Pros/cons of the non-Enphase microinverters

[zanydroid]

I wanted to get some more information about all of the different non-Enphase microinverters. Both from a cost/performance standpoint as well as things like ecosystem and support

Manufacturer list

• AP Systems
• Hoymiles
• Others?

In terms of what I've found so far unique to the above.

I can see that both provide minimal direct support for DIY. AP Systems actively discourages it, but still has instructions on how to enroll as an individual. Is this discouragement more of a covering-their-ass / avoiding the support costs, or is it stronger than that? As far as I can tell, Hoymiles just ignores DIY and pretends it doesn't exist (haven't stumbled across any DIY info on their website).

Hoymiles still has 4 port microinverters while AP systems is 2 port only. My understanding is that the cost is significantly less due to the simpler circuitry vs the 2-port ones -- 4 DC inputs but only 2 MPPTs instead of 4. So there is an additional dimension for value engineering.

Hoymiles has zero-export capability with control over RS-485, and there's experience on this board on how to deal with it. I found the ModBus command list here. Note the 2-100% direct throttling (though after this it needs to be forwarded to the micros over wireless). I'm not an expert in ModBus nor power systems, but it looks pretty well thought out to my software engineer eyes.

That said, I have noticed some typos in the register map already from just looking at it for 10 min, which is a bit disappointing.

APSystems also has RS-485, I haven't found the official command list document but it is documented to some extent in the HomeAssistant community.

Random questions
Which company is more likely to have partnerships / expansion of product line to say AC coupled batteries or V2H?

 

[svetz]

Other microinverters brands I've heard of: Sol-Ark, Generac (previously Chilicon), SolarBridge, CyboEnergy, Blue Frog, HiQ,

First off, let me apologize as I pretty much only know about Enphase as it is what I have...

Enphase Pros
Enphase University, support, monitoring in at insane levels, API, battery integration, Excellent warranty and stable company, highly efficient, Sunlight (batteryless) backup, American based company.

Enphase Cons
For local API monitoring you need remote access to get a token (although the token can be good for up to 6 months).

EnphaseThoughts
Extremely well-engineered with things that aren't immediately obvious but desirable in most instances (e.g., the Enpower has a highly efficient neutral forming transformer built in (balances L1 and L2 loads automatically)). Enphase is a top-tier product, the price is higher than a string system if you compare inverter to inverter, but as microinverters include things like mppt, arc fault and RSD it's not as bad as most imagine. Factor in the life-cycle costs and they're generally a bargain.

After having them on the rooftop for a while I really really like microinverters:

Microinverters are superior in most ways to String inverters

• No single point of failure, If one fails you're only out a panel's input until you can get it replaced.
• More reliable with longer warranties (25 years) than String inverters (~5-12 years)
• Replacement shipping costs are less because they're small and light.
• Typically they have high conversion efficiencies
• They are silent
• They are corrosion resistant
• They have very low self-consumption (e.g., not running cooling fans)
• Per panel MPPTs have individual panel optimization and will almost always outperform a String system
• Shade on one panel doesn't cause loss from the rest of the panels
• They sit underneath the panel and don't take up additional space (this was important for me as I couldn't fit an inverter under the house due to the building codes, but the rooftop was well above the flood elevation).
• Lightweight, quick, and easy for one person to install, but one per panel means more time overall spent on it
• Typically you can get per-panel diagnostics which makes it much easier to diagnose problems
• Lower (240V in the U.S.) voltage off the roof.
• Don't need to do string designs or calculate string voltages/amperages (although see Microinverter Voltage Rise Design Issue)
• Easily expandible over time, they allow for different panels of different types and ages to work together without loss
• The RSD & Arc-Fault is built-in
• Having a smaller wattage range, the efficiency sweet spot is wider (e.g., pulling low amps on a big inverter can fall outside the sweet spot)

Dual Vs. Single Port Microinverters

...still has 4 port microinverters while AP systems is 2 port only. My understanding is that the cost is significantly less due to the simpler circuitry vs the 2-port ones -- 4 DC inputs but only 2 MPPTs instead of 4. So there is an additional dimension for value engineering.

I've heard that Enphase abandoned dual-input microinverters based on field experiences, not sure what they were though. @ncsolarelectric also recently reported some interesting discoveries about a Hoymiles Quad unit (although to be fair he was using it in a way not intended).

Hoymiles just ignores DIY and pretends it doesn't exist (haven't stumbled across any DIY info on their website)...has zero-export capability with control over RS-485...

@ncsolarelectric has done a lot of DIY with them, check his thread out.

Random questions
Which company is more likely to have partnerships / expansion of product line to say AC coupled batteries or V2H?

V2H
Enphase purchased ClipperCreek in 1/22 to get a jump start on V2H. But, I don't believe the EV standard in the US (SAE J1772 connector) can support bidirectional charging, so unless you have a tesla or a CSS style plug (e.g., Ford Lightning) it won't make a difference.

ESS
Enphase is the least likely to make it easy for 3rd party systems to connect to them since they want to offer a full solution suite. But, as they have so many installs it's the company most others will integrate with (e.g., Tesla, Outback, Victron, Sol-Ark, and Schneider all have solutions that work with Enphase).

Other
Enphase has a new device that controls loads (e.g., air conditioners, ev chargers, hot water tanks) based on rules (e.g., turn ev charger on when battery is > 90% and off when < 75%).

Hope that's of some help, sorry it's mostly Enphase stuff. I think Generac will be a big player in a few years (see Generac: Cheaper alternative to Enphase?). They have the capital to probably get ahead of the other companies and compete with Enphase in this category.

 

[zanydroid]

Thanks! Really appreciate the info. I had originally wanted to keep this focused on the smaller players but the first thing on the decision flow chart is probably Enphase or not.

Arguably, string inverter is more future proof / easier to swap out than microinverters, since you only have low complexity RSD devices up there, and if you need to go from solar-only to hybrid, or swap to a hybrid that has a CCS connector, you only need to make changes on the ground, you don't have to go lifting every panel and swapping out the MLPE. With the non-Enphase microinverters, esp the HM1500 with the simplified power architecture, the cost difference between string + RSD + extra company's support hardware to make RSD work (looked at TIGO last night) it's not that much.

V2H
Enphase purchased ClipperCreek in 1/22 to get a jump start on V2H. But, I don't believe the EV standard in the US (SAE J1772 connector) can support bidirectional charging, so unless you have a tesla or a CSS style plug (e.g., Ford Lightning) it won't make a difference.

J1772 needs a bigger architecture change than CCS on the card side to support bidirectional -- the power electronics need to be able to invert in both directions, standard transistor topology won't be enough.

CCS just to work is a software change -- basically just need to close the contactor and let the battery connect to the outside world. The V2H load is surely less stress than driving the traction motor to merge onto a highway. However, I'd love to hear from engineers on how much needs to be added, if any, to make it safe.

Companies I'm watching for CCS are dcbel, Emporia, and whoever makes the one for Ford. For hybrid inverters, basically to integrate CCS you get a CCS connector box that can communicate with the car and DC couple into the inverter.

ESS
Enphase is the least likely to make it easy for 3rd party systems to connect to them since they want to offer a full solution suite. But, as they have so many installs it's the company most others will integrate with (e.g., Tesla, Outback, Victron, Sol-Ark, and Schneider all have solutions that work with Enphase).

Yes, that makes sense. My assumption is that you can't really go wrong with Enphase in terms of AC-coupled future proofing. It's more a question of finances.

Other
Enphase has a new device that controls loads (e.g., air conditioners, ev chargers, hot water tanks) based on rules (e.g., turn ev charger on when battery is > 90% and off when < 75%).

Hmm, this is very cool, although just being able to connect/disconnect things isn't really rocket science from a electrical engineering perspective (the MLPE, inverters, etc are much more interesting), it's just a box of contactors. The interesting part is the software integration with the production and making it easy for the user. Also, you really want it to communicate with the thermostat and dump load controller in a smart way. Better to tell the thermostat to ramp up the heatpump inverter compressor by 500W than to brutishly turn the power on or off (and that's probably not great for an expensive heatpump either). Same for a hybrid water heater, ideally you would select between turning on/off the heat pump stage, and turning on the resistive stage. If you're on, say, NEM3.0, you probably want the heat pump stage on all the time, and if the math of surplus power works out, turn on the resistive stage as a massive dump load even though it's not the most efficient use of electricity.

Hope that's of some help, sorry it's mostly Enphase stuff. I think Generac will be a big player in a few years (see Generac: Cheaper alternative to Enphase?). They have the capital to probably get ahead of the other companies and compete with Enphase in this category.

Hmm, just going by market cap Generac is basically the same size as Hoymiles, which IPO'ed last year.

 

[svetz]

... string inverter is more future proof / easier to swap out than microinverters,

There's never a reason to swap out a microinverter (unless it died and onesie/twosies are easy) ... It's not like the panel will start producing more power where you'd need to upgrade it.

But the flip side with a string inverter is a nightmare. If you buy a 6kW inverter and later decide you need a 7 kW, you have to toss/eBay the 6 kW at a loss.

Even worse, if you need more panels 5 years down the road chances are good the new panels won't play nicely in old strings so you'll need to redo the strings as well. Whereas future Microinverters don't care, mix/match any brand of any power, shade or no shade. You can even put hoymiles in with Enphase (although that gives you two monitoring systems so I wouldn't recommend it).

J1772 needs a bigger architecture change than CCS...

My bad... I was thinking the CCS1 or 2 that has direct access to the battery. I do like the architecture of the charger in the car, but I'm still ambivalent about the inverter in the car/truck.

Hmm, just going by market cap Generac is basically the same size as Hoymiles, which IPO'ed last year.

I prefer products from countries with strong protection laws and like to support my country and Vets when I can (e.g., Sol-Arc). Hoymiles is headquartered in Hangzhou, Zhejiang, China. Generac is in Wisconsin. Enphase is in California. I hear you though, I too am always looking for ways to cut costs (e.g., DIY ESS for Enphase).

Nice chatting with you Z!

 

[zanydroid]

There's never a reason to swap out a microinverter (unless it died and onesie/twosies are easy) ... It's not like the panel will start producing more power where you'd need to upgrade it.

Yes, definitely micros are better for incremental change and for redundancy. But here are a few scenarios where there could be problems.

What if it doesn't AC couple the way you want, or you don't have access to change the grid profiles? For instance, if you have old microinverters that don't support one of the UL1741 ramp back profiles, then you only have off/on control for AC coupling. Similarly you could have an issue like, too much interference for wireless ramp back (which Hoymiles does) to work, and decide you want to switch to a different module that uses PLC.

If you have limited space to install solar panels, you can't just add more panels and microinverters to update to latest generation capabilities, you have to the inverters.

My bad... I was thinking the CCS1 or 2 that has direct access to the battery. I do like the architecture of the charger in the car, but I'm still ambivalent about the inverter in the car/truck.

CCS1 can probably be tricked into providing direct access, but to my knowledge there are more posts about knowledgeable people telling people interested in details of how to implement it, than there are posts saying that people have unlocked it. People have gotten Chademo to work though.

CCS2 has supported direct access, I'm not sure if it's required or just standardized. Volkswagen made some noise about providing access via a firmware update on ID.4, which was exciting back in April, but I'm now just assuming that is hot air.

I wrote about the Ford Charge Station Pro CCS2 connector here (links to my reddit thread). To my knowledge it's the first CCS2 solution. This box is interesting, it's a J1772 AC charger in the same box as a CCS2 DC backfeed. There's a RS485 control channel from this to their $5000 support hardware to make this all work for house backup.

240V V2L from North American Ioniq 5

Some hard facts in no particular order (some of which repeat zanydroid already said, sorry). I own an EV6, and have personally tested/measured/opened and observed all of these things except where indiciated. I have a USA car, so keep that in mind also: The EV6/Ioniq 5 use a bidirectional ICCU...

diysolarforum.com

Bidirectional J1772 would require a neutral-forming transformer on the shore side for V2H. Basically all the usual complexity of AC coupling. It's not that different from what you need to go off-grid with microinverters (IE Enphase Sunlight Backup has this transformer as part of that $5000 hardware add-on to enable it).

Communications isn't commonly implemented for J1772 even though it's in the standard, and without communications this inverter would need to be controlled via UL1741 or something like that.

My opinion is that, on the one hand, the car companies won't want to be in the business of implementing UL1741 and delegate that to a partner (which the CCS2 coupling is well suited for). On the other hand, they have an opportunity to compete on cost for storage and inverters, since that hardware is just sitting there in the car already. So there might well be a way for them to profit.

I prefer products from countries with strong protection laws and like to support my country and Vets when I can (e.g., Sol-Arc). Hoymiles is headquartered in Hangzhou, Zhejiang, China. Generac is in Wisconsin. Enphase is in California.

I think the big difference is in what the local market tech support will be and who takes what cut of the value-add chain. Generac and Enphase shareholders will take like a chunk of the manufacturer's profit cut that would otherwise go to a Chinese company. Sol-Ark I assume takes a similar cut as an integrator for Deye.

I think Enphase pulled some of the manufacturing back to North America which is nice.

 

[svetz]

What if ... you have old microinverters that don't support one of the UL1741 ramp back profiles

In this particular case you're grandfathered in; laws and requirements change all the time, fortunately, they only apply to stuff going forward. But sure, if you're doing for non-legal reasons then it's a pain. Same as when you redo the roof and need to take all the panels off. In general, for the normal cases, I still think microinverters are a lot more flexible and easier - but that's just my opinion.

I wrote about the Ford Charge Station Pro CCS2 connector here (links to my reddit thread). ...

Cool! Thanks!

 

[zanydroid]

In this particular case you're grandfathered in; laws and requirements change all the time, fortunately, they only apply to stuff going forward. But sure, if you're doing for non-legal reasons then it's a pain. Same as when you redo the roof and need to take all the panels off. In general, for the normal cases, I still think microinverters are a lot more flexible and easier - but that's just my opinion.

I was referring to AC coupling the microinverters to a grid-forming battery inverter/charge controller. The historical example would be, if you installed microinverters from before frequency shift was implemented and wanted to add storage. I don't know what kind of examples might come up in the future.

 

[svetz]

...The historical example would be, if you installed microinverters from before frequency shift was implemented and wanted to add storage...

Even then there will typically be options for Microinverters due to their popularity. For example, last time I checked neither Tesla Powerwall nor the Outback Radian required the use of frequency shifting to control microinverters. Instead, they can use an internal relay to disconnect the panels when the battery is over a certain percentage. Those AC coupling solutions are compatible with older microinverters.

Although (and I haven't checked) if I recall correctly, the frequency shifting frequencies fall into the old anti-islanding frequency ranges - so even the newer ones with frequency shifting might work safely (although they probably wouldn't be as efficient).

... I don't know what kind of examples might come up in the future.

Yep, always a risk with anything. I suppose that's one advantage a 10-year string inverter has over a 25-year microinverter; you can get the newer/better replacement sooner.

 

[Shimmy]

What if it doesn't AC couple the way you want, or you don't have access to change the grid profiles? For instance, if you have old microinverters that don't support one of the UL1741 ramp back profiles, then you only have off/on control for AC coupling. Similarly you could have an issue like, too much interference for wireless ramp back (which Hoymiles does) to work, and decide you want to switch to a different module that uses PLC.

Functionally I don't think that is a huge deal; you might replace a portion of your microinverters with an alternate that supports the newest bells and whistles to accommodate either your base load or some similar increment. The bigger, real challenge is needing/wanting to modify the AC wiring on the roof to be able to control individual circuits for one purpose or another.

For my system, I like that I have the option to upgrade my old M215's and up-size a few of my panels to maximize production.

 

[zanydroid]

The bigger, real challenge is needing/wanting to modify the AC wiring on the roof to be able to control individual circuits for one purpose or another.

Interesting, what are some of the use cases for this?

I thought about this in terms of AC coupling into different portions of a system, like some in grid tie only mode and others for backup power, but no solid ideas behind that.

Having cable going into conduits with some generous extra size for derating may help.

 

[Shimmy]

Interesting, what are some of the use cases for this?

If you want to curtail microinverters an AC relay is the easy approach. Or, if you want to add panels they might be in the wrong spot for where the branch circuit has excess capacity. All seems to end up pushing for re-wiring.

 

[zanydroid]

Ok, that was sort of similar to what I was thinking. For curtailing if you can modulate by modbus etc that seems more convenient.

 

[foggysail]

My system uses the Enphase equipment which involves more than just inverters. So far things are relatively fine. I did not want a string inverter in place of micros for a lot of reasons especially those related to DC.

Do today's string inverters provide safety shutdowns of panels to comply with the NEC or do they require other devices such as optimizers for that function?

 

[Shimmy]

My system uses the Enphase equipment which involves more than just inverters. So far things are relatively fine. I did not want a string inverter in place of micros for a lot of reasons especially those related to DC.

Do today's string inverters provide safety shutdowns of panels to comply with the NEC or do they require other devices such as optimizers for that function?

You do require either Tigo's or optimizers for compliance.

 

[foggysail]

You do require either Tigo's or optimizers for compliance.

Thanks---

 

[zanydroid]

Yes you need electronics up in the array. The requirement is shutdown to below 60V. Since most panels are Voc 60 that means at best you can maybe cut off part of a parallel string, if you are trying to reduce parts.

Then you need either a separate rapid shutdown box or have the string inverters talk natively to the RSDs. TIGO compatibility is baked into some

 

[foggysail]

And those extra ''electronics'' add to system cost maybe to a point where individual inverters are attractive. In life nothing is simple. Even if a bunch of micros appear attractive to strings one cannot overlook that even the micros need extra electronics. For Enphase it is their Combiners with internal Envoys and that adds a bunch of bucks to system cost. I have no knowledge of what other micros require for system support.

 

[foggysail]

I thought of another question on strings concerning shutdowns. The thought question was why string inverters could not just short circuit their input because panels have no problem working into direct shorts. Then I realized the answer. Yeah, that would work but only if the panels were connected in parallel. I am done---

 

[Danke]

Charging batteries and EV (future?). Having DC coming from panels may be more efficient.

 

[zanydroid]

If you only want Grid-Tie, I think microinverters are pretty good esp when you factor in the extra overhead of commissioning a system now that RSD is required.

If you want hybrid, DC definitely has advantages b/c the stuff on the roof is mostly dumb, you can send it into an appropriately selected AIO, etc, without having to coordinate microinverters & do AC coupling.

With regards to RSD vs micros. I think the RSD only systems you don't really need the extra control box (TIGO has partnership with a couple of string inverter companies to have the RSD initiator built into the inverter). But if you want the optimizer side you do need the control box for TIGO, if you search around you can see the drama of how the optimizers have a really high chance of doing the wrong thing if you don't have the control box, and TIGO can't even debug them unless you install a cloud control box and give them access to grab the telemetry and potentially apply a firmware update. That's a lot of technician time that you don't need to spend with microinverters.

One of the costs of micros is more custom fittings. I just ordered some micros and I got a bit of sticker shock on the proprietary trunk connectors. Enphase Q - $18 for portrait for single retail quantity (not sure how much installers can get for them). Hoymiles - $50 for one, which will let you hook up 2 or 4 panels depending on which micro.

By comparison Tigo RSD module is $30 for a single and $50 for a double, so it's in the same ballpark as the custom connectors. Their basic optimizer (which is buck only with a limited amps multiplication ability, not buck/boost with MPPT) is another $20 per panel. Almost like they designed the pricing that way.

The RSD basically just need a radio or PLC modulator connected to some NVRAM and a solid state relay, so there's a big reduction in complexity vs a microinverter which need MPPT, inverter, and UL1741 following logic.