• Have you tried out dark mode?! Scroll to the bottom of any page to find a sun or moon icon to turn dark mode on or off!

diy solar

diy solar

Microinverter power distribution basic overview

doc3g

New Member
Joined
Oct 9, 2024
Messages
76
Location
Texas
I can't seem to find the direct answer to this question but after watching lots of videos, the basic overview I get for power distribution looks like:

PV panel --> Microinverter --> double pole circuit breaker in main load center to back-feed the panel.

I recognize that Enphase and others have additional components for monitoring, forming microgrids, etc., but at the end of the day, the power is going from DC at the panels to AC 220 from the microinverters to feed the panel (assuming POCO agreements, etc.).

Is this the general idea of power distribution w/ micro inverters?
 
That is the general idea but there are a lot of little details.

At the lowest level, single phase AC loads have sine squared power profile. The power pushed to the single-phase AC grid from PV grid tied inverter is 120 'squirts' per second, for 60 Hz single phase.

To yield the maximum power from PV panels the load must be constant DC current at the given illumination current generated by panel. PV panel will lose efficiency if the single-phase AC ripple power makes it to the PV panel loading.

There must be an energy storage device to deliver the 120 Hz sine squared 'squirts' of power to AC grid while maintaining an even, constant power flow from PV panels. For a GT battery-less PV inverter this energy storage is in a bank of HVDC capacitors.

Above this, there are a lot of 'conditions' that determine when, how, and for how long a PV GT inverter interface is allowed to connect.
CAL Rule 21 are latest rules that tries to manage the conditions of large amounts of PV power pushed to grid to prevent unstable situations developing on power grid that can put the whole grid network at risk.

Hawaii is on the frontlines of having to face large percentage of distributed PV power injected into relatively small power grid. Each Hawaiian Island has at least its own power grid and some islands have multiple microgrids. This makes managing PV production very important,
 
Thanks for the explanation! I ask because the broad concept of microinverters seems like it could be a viable approach to scaling additional power into an off-grid system.

For example, if I have a 5KW string inverter w/ 6KW of solar coming in and on an average day I'm still consuming ~1kw more during my sun hours, then that extra KW would come from the battery (or a generator, etc).

If, for example, I find a great deal 3 x 405w panels and I wanted to add those to my current system I can either:
A) buy a bigger inverter to handle the additional 1200w
B) buy a second inverter to run in parallel
or, at least hypothetically
C) buy 2-3 used microinverters (Enphase, NEP, Hoymiles, etc) for ~ $50 each and run a breaker to my off-grid panel?
 
OP missed just getting a solar charge controller (appropriate for low voltage 3S strings) on that list. That would be most straightforward

Also in 2024 the chances of random AIO or hybrid you already own from a previous year’s purchase being AC couple capable are very low

In the future, it will be better
 
OP missed just getting a solar charge controller (appropriate for low voltage 3S strings) on that list. That would be most straightforward

Also in 2024 the chances of random AIO or hybrid you already own from a previous year’s purchase being AC couple capable are very low

In the future, it will be better

So where does the DC output from that inexpensive solar charge controller feed to?





Screenshot_575.png
 
Battery bus bar of course! A solar charge controller is by definition a charger that takes as input a DC solar string and as output charged a battery

Which AIO is it? Maybe it is an AC coupling capable unicorn.
 
Your microinverter diagram has more problems:
- 240V micro won’t activate off a 120v AIO
- 120V micros are a niche item and often have lying specs or marketed towards people doing something weird or don’t know what they’re doing
- Microinverters are never sanctioned by AIO manufacturers in the config you have, where it eventually reaches a load terminal of the AIO. They’re supposed to be fed through a relay controlled terminal on the AIO so it can pull an OHSHIT!!! disconnect to protect itself from the micro

Is the purpose of this thread to learn or to make something work? The latter is unlikely to happen outside of stacking a compatible AIO or SCC
 
Your microinverter diagram has more problems:
- 240V micro won’t activate off a 120v AIO
- 120V micros are a niche item and often have lying specs or marketed towards people doing something weird or don’t know what they’re doing
- Microinverters are never sanctioned by AIO manufacturers in the config you have, where it eventually reaches a load terminal of the AIO. They’re supposed to be fed through a relay controlled terminal on the AIO so it can pull an OHSHIT!!! disconnect to protect itself from the micro

Is the purpose of this thread to learn or to make something work? The latter is unlikely to happen outside of stacking a compatible AIO or SCC

- The current AIO I have sitting unopened in a box is a SunGoldPower SPH5048, so no, there is no AC coupling. It's still in the box because I have until Jan 31st to return if I realize I need something different, but I wanted to buy it at black Friday prices (saved $40).
The point of the thread is that I'm trying to learn how stuff works so that I can make informed decisions about what to buy. I see good deals on (apparently) good-quality used microinverters on eBay. When I weed through all the videos, it seems microinverters just change DC to AC and dump it into a panel. The caveat is that these tend to be people who have ponied up 10's of thousands of dollars and had a professional install including an agreement with POCO so that they can dump energy into the main load center and not have to worry about backfeeding.
I am trying to build a system that doesn't connect to my main panel (at least directly) and move over a few circuits at a time. The goal is to save a few bucks while learning about solar in anticipation of putting in a big system at our next home (~ 3-4 yrs).
My basic understanding of the microinverters is that they have an "anti-islanding" feature that means if the grid goes down, they sense the power loss and stop injecting power (with the exception of the new microgrid forming setups).
Looking at my diagram I see now how the 220v microinverters wouldn't work in that scenario, but for hypothetical purposes, assume that I had parallel AIO's, feeding each leg of the load center. Would it provide the feedback the 220 microinverter was looking for?

With the 110 grid tie microinverters, my understanding is that the peak voltage is slightly higher than the nominal peak voltage coming into the load center (per this video:
). So, again, for hypothetical purposes, assume a reputable company comes out tomorrow with a UL listed 110 grid tie with all the safety features such anti islanding. Is there a technical limitation preventing a connection to a load center as I sketched? Trying to understand where the magic device is that makes a turnkey enphase system work but that renders a single IQ7 essentially worthless for any purpose other than expanding an existing enphase system.
 
OK Cool.
The point of the thread is that I'm trying to learn how stuff works so that I can make informed decisions about what to buy. I see good deals on (apparently) good-quality used microinverters on eBay. When I weed through all the videos, it seems microinverters just change DC to AC and dump it into a panel.
For microinverters you would look for 1741SA ones, those will support frequency shift shutoff (and I guess ramp-down but i don't believe in the value of that vs just shutting off). Otherwise you are stuck with really clunky older ways to shut off.

And 1741SA microinverters have been on the market for a while.

Lower powered microinverters command a better savings IMO.

You are not eligible for 30% tax credit when used, so factor that in.

I like microinverters in a new build if you have a lot of small roof planes, or want to simplify away from DC wiring. Otherwise hybrid string inverter is better. Microinverters also have developed faster than solar charge controllers (more volume / money in it)
The caveat is that these tend to be people who have ponied up 10's of thousands of dollars and had a professional install including an agreement with POCO so that they can dump energy into the main load center and not have to worry about backfeeding.
In my state it's pretty straightforward to DIY grid tie agreement. $150 for interconnect, $200 for permits, $400 for plans, is what it cost me.

My basic understanding of the microinverters is that they have an "anti-islanding" feature that means if the grid goes down, they sense the power loss and stop injecting power (with the exception of the new microgrid forming setups).
All grid-tie inverters have anti-islanding, yes. In the absence of an interconnect agreement I'm not sure it's worth getting a grid-tie string inverter, as opposed to microinverters, to learn and play with. Since grid tie strings require rapid shutdown (if you want to be 100% code compliant), need DC wiring technique, and are less scalable in terms of reducing number of panels. There's minimum voltage for MPPT to activate and minimum number of series in string for the RSDs.

WRT trying a small system. I think an AIO is best since it forms a grid and does not have a chance to backfeed. You might consider a hybrid with grid connection air gapped; you can reuse this when you get an interconnection agreement. Like the SRNE and Growatt hybrids, which have 1741 certifications at a $1500-2000 price point.

However, these hybrids may not have a strict off-grid mode, which is why you want the air gap. Some hybrids do have strict off-grid mode which can be directly tied to grid.
Looking at my diagram I see now how the 220v microinverters wouldn't work in that scenario, but for hypothetical purposes, assume that I had parallel AIO's, feeding each leg of the load center. Would it provide the feedback the 220 microinverter was looking for?
In this case you need properly parallel stacked AIOs, since that ensures the voltage are synchronized in frequency and phase. This is just as important as AC coupling capability. Randomly shifting phases can blow up equipment plugged into that panel, and confuse or blow up the microinverters. AIO without AC coupling can blow up the AIO.

It's MUCH easier these days to just buy a split-phase inverter. Stacking 120V for split phase is only recommended in a pretty limited set of cases. So the preoccupation with 120V micro setup is kind of oddball

With the 110 grid tie microinverters, my understanding is that the peak voltage is slightly higher than the nominal peak voltage coming into the load center (per this video:
Same applies for 220V (also the system is 120/240 not 110/220)

110 grid tie with all the safety features such anti islanding. Is there a technical limitation preventing a connection to a load center as I sketched? Trying to understand where the magic device is that makes a turnkey enphase system work but that renders a single IQ7 essentially worthless for any purpose other than expanding an existing enphase system.
I'll probably not cover all the potential issues with 110, since it's pretty uncommon. Off hand, you can't power the loads on the other leg directly; you'll need an autotransformer to shift the power across. You double the wiring / have higher losses if you want a balanced system (so usually with 240V micro you just have a string that spans both legs, and that's it for a house. Easy. 120V micros means you have to plan it like a 3phase european or american commercial set up. Not rocket science, but also harder to find help with).

The magic device is the grid forming, AC coupling capable battery (like an IQ5), in combination with the Microgrid Interconnection Device, the System Controller. The System Controller has an ATS that will disconnect from grid before sending grid forming signals to the microinverters. It also has a neutral forming autotransformer to convert the 240V only microinverters to a 120/240V system. When on-grid, you use the utility transformer to do this (IE, you send 240V out your service ,to that transformer, then any neutral current you need is created here and sent back to your house)
 
The magic device is the grid forming, AC coupling capable battery (like an IQ5), in combination with the Microgrid Interconnection Device, the System Controller. The System Controller has an ATS that will disconnect from grid before sending grid forming signals to the microinverters. It also has a neutral forming autotransformer to convert the 240V only microinverters to a 120/240V system. When on-grid, you use the utility transformer to do this (IE, you send 240V out your service ,to that transformer, then any neutral current you need is created here and sent back to your house)

Thanks for the detailed response, especially the above, because that makes understanding the value of the whole enphase ecosystem more palatable.

Regarding the split phase vs paralleled AIO's, is it that stacking a pair of SPH4058's is going to cause headaches? One of the reasons I picked this unit was because of the ability to expand it to paralleled with a second unit in the future. It seems the starting point for split phase is about 2x the cost. I also only have ~5000w of solar currently, so something like the 10kw splitphase version of my AIO wouldn't be useful unless I also spend another $1200 on panels. Of course, if paralleling isn't truly feasible, then I guess the only option would be to go with a split phase AIO from the beginning.
 
Can you link the inverter?

Often a split-phase will have two single phase inside it, so 2x is a fair starting point.

You also want to see what the standby consumption power of a 2x 120V stack vs 1x 120/240 inverter.

Other value besides MPPT capacity
  • You will save wall space and wiring with a split phase unit
  • There are more ways to fat finger and mess stuff up with stacking
  • There is more stuff you have to look up / learn to build split phase with stacking
  • 10kW unit has more surge if you want to start some inductive loads
  • If you get 2 server rack batteries (for 10kWh total or higher), you can leverage the 10kW-AC output of a 10kW unit
 
If you want to experiment with microinverter AC coupling, I would recommend getting one of those $1500-ish split-phase inverters with AC coupling. 6000XP has it in some modes. You can check on the SRNE threads to see if those have it
 
Can you link the inverter?

Often a split-phase will have two single phase inside it, so 2x is a fair starting point.

You also want to see what the standby consumption power of a 2x 120V stack vs 1x 120/240 inverter.

Other value besides MPPT capacity
  • You will save wall space and wiring with a split phase unit
  • There are more ways to fat finger and mess stuff up with stacking
  • There is more stuff you have to look up / learn to build split phase with stacking
  • 10kW unit has more surge if you want to start some inductive loads
  • If you get 2 server rack batteries (for 10kWh total or higher), you can leverage the 10kW-AC output of a 10kW unit

This is the inverter.
- https://sungoldpower.com/products/5000w-48v-solar-charger-inverter

One of my other interests in microinverters is because of shading. I've got a short run of yard where I don't expect any shading, so it seems like a string inverter will work just fine there. However, I'd also like to put some panels on the roof eventually, but, due to the chimney and my neighbor's massive loblolly pines, there are lots of periods of shading. I realize I could put optimizers on those panels but, when I get down to comparing optimizers vs used microinverters to deal with the shading, it seems that the microinverters are a better value per panel.

Which brings me back around to wonder if I should just start with a microinverter system and scale up. I realize that means I have to get permits and a POCO and inspections, but the original point of trying to avoid all that was due to cost efficiency. If, however, I'm considering spending closer to $1500 for an inverter, then perhaps I'm in the ballpark of a DIY install grid-tied enphase or similar system.

Pictures for reference of east facing roof Vs. the south-facing fenceline where I plan to install the ground mount (for size context, the larger panel is a Sunpower 405W, and the smaller panel is an Eco-Worth 200W)

east roof.png
south fence.png
 
One of my other interests in microinverters is because of shading. I've got a short run of yard where I don't expect any shading, so it seems like a string inverter will work just fine there. However, I'd also like to put some panels on the roof eventually, but, due to the chimney and my neighbor's massive loblolly pines, there are lots of periods of shading. I realize I could put optimizers on those panels but, when I get down to comparing optimizers vs used microinverters to deal with the shading, it seems that the microinverters are a better value per panel.

Which brings me back around to wonder if I should just start with a microinverter system and scale up. I realize that means I have to get permits and a POCO and inspections, but the original point of trying to avoid all that was due to cost efficiency. If, however, I'm considering spending closer to $1500 for an inverter, then perhaps I'm in the ballpark of a DIY install grid-tied enphase or similar system.
You are still missing a few of the cost items for AC coupling.

To make the Enphase microinverters work off-grid, you will need at least a $1000-1500 AIO or hybrid and a properly sized battery. Cross-brand AC coupling "by the book" also requires that the grid forming inverter be capable of absorbing the full AC output of the microinverters in case load drops. Within Enphase you can get away with fewer batteries (on IQ8 and higher IIRC) because the IQ8s have a mode where they can automatically ramp down according to load (this capability is the same as used for Sunlight Backup [which is a dumb feature; it's more of a bonus]). But standard microinverters are incapable of ramping down automatically or quickly.

Thus, one 5kWh battery, capable of 1C charging -> ~4.5 kW probably after losses, is needed per 4.5 kW-AC worth of microinverters. That's something like $1000 per 15 microinverters. You might also clip your DC coupled solar if your microinverters are going full-out, and the batteries cannot accept the charge current.

I'm not a fan of optimizers because they are poorly documented. Tigo optimizers are the only cross-brand optimizers on the market right now, and they can only bridge something like 20-30% worth of voltage or current difference. They also don't have individual MPPTs.

Another way to deal with shading is solar panel cell layout + regular bypass diodes. For instance half-cut and shingled panels have more internal parallelism to route around shade within a single panel, so there are more shadow geometries where the production is not a complete disaster. It's still going to be really bad though, and the geometry + electrical design is more expensive of your time to learn to do manually (or get string design software for), than it is to just get microinverters.

Pictures for reference of east facing roof Vs. the south-facing fenceline where I plan to install the ground mount (for size context, the larger panel is a Sunpower 405W, and the smaller panel is an Eco-Worth 200W)
200W panels are more for learning/portable use cases IMO. Though you can probably 2P or 2S them into a single microinverter or optimizer. Look carefully at the voltage and current. After you get the 400W panels on the roof, they are less annoying to deal with than a bunch of smaller ones (which require more wires, clamps, rails)
 
You are still missing a few of the cost items for AC coupling.

To make the Enphase microinverters work off-grid, you will need at least a $1000-1500 AIO or hybrid and a properly sized battery. Cross-brand AC coupling "by the book" also requires that the grid forming inverter be capable of absorbing the full AC output of the microinverters in case load drops. Within Enphase you can get away with fewer batteries (on IQ8 and higher IIRC) because the IQ8s have a mode where they can automatically ramp down according to load (this capability is the same as used for Sunlight Backup [which is a dumb feature; it's more of a bonus]). But standard microinverters are incapable of ramping down automatically or quickly.

200W panels are more for learning/portable use cases IMO. Though you can probably 2P or 2S them into a single microinverter or optimizer. Look carefully at the voltage and current. After you get the 400W panels on the roof, they are less annoying to deal with than a bunch of smaller ones (which require more wires, clamps, rails)

Yes, the 200W panels were exactly for that, learning (bought them after a thread about them on ebay was posted, and, after they refunded me 30% for trying to send them back, came out to $67 each for 200W bifacial panels). I'm planning on putting them on my shed for a 12v system to do lights.

So, as impressive as Enphase looks, I've realized that using their stuff requires lots of pricey magic boxes like the IQ system controller. However, now that I've started down this microinverter rabbit hole, grid tieing w/ NEP or Hoymiles. I'd appreciate your input if I'm missing something, but, for example, to power my 10 405w panels, I could get 5 NEP BDM-800s (~ $1k). A couple of junction boxes + an inspection and I could grid tie, which, from my perspective, would maximize my ability to use all of my current solar. Up until now, I have been planning to avoid grid tieing because of the complexity and cost of having numerous components fit together in a manner that would pass inspection. However, it seems that with at least the NEP setup, their ecosystem seems to function pretty much as I thought originally, where solar DC is converted to AC 240 and fed into the main (grid-powered) existing load center.
 
So, as impressive as Enphase looks, I've realized that using their stuff requires lots of pricey magic boxes like the IQ system controller.
Microinverters will generate as soon as you plug into grid.

I believe to get full warranty you just need combiner. (combiner is also needed as admin interface to push settings, I don't think there are any microinverters that are manageable without some kind of box). And for on-grid self-consumption on batteries you also just need combiner.

System Controller is for backup power. An IQ5P is not a terrible price for an ESS if you are under NEM3 and need to comply to latest code.

The difference between each of the Enphase, NEP, and Hoymiles is probably what you need for monitoring and what you need for a minimum manufacturer compliant installation / warranty. The advantage of Enphase is probably surviving trade war better. It's possible a POCO or AHJ will request that you show them the grid settings, and you can only do that if you have monitoring.

For HoyMiles (which I have) monitoring is optional, though it's not a terribly priced box ($200-250), and there is something called OpenDTU that can do it with a $15 OpenDTU hardware from AliExpress.

With one of my Hoymiles setups (I have two) I actually have two strings feeding into a quad breaker at the opposite end of my load center (to avail of 120% rule). You can do 30A + 30A since Hoymiles allows up to 30A branch circuits.

I'd appreciate your input if I'm missing something, but, for example, to power my 10 405w panels, I could get 5 NEP BDM-800s (~ $1k). A couple of junction boxes + an inspection and I could grid tie, which, from my perspective, would maximize my ability to use all of my current solar. Up until now, I have been planning to avoid grid tieing because of the complexity and cost of having numerous components fit together in a manner that would pass inspection. However, it seems that with at least the NEP setup, their ecosystem seems to function pretty much as I thought originally, where solar DC is converted to AC 240 and fed into the main (grid-powered) existing load center.
Yes, the price sounds right and costs the same or less than just one of the big ticket components in the AC coupled off grid system.

You can also look at Hoymiles HMS-1600 and HMS-800 (you can even consider HMS-1600 and leave some ports open for expansion or something). These can go straight into backfeed breaker without a combiner box (depends on AHJ/POCO requirements, I'm allowed to do this). 5x BDM800 and 5x HMS800 will slightly go over a 20A branch circuit, so you will have to pull #10 for a 30A circuit.

With microinverters you can use any wiring method including Romex, and you don't have to learn how to hook up rapid shutdown etc.
 
I believe to get full warranty you just need combiner. (combiner is also needed as admin interface to push settings, I don't think there are any microinverters that are manageable without some kind of box). And for on-grid self-consumption on batteries you also just need combiner.
Screenshot_585.png

This is the wire diagram from SS for the BDM-800.
- The combiner box is basically a waterproof box with some components on din rails? (surge arrester, etc).
It seems the combiner box in the above setup would need less components than if I were trying to wire 4 strings into an AIO?

System Controller is for backup power. An IQ5P is not a terrible price for an ESS if you are under NEM3 and need to comply to latest code.
I got 3 server rack batteries during black Friday for $88 Kwh! Though, from my standpoint, battery capacity only matters if I get to the point where I'm producing more than I can consume, which I won't do with my current # of panels.

The difference between each of the Enphase, NEP, and Hoymiles is probably what you need for monitoring and what you need for a minimum manufacturer compliant installation / warranty. The advantage of Enphase is probably surviving trade war better. It's possible a POCO or AHJ will request that you show them the grid settings, and you can only do that if you have monitoring.
I was reading this thread, and it seems to indicate that Hoymiles, at least for the time, is a solid bet
https://pv-magazine-usa.com/2024/08...facturer-financial-stability-ranking-updated/

For HoyMiles (which I have) monitoring is optional, though it's not a terribly priced box ($200-250), and there is something called OpenDTU that can do it with a $15 OpenDTU hardware from AliExpress.
- I've seen these on eBay for good prices, and even at full retail, they are reasonably priced compared to the $2500 for an Enphase IQ3

With one of my Hoymiles setups (I have two) I actually have two strings feeding into a quad breaker at the opposite end of my load center (to avail of 120% rule). You can do 30A + 30A since Hoymiles allows up to 30A branch circuits.
- OK, you lost me on that one, I have to go look up the 120% rule.

You can also look at Hoymiles HMS-1600 and HMS-800 (you can even consider HMS-1600 and leave some ports open for expansion or something). These can go straight into backfeed breaker without a combiner box (depends on AHJ/POCO requirements, I'm allowed to do this). 5x BDM800 and 5x HMS800 will slightly go over a 20A branch circuit, so you will have to pull #10 for a 30A circuit.

- I was watching a battle thread today on the FB group about DC needing to be in metal conduit inside of a structure and was thinking, "it would be nice if I could just run Romex"

With microinverters you can use any wiring method including Romex, and you don't have to learn how to hook up rapid shutdown etc.
 
This is the wire diagram from SS for the BDM-800.
- The combiner box is basically a waterproof box with some components on din rails? (surge arrester, etc).
It seems the combiner box in the above setup would need less components than if I were trying to wire 4 strings into an AIO?
For microinverters in the U.S. a combiner box is just a regular load center with standard breakers

If wiring 4 strings into an AIO you may need a lot of safety disconnects and conduit / box fittings suitable for DC.

I got 3 server rack batteries during black Friday for $88 Kwh! Though, from my standpoint, battery capacity only matters if I get to the point where I'm producing more than I can consume, which I won't do with my current # of panels.
OK, what I said above was setting up batteries as a dump load for AC coupled output, and most / all AC coupled inverters are 48V

> 25% for IQ3

Is that system Controller 3? That is a much more complex/feature rich thing than a DTU-S. The IQ combiner is closer in capability, but not really, because it also adds CTs and plug on load center busbar. You put 20A standard breakers onto this, into which each Enphase branch string lands

The DTU-S in some markets can accept external CTs but I believe it is less capable than Enphase. And it requires like $200 of extra stuff

Hoymiles is way cheaper on the microinverter side though

If you like the option to program your own stuff, Hoymiles serial API is well documented. That is why OpenDTU exists

> 120% rule

This is NEC 705.12. It is the most common backfeed limit calculation used for grid tie
 
Last edited:
For microinverters in the U.S. a combiner box is just a regular load center with standard breakers

If wiring 4 strings into an AIO you may need a lot of safety disconnects and conduit / box fittings suitable for DC.


OK, what I said above was setting up batteries as a dump load for AC coupled output, and most / all AC coupled inverters are 48V

> 25% for IQ3

Is that system Controller 3? That is a much more complex/feature rich thing than a DTU-S. The IQ combiner is closer in capability, but not really, because it also adds CTs and plug on load center busbar. You put 20A standard breakers onto this, into which each Enphase branch string lands

The DTU-S in some markets can accept external CTs but I believe it is less capable than Enphase. And it requires like $200 of extra stuff

Hoymiles is way cheaper on the microinverter side though

If you like the option to program your own stuff, Hoymiles serial API is well documented. That is why OpenDTU exists

> 120% rule

This is NEC 705.12. It is the most common backfeed limit calculation used for grid tie

Thanks for the info!

So, I'm down to deciding 1 of 2 paths at this point

1. Microinverters (grid tied)
- likely Hoymiles or NEP
- Dual/Quad input units get the price down to under $100 per panel depending on the deal and a full price DTU adds another couple hundred bucks
- The upside is that the installation would be simpler, and I would be connected to the grid, so I could parallel the grid and, therefore, utilize most of the solar energy I produce.
- The downside is that permitting and interconnect agreement is still a big black hole as far as cost. I see people saying they were able to do it for $300-$400 while others are at $1k plus

2. String inverter and not grid-tied
- Looks like the smartest approach would be to send back my single phase AIO and purchase a split phase, so that drops me into the same $1000 bracket as the microinverters
- Fewer unknowns in the price workup, but it is difficult to say if this ends up being more or less than the microinverter setup
- Overall, this options seems like it would be more interesting to build and more flexible, but also more room for errors

To move forward I need to go to the city and see what all they would require for a permit. I would prefer to find a service that could outline everything and assemble it into a packet for submission.
 
The big items are:
- interconnection paperwork fee
- permit fee
- whether you need engineering stamps

To hit $300 you would need to draw your own plans and/or use SolarApp+ equivalent for your state to avoid having to submit as much stuff

I would say $600-1000 is most likely.

Note that visible solar panels in some places can result in being forced to document to POCO that they are off grid or had permits pulled.

To move forward I need to go to the city and see what all they would require for a permit. I would prefer to find a service that could outline everything and assemble it into a packet for submission.
A plan service will do it submission ready, you just need to sign the pages. It covers electrical and panel layout, all codes. The panel layout needs to be aware of your local roof access path rules, etc.

It will also save you some work in learning the installation rules for the roof racking you pick. My service was UniRac and they put rail cut and rail feet positioning info. Granted I still looked up the install instructions in case they were wrong (no mistakes other than me giving them wrong dimensions for one part of the roof, so I was over by one solar panel. I have a spare now which is a good idea to have anyway.
 
FWIW the soft costs for grid tie are fixed so you should probably just scale it for your full future needs. You have to pay the soft costs again every time (well you might be able to modify the plans without paying service again, but that’s only half of them)

Also if you buy panels again later for an expansion they will be different panels, which may look ugly. I have two different kinds of panels on one side of my house, kind of cringe
 
Everything is going to be in the back yard, so not visible to the man. I was really excited about this ground mount/microinverter/grid-tied prospect but I'm running into a huge brick wall for the ground mount. I need to talk to the city but if I have to have a ground mount w/ an engineering stamp then the install costs blow up.

I also need to look into this engineering stamp thing, because honestly, there has to be a freelance engineer who can do this for a nominal fee as long as you are making a reasonable structure.
 
Everything is going to be in the back yard, so not visible to the man. I was really excited about this ground mount/microinverter/grid-tied prospect but I'm running into a huge brick wall for the ground mount. I need to talk to the city but if I have to have a ground mount w/ an engineering stamp then the install costs blow up.

So I don’t know much about ground mount permitting. There is a weird situation where
DIYers super love them but some contractors I’ve randomly talked to on Reddit say they add a lot of permitting and site planning cost compared to roof mounts. You might search here for folks that did permitted ground mounts. Also the extra red tape PITA of ground mounts may point to just doing off grid

Roof mounts require you to go on a roof and make some holes/share fate with your roofing situation, but they also inherit wood construction rules (IE, your house is properly bolted to a foundation, and it has properties from standard framing, and the weight of your whole house is helping to hold down the panels) / are super streamlined and optimized from an engineering and review POV. Roof mounts also have a higher chance of random complaints and scrutiny (but it’s easier to get them 100% kosher so in my mind it cancels out).

The man can see stuff from drones pretty easily, but it’s probably less likely for you to hit a drive by snitching situation for a ground mounts vs a roof mount

I also need to look into this engineering stamp thing, because honestly, there has to be a freelance engineer who can do this for a nominal fee as long as you are making a reasonable structure.
It’s $300 for a roof stamp when I did it, the engineering might be more streamlined because you can use standard wood building manual for this (and in my town they actually don’t care about stamps if the framing is above a certain strength, and accept simply the pre-engineered tables from the racking manual).

Remember the engineer is going to cover their asses since they have professional liability
 

diy solar

diy solar
Back
Top