buck-boost circuit. What they don't mention is that the optimizer losses must be added to the supporting
inverter losses, bringing them down to or below micro efficiency. The system design certainly gets more
complex, but in a varying shading or clipping situation, the optimizers should do better.

There has been the thought of improvising a home brew optimizer, for a complete string subject
to earlier gradual shading. But so far it has been easier to just cut the tree. Bruce Roe

[ButchDeal]

Can't tell for sure, but the quite high efficiency and limited # of parts implies a non isolated buck or
buck-boost circuit. What they don't mention is that the optimizer losses must be added to the supporting
inverter losses, bringing them down to or below micro efficiency. The system design certainly gets more
complex, but in a varying shading or clipping situation, the optimizers should do better.

The current optimizers are 99% efficient and the HDWave inverters are 99% efficient so the combined efficiency is above the micro and most string inverters.
The buck-boost system allows them to have a much wider operating range.

There has been the thought of improvising a home brew optimizer, for a complete string subject
to earlier gradual shading. But so far it has been easier to just cut the tree. Bruce Roe

No shadow outperforms optimized with shadow any day.

[jflorey2]

Can't tell for sure, but the quite high efficiency and limited # of parts implies a non isolated buck or
buck-boost circuit.

Yeah, there are a few reference designs out there for solar optimizers based on the Solaredge products. They are all four switch buck-boosts.

[bcroe]

The situation here might be a string some 30' long, and the panels get shaded one by one. A
"string optimizer" might keep the remaining unshaded panels working a while longer at their
full efficiency; the shaded ones going to bypass. Not every string here needs this. Bruce Roe

[jflorey2]

The situation here might be a string some 30' long, and the panels get shaded one by one. A
"string optimizer" might keep the remaining unshaded panels working a while longer at their
full efficiency; the shaded ones going to bypass. Not every string here needs this. Bruce Roe

Have you looked at the Tigo optimizers? They can be used on a single string and the output still connected to a string inverter.

[steveholtam]

I've never liked the argument that it's hot on the roof, and the parts will die. These electronic boxes are designed to withstand the heat. Heat on a roof is not an unknown factor. Car manufactures have all sorts of electronics under the hood that are punished by heat for decades and they keep on working. Only time will tell, but so far there does not appear to be heat issues with the newer models.

[bcroe]

Have you looked at the Tigo optimizers? They can be used on a single string and the output still connected to a string inverter.

That will something to consider, thanks. First I'll finish moving/adding panels, and removing shade. Then see
just what benefit might be possible. It seemed like a fun project, the strings are 3KW, a little bigger than any
switcher built here before. The idea of uneven size strings is interesting, now all strings have 720 cells. Bruce Roe

[solarix]

The original SolarEdge optimizers had no electrolytic caps in them. I checked before getting involved with them and was the reason I thought they would be better than the Enphase design. Still had a bunch of them fail within two years. Oh ya, car electronics and high under hood temps. That's a good one - how long do auto makers warranty their stuff?

[Mike90250]

car electronics and high under hood temps

A short warranty (5 years) and low power (as in 10's of watts) vs 100w+ And the complex parts are usually in the passenger cabin

[jflorey2]

I've never liked the argument that it's hot on the roof, and the parts will die. These electronic boxes are designed to withstand the heat.

Failure rates double per 10C increase in temperatures.

Heat on a roof is not an unknown factor. Car manufactures have all sorts of electronics under the hood that are punished by heat for decades and they keep on working.

Nowadays high temperature/high power power components under the hood (i.e. inverter FETs, rectifiers, inductors) are actively cooled either with fans or liquid coolant systems. On-roof systems usually do not have those sorts of mitigations.

Note that there are systems (down-hole monitoring for example) that do indeed work at extreme temperatures. They often have lifetimes of weeks to months.

[ButchDeal]

The original SolarEdge optimizers had no electrolytic caps in them. I checked before getting involved with them and was the reason I thought they would be better than the Enphase design. Still had a bunch of them fail within two years.

The newer optimizers have been out for some time now. We have have very few failures. The failures we have had can easily be traced to a few installers with poor practices, easily corrected with training.
The equipment is improved and proven, Do you base your auto purchase decisions on the reliability of cars made in the 1970s?

[ButchDeal]

low power (as in 10's of watts) vs 100w+ And the complex parts are usually in the passenger cabin

neither is typically true of any hybrid system.