diy solar

diy solar

What I'm learning about DIY Enphase and Hoymiles Microinverters

Merry Christmas, y'all!

We had a brief power outage two nights ago, on the coldest night of the year. Temperatures got down to 5 deg-F, with a wind chill of -5! My DIY battery system combined with my Enphase IQ Battery 3T worked great!
You mean:
Your Enphase IQ battery 3T generated the micro grid and your AC coupled DIY hoymiles inverters (powered by batteries) helped.
 
You mean:
Your Enphase IQ battery 3T generated the micro grid and your AC coupled DIY hoymiles inverters (powered by batteries) helped.
Yes, and our fear was that the outage would last all night and my IQ Battery 3T would only last about 2-1/2 hours with the heater running. We only have electric heat here. The DIY battery setup allowed me to run the Enphase battery at only 200W so that it could last all night.
 
Yes, and our fear was that the outage would last all night and my IQ Battery 3T would only last about 2-1/2 hours with the heater running. We only have electric heat here. The DIY battery setup allowed me to run the Enphase battery at only 200W so that it could last all night.

It's great that your ad-hoc AC battery worked but it really seems like a different solution without iQ battery would have been simpler/cheaper.
 
It's great that your ad-hoc AC battery worked but it really seems like a different solution without iQ battery would have been simpler/cheaper.
Probably, but I fell for Enphase and already had it in place. However, there are not many choices for UL 1741 "SA" qualified off-grid inverters that can do Freq/Watt control of the output. The new EG4 Hybrid says it is, so I would like to try this with one of those, but I can't afford one right now. I don't have enough subscribers to source it yet. :)
 
Probably, but I fell for Enphase and already had it in place. However, there are not many choices for UL 1741 "SA" qualified off-grid inverters that can do Freq/Watt control of the output. The new EG4 Hybrid says it is, so I would like to try this with one of those, but I can't afford one right now. I don't have enough subscribers to source it yet. :)
Which one do you want? Maybe signature solar will sponsor you by sending you one to test out because obviously if you can make it work they will make them sell.
 
Which one do you want? Maybe signature solar will sponsor you by sending you one to test out because obviously if you can make it work they will make them sell.
The new Hybrid 8kW hybrid is the only one that might work because it is 1741 "SA" certified. The 6kW doesn't say one way or the other.
 
What is ‘SA’ and why might one want it?

These say: ‘Globally certified for UL1741, SAA[/b{, TUV, VDE-AR-N 4105, VDE 0126, G83/2, CEL 021, IEC61727, EN50438, ABNT NBR 6149/16150’

Is that the same thing?
UL1741SA has the specifications included for Freq/Watt and Volt/Watt control, that 1741 didn't have.
 
UL1741SA has the specifications included for Freq/Watt and Volt/Watt control, that 1741 didn't have.
So 1741SA is just basic throttling of output through frequency shift (or voltage shift)?

I thought for several years now all Microinverters approved for sale in California needed to support that.

The APS-DS3: https://northernep.com/downloads/technical-sheet/BDM-600-Microinverter.pdf

also says: ‘CA Rule 21 (UL 1741 SA)’

While the detailed table of the NEP BDM600 states:

‘UL 1741 CSA C22.2 No. 107.1 IEEE 154’

So don’t all of these Microinverters support frequency shift and is there any reason either of them would not work in your setup for backup power as well as your Enphase Microinverters?
 
You're missing the point. Grid-tied inverters have UL1741SA. Off-grid inverters mostly don't have UL1741 at all, or only have UL1741 without the SA, because they are not grid-tied. The new EG4 8kW is the ONLY one I've seen that is both, outside of Tesla, Enphase, & Generac.
 
You're missing the point. Grid-tied inverters have UL1741SA. Off-grid inverters mostly don't have UL1741 at all, or only have UL1741 without the SA, because they are not grid-tied. The new EG4 8kW is the ONLY one I've seen that is both, outside of Tesla, Enphase, & Generac.
Ahhh, gotcha. I though you were referring to the Microinverters themselves, not the hybrid inverters that can control them for backup power when the grid is down.

That makes slot more sense and thanks for clarifying.
 
@ncsolarelectric I'm curious about a related, possibly "opposite" application of Hoymiles export management. IIRC, you use it here to force the microinverters to behave like AC coupled batteries in local consumption only mode.

AC coupled systems typically have a recommended ratio of (solar AC output / battery inverter output) and and (solar AC power / battery capacity) when running off-grid. Do you think the export management can be used to adjust the output of a Hoymiles string to conform with this ratio? One worry is the 15-30 second response time you list above in the export management. I'm not sure how much value it adds on top of the frequency shift.
 
Hi Z,
All you can do is set the maximum export power limit, and determine where to put your meter CTs, per the diagram shown.
I've seen the inverters frequency shift the output current on their own to prevent over voltage on the microgrid, so that is apparently a hardware safety that adjusts the power factor much faster than 15-30 seconds. But true export control by reducing the output current is much slower as it depends on the RS485 and Wifi network to get the message from the Meter => DTU => Inverters. It takes time.


Screen Shot 2023-02-23 at 10.41.00 AM.png
 
Thanks for the UI pic, that is helpful. IIRC this can be set over their modbus API too.

Putting down a couple thoughts for anyone who might come across this thread in the future.

I think the export management can help in a few ways —
1) don’t have to frequency shift as much if the overproduction occurs over a slow timescale. This leaves more “precision” available to make small adjustments
2) another knob to tweak if AC coupling is not as stable as desired
3) provides redundancy on top of frequency shift in case one or the other mechanism gets confused

If this is just used to clip the mid day production bump, 15-30 seconds is probably fine since the change is pretty slow.

Can’t see how it would hurt other than clipping more than needed, compared to if the solar panels were DC coupled or using faster responding microinverters (maybe the PLC ones are quicker).

In the normal export control case, 15-30 seconds will not respond quickly enough to hide appliances turning off, so it depends on the POCO export agreement as to whether averaging it out over time is acceptable. If it’s not, then this is not the right system. I believe Hoymiles targets Australia and Germany and has examples in their application notes for configuring export limits appropriate to those markets.

15-30 seconds with 2.4G radio control is probably not good enough to qualify for reducing the size of circuit conductors, and I’m not sure this feature is listed for that anyway.
 
Yup. I received approval and final certification this morning. Yay! :cool:
I also learned that they use an inaccurate rule of thumb regarding NEC 2017, 710.15(A). They said regarding my SLD that the ESS must provide less than or equal to the largest circuit breaker in the backup sub-panel. That is not what the code says. The code says, that the capacity of the ESS "...shall be equal to or greater than the load posed by the largest single utilization equipment connected to the system." It is clear throughout the code book that "utilization equipment" is referring to the actual loads themselves and not the size of the circuit breakers in the panel. Clearly, such a restriction would be unreasonable considering Eaton, SQD, GE do not make AFCI breakers of values less than 15A, or breakers less than 10A in general. The wiring must be sized appropriately for the breaker, but the breaker value is not what NEC 710.15 is looking for. They almost had me convinced I was screwed using an IQ Battery 3T, because I can't get 6A breakers! Thankfully, I verified exactly what the code says and I'm good with my breaker ratings as long as my loads don't pull more than 5.33A continuous, or 6.33 surge at 240V.
One more subtle way for Enphase to sell more batteries.
You cought it, many will miss it and buy another 3T.
Cheers!
 
Yes, that's close. I have the CT on the Grid input to my main panel to cap at zero-export for the HM inverters. I also have a relay where you have the CT, to turn off the inverter, and I have my current controller between the battery and the inverter that also controls the relay, low battery shutdown, and normal battery restart.

Is that smaller PCB on lower right your current controller?

Is the design finalized?

Has it gone on for UL certification?

Are there any specs you can share?

Interested in any update you can provide.
I'll be din-rail mounting the current controllers next to those 10A DC breakers. There is a 15A battery charger sitting in the bottom, that has no mounting brackets or available screw holes to mount it. :unsure: Above that are my two battery banks, about ~7.5kWh. I'm waiting on the final production sample to arrive next week and 6 x HM-350NT inverters. I should have the whole thing assembled and hooked up in my garage in a week or two.
Interested to understand whether those December 2022 final production samples actually ended up being final as well as any schematic or drawing you might have that you can share of how your whole rig is wired up…
1. What I learned recently is, the monitoring doesn't work with the Dual and Quad inverters when all the Neg DC inputs are ganged together at the battery, because they are already internally ganged together inside the Dual microinverter. Due to this, the return current is not equally shared and takes the path of least resistance, causing errors in the current and power measurements as shown on the software. This does not happen when I was running two single microinverters, so I ordered all single HM-350NTs to use on this. That solves the problem.

2. Next I have some HMS-2000 2kW Quad microinverters coming in that are going to be tested with 455W modules. I wish I had bigger modules to test them with. They can output up to 475W per quadrant, so the 455 should have no clipping at all.

View attachment 119840
 
Back
Top