Open Source Microinverter: Epilogue?

That’s part of it.

Anti-islanding protects the folks working on the high voltage lines.

If you backfeed a transformer that normaly steps down 2400/7200/12470 to 240 Volts, it will
put that 2400/7200/12470 on the line its primary side is tied to. A lineman performing a repair
isn’t likely aware the line is live. (although proper training says treat ALL lines as live)

@Bill.Thomson Do you know if the Enphase IQ microconverters handler the anti-islanding sensing and regulation, or is that handled by another Enphase component downstream? I believe the IQ microvonverters can be signaled to go offline, but I imagined a controller downstream was managing and monitoring the grid status. It seems to me the monitoring of grid status and preventing backfeed when the grid is down ought to be a mature technology since before microconverters, you had converters for a series of panels that would feed into the grid. We installed a Sun Power system in 2013 and I remember that back-feed safety issue. I there there may be two issues: 1) backfeed, and 2) emergency stop (for firefighters). The emergency stop is something that was recently introduced in the last couple of years and I’m not sure all states in the United States have implemented that requirement into their code.

What I’m seeing with Enphase is you can have option A) generate with microconverters and they only work if connected to an active grid, B) if you want to have the ability to draw on the microconverters when the grid is down, you have to pay approximately $3,000 to isolate your “island”.

ALL of the inverters Enphase makes, and AFAIK, all they’ve ever made have anti-islanding
capability. So the answer to your question about the IQ8 and anti-islanding, is yes.

The IQ8 is a special animal compared to the rest of the Enphase product line.
It can be throttled, and it can support a micro grid.

More info about the IQ8 here:

The Sunpower system you desribed sounds like it’s a string inverter system.
i.e. one inverter fed by one or more series-connected “strings.” The downside to that scenario
as far as firefighters are concerned, is the high voltage DC produced by the PV modules regardless
of the state the grid is in. That DC voltage can range from 300 to 1000 volts, so I can understand
why they’d be concerned about it. WRT a string inverter system, an emergency stop should disconnect the DC side of the inverter as well as the AC side.

Given the requirement is part of the National Electrical Code, I’d say it’s a safe bet to say all 50
states have that requirement in place. Although it’s true the NEC isn’t law, AIUI, the vast majority
of AHJs (Authority Having Jurisdiction) follow the NEC. (The AHJ is the lowest level that has control
over the grid at the local level. It might be a city or town, an eletrical co-op, a corporate concern
or other entity)

Because of the anti-islanding requirement, all of the the microinverter manufacturers that sell in the
US market legally, produce products that have anti-islanding/rapid-shutdown capability.

The El Cheapo micros from China have historically NOT had anti-islanding capability.
However, it seems that the latest offerings from China do. Whether or not it actually works,
I cannot say.

That sounds like the Ephase Ensemble.
I’ve never looked at it, so I don’t know a great deal about it. But I can definitely believe the 3k-buck
price tag. I wouldn’t be a bit surprised if it was more that that.

Point of clarification for anyone reading this thread and NEC and law. The National Electric Code (reg. trademark by National Fire Protective Association (“NFPA”)) (“NEC”) is a model code, each state decides which version and to what extent it may adopt a model code. Just need to be clear that the NEC, but itself, do not have the force of law, it is a template for states to adopt and I’m guessing most do without much alteration, though it appears various version are currently in force in various states. I think the island safety standard is also known as Rule 21, however, I have to give up my privacy to the NFPA for “free” access to the code – not something I choose to do because I abhor organizations that set standards and then try to encourage governments to adopt the standards so they have force of law, but you have to purchase their editions. (It used to be that in order to know what the code was, you had to purchase the code from the organization, but that recently changed in California.) See Updated 2020 NEC Code Regulations by State | HomElectrical.com

Which is exactly what I was referring to when I said

It isn’t just the state that has a say in electrical energy supply matters.
Hence my mention of the Authority Having Jurisdiction (AHJ)

A state may have no law on their books prohibiting something, but an AHJ could have rules/regs
tighter than those of the state, effectively prohibiting that same “something” the state allows.

For anyone outside the USA reading this, do be aware of your local regulations.

In the UK, the “IEE Regs”, otherwise British Standard 7671 is invoked by Part 5 of Building Regulations, and those are law.

Anti-islanding is a requirement before you can legally connect any private generation to the grid.

On Friday, Jun 24, 3:32 PM PDT, I sent an email to Ken Boak. @TrystanLea had kindly provided me Ken’s email. I have not heard back, as yet.

Here is a quote ($18,400) from a company based in California selling Enphase products. I presume they have the necessary credentials to allow a customer’s self-built system to become activated assuming your purchase the “kit” and pay for the extra services. I had specified: 1) must not have batteries, 2) must work when grid is down (the predicted aftermath of Cascadia), must be something I can repair if damaged (by earthquake) and maintain.

This proposal looked tempting at first encounter, but the more I considered the design and started looking into the components, the more I became concerned about black box technology. I’ve witnessed how exploiters of some LoRa patents have positioned themselves in the market place to sell reasonably priced sensors using LoRa technology, but in order for sensors to work, you have to go through their cloud service. Looks like Enphase is doing the same thing – positioning itself between the mechanical part and in effect “renting” you the right to access and use your components as designed. John Deere tractors… ho ho, Russia commandeered some John Deere equipment in Ukraine and now they cannot operate it because John Deere manufactured the equipment to tie into their cloud.

So back to this topic: open source microconverter, it looks like if it could be approved for use with a grid, there are substantial hurdles to be overcome if you are to integrate such a home-built microconverter into an established electrical grid. It’s becoming tempting to just skip the goal of grid tie-in because of the barriers to entry.

I still would be interested in hearing from Ken in case there were some technical challenges he felt were insurmountable.

This has been a very engaging discussion. I’m reflecting on this and thinking: there needs to be an open source generic device that 1) assures synchronization to the specifications of the grid it attaches to and 2) acts as the shutoff to a grid.

What feeds into such a device, solar, wind, whatever, doesn’t matter and can be left to another component. Assume @TrystanLea and Ken’s project that caught me eye were working as desired, they still would have the barrier of legitimately interfacing into a grid. Other projects such as wind power generation face the same issue, once they’ve created DC, how can they export that into AC in a fashion that is compatible with the grid specifications and safety requirements. It would not be in a for-profit entity’s best interest to make a generic tie-in device, better to bury such design into the product in such as way that they only function with their product; this way they have a closed market with no competitors.

It’s worth noting that the discussed design above was only ever a very early stage design, grid synchronisation was only ever discussed as a wish list item in one of Ken’s blog posts and not something implemented in the design. The first goal would have been to complete a version that would have worked in an off grid context.

Talking of anti-islanding though, here’s a hackaday grid tie inverter project that specifically discusses this with an extended discussion in the comment section about safety Build Your Own Grid Tie Inverter | Hackaday

It’s probably worth looking around for other inverter projects as well as there may be some more mature projects around by now?

Care as ever is going to be really important with any mains voltage inverter design given the potentially lethal voltages involved and ultimately a good understanding of the relevant regulations involved.

It’s great that you are interested in this @JohnPoole and don’t wish to discourage you, but I would emphasise that our work in the above project was very much an early stage design and a long way from anything that ready for general use I would have thought? There’s a ton of knowledge, research, experience and as many are discussing here regulation compliance! built into commercially available inverters.

If your interested in potentially lower cost DIY solar systems using off the shelf commercial components this website from a UK supplier may have some inspiration https://www.bimblesolar.com/offgrid/complete-packages there’s probably a US company doing similar?

There are quite a few open source solar related projects e.g:

I visited the Bimble Solar site and in looking at the kits, I realized what intrigues me is the approach of a microinverter in that you can have different panels in the array and each panel operates independently of the other in terms of output. Another way of stating it is if panel A is shaded and only putting out 20% of what it could with direct sunlight, panel B which is not shaded still contributes 100% of its output to the “island” grid. My understanding is that in a DC series, the entire system is only as strong as its weakest links, so A in series with B will only contribute a net of A*.2 + B*.2, whereas as in parallel through the use of microconverters, you have A*.2 + B*1.0. So microconverts let you have an array with different shading topologies without the entire system being limited to its weakest performer.

That parallel contribution feature is what lead me on the hunt for microconverter alternatives and to yours and Ken’s project. I visited The Back Shed side which is based out of Australia and they are very much focused on the DC series, one of its prominent contributors confessed he had not considered microconverters until I posted.

I have a situation where there a several planes (different rooftops) for panels, but each has a different shading topology, so using a DC series system would mean to combine units within each unique topology and then have an inverter dedicated to that group. The microinverter on each panel approach would allow

I’m trying to assess the state of development in open source vis-a-vis what Enphase IQ8 offers. I believe there may be money either from government or well-funded charities that could fund further development were it open source. I really do appreciate the time you and others are giving me here.

1 Like

Yes I agree, it would be great to see a project pursue such a goal. Im sure if someone designed such an open source product there would be a lot of demand for it!

I have a self-built system and had no problems getting it activated. That was in 2013.

Last year, I helped a friend activate their Enphase system. No problems there, either.
Both systems were activated at zero cost.

Enphase stopped charging for access to Enlighten several years ago. When we activated my friend’s
system, there were no costs involved. We simply went to the Enphase website and activated his system.

An Enphase system will produce energy without an Envoy. (I operated mine sans an Envoy for 3 years. Searching for an alternative led me to OpenEnergyMonitor) Of course there was no module level monitoring, but I couldn’t see shelling out the 580 bucks Enphase wanted for an Envoy at the time.
As with many things, there are exceptions. If an Enphase IQ Battery or IQ System Controller is added to
a system, the Envoy becomes mandatory.

I’d say that’s probably the reason why it hasn’t been done.

In a manner of speaking, there is money from the government in the form of Federal tax credits.
Some states offer tax credits too.

My 9-year-old PV system is based upon Enecsys microinverters. The Enecsys company went bust shortly after my install so when problems occurred with some of the electronics, I reverse engineered the design so I could carry out some repairs. The result is attached as a pdf file which is not copyright as it is not based on the original circuit diagrams. It is not perfect, there are errors but it helps a great deal when trying to understand how these thing work.
full_schematic_11_12_2021.pdf (606.2 KB)

@BrianD Your PDF: wow! That is going to take some time and effort to digest and it very tempting to try and understand the entire document. What happened with Enecsys looks to be a great case-study, so I did some searching and found Q&A: Clever monitoring for Enecsys inverter owners left in the dark . Did you ever collaborate with René Sandmann? Seems like you and René Sandmann were is the same position, but working in parallel.

@JohnPoole I mostly worked alone although a chap in the USA collaborated with me, he worked on the software while I stuck to the hardware. Lots of other people tried to do something with the old inverters but they mostly fell by the wayside as it’s not easy.

I have a RPI running Emoncms to capture data and another RPI which runs a Pylon script to upload real-time data to a site in Australia. An Arduino diverts some surplus power to water heating and OpenEVSE feeds the remainder to an EV.

If you have questions feel free to ask as I have a lot of accumulated info’ as you can probably imagine.

Our PV system here in the UK uses Enphase microinverters and an Envoy. It was installed in 2014 or 2015 (don’t remember which) and went live on 20 May 2015.

The anti-islanding is done by the microinverters themselves - the SWA cable lay in a coil on the floor for several months and the installer told me I’d be completely safe to put the end in my mouth! I didn’t but I did check there was no voltage on the cable.

The G98 3.68 kW export limit is I think enforced by the Envoy, since it’s the only thing (apart from the Enphase factory) that knows what the total power being generated is. So I believe an Envoy is a legal requirement for a grid-connected system at least in this country.

On June 27, 2022, I ordered an Enphase Q8+ microinverter through Amazon Prime for $164.99. The component shipped from Miami, FL, and arrived July 2nd via United States Postal Service (“USPS”) in a USPS box.

I wanted to dismantel the microinverter to see what was on the circuit board and learn about what components were involved. When I learned that the circuit board will be encased with “potting” material, a substance which the circuit board is submerged into, or is injected into a mold to completely encase the circuit board, which can be made of a variety of materials, many of which are epoxy based, I learned I was looking at a significant effort to clean off the material – like many hours, if at all possible. I researched various ways to remove the potting material and because of the variety of materials used, many were try-it-and-see-if-works. I even found there is a U.S. government specification for potting material to make it as difficult as possible to reverse engineer the circuitry, so I concluded potting material not only serves a purpose of keeping things from movement and being a heat sink, it also thwarts prying eyes. I tried searching for a business which could removed the material for me and did not find such a service despite one form participant suggesting Google could provide a listing of such services.

I contacted Enphase requesting their Material Data Safety Sheet (“MDSS”) which in the United States is required to be given on request for materials that may pose hazards. While Enphase has an MDSS for their battery product available on their website, they do not for the Enphase Q8+. So I made a request to Enphase to supply me an MDSS, and I received an email from Surya N with an attached PDF: E341165-20210317-IQ8PV series-R21-CofC.pdf. The PDF is nine sheet document entitled “CERTIFICATE OF COMPLIANCE” which has nothing to do with material safety, but concerns compliance with various standards for safety promulgated by Underwriters Laboratories, IEEE and other organizations. I was hoping for something which would indicated what the potting material is made of, that way I could use the appropriate solvent.

I reached out to a couple of chemical analysis labs, one of which was agricultural based and proved to be a not productive lead, the other looked to fit the bill and they used a calendaring system to schedule a discussion. I submitted my issue and obtained a date and was never contacted.

In the meantime, I started searching around for open source attempts at microinverters and came upon Texas Instruments development kit. It is the discovery of this development kit that caused me to realize this is precisely what I was searching for – and it was all designed to encourage development.

https://www.ti.com/tool/TIEVM-HV-1PH-DCAC?keyMatch=DEVELOPMENT%20KIT%20DCAC

TIEVM-HV-1PH-DCAC
Single phase inverter development kit with voltage source and grid connected modes

$848 - $908 USD

Description: This reference design implements single phase inverter (DC-AC) control using the C2000™ F2837xD and F28004x microcontrollers. Design supports two modes of operation for the inverter. First is voltage source mode using an output LC filter, this control mode is typically used in Uninterrupted Power supplies. Second is Grid connected mode with an output LCL Filter, which is typically used in solar inverters. Firmware for the design is supported under powerSUITE framework which enables adaptation using the Solution Adapter and enables tuning of the control loop using Compensation Designer and SFRA. High efficiency, low THD and intuitive software make this design attractive for engineers working on inverter design for UPS and alternative energy applications such as PV Inverters, Grid Storage, micro grids.

base hardware only: $749
One of either daughter boards:
$99 TMDSCNCD280049C — F280049C controlCARD Evaluation Module
$159 TMDSCNCD28379D — F28379D controlCARD for C2000 Real time control development kits

Required software (looks to be available at no charge):
C2000WARE-DIGITALPOWER-SDK — DigitalPower SDK for C2000 Real-time Controllers
CCSTUDIO — Code Composer Studio™ integrated development environment (IDE)

The Texas Instrument kit provides everything I could hope for and have a reasonable chance of understanding the engineering. So I’m returning the Enphase – untouched, for a refund at Amazon charged me about $7 for the return shipping vis UPS.

I reached out Felix Rusu at https://lowpowerlab.com/ for his opinion on what it would take to make a circuit board similar to the TI Development Kit. He noted that with the parts shortage: probably impossible at this time. Moreover, the Development Kit circuit board was larger than he could handle. His practical advice was two prong: an open source product would have to either be 1) significantly cheaper than currently available on the market, and/or 2) provide features that are highly desirable and not offered by current producers. It is my opinion that there is no way to undercut commercial offering in price, so that leave only the second prong: “not currently offered by current producers”.

My hope was to look into an open source solution that provides microinverters that can be hooked up to the grid AND work when the grid is down. I’m concluding that to hook up to the grid, the certification phase would be an almost impossible hurdle to overcome. Underwriter Labs certification? Probably several hundreds of thousands of dollars – and that would be for a fixed design. With an open source project which would have design changes at a rapid pace, it just doesn’t seem to possible unless there was a significant demand. That brings me to another point: demand. Working in the area of conversion from DC to AC acceptable to a grid probably is understood by less than 1/10,000 or 1/100,000 people. I consider myself somewhat sophisticated and I did not fully understand as I was dealing with Sun Power on a bid for a 7k system that would feed into the grid and work off grid. It wasn’t until well into the phase where I felt I might sign a contract that as a last minute disclosure, they said (not in writing) that I would have to purchase a battery system if I wanted power when the grid was done. How much? Response: $20,000 - 30,000. Discussion quickly came to a closure. Lastly, if you build system, who is going to service it when you are not available? I can understand why companies installing products known and vetted would not want to touch a home-built solution. I’m 65 and if I build something, it’s going to have to be extremely simple lest I forget as I grow older, or something happens to me and my successors are left with a contraption that nobody want to tinker with. So getting back to demand, I think demand for an open source microinverter would have to be knowledgable demand and I do not see there is sufficient enough of such. It would take a government sponsored program move something like an open source hardware microinverter certified for grid connection to have enough momentum… and then who would go the do-it-yourself route to use it? And with the disclosure of how Uber manipulates government officials in Europe, I can only imagine that companies making and selling microinverters would have formidable political clout to kill any open source hardware project that would threaten their revenues.

I did reach out to a Senator from Oregon expressing my concern about solar power generation after Cascadia and is there any program which might fund a research project having open source hardware and I received back a machine-like generated response extolling upon the Senator’s achievements in green technology. That was a waste of my time and the Senators mail response machine.

I still may try to do something off-grid, but with microconverters on one or two panels, but we’ll see. The concept of micro inversions so you can have a mix or panels without reducing the power of the entire network is really seductive.

Thank you to everyone to pitched in with comments and suggestions.