I'm an engineer working at Enphase, working on the envoy's software. It just feels so cool to see Dave talk about the Enphase system.

I work on macro inverters. Utility scale. Up to 4MW for a single inverter in some cases. And on those puppies, they disconnect the DC input and AC output from the inverter via contactors when the inverter is idle, for exactly this reason. Otherwise they would be a giant VAR suck. However, sometimes you can order them with what we call NIGHT SVC (Static Var Compensator) mode, which allows you to send a signal to the inverter to reconnect its AC output so that the output caps are back in the circuit, and thus the inverter can act as a capacitor bank at night.

⚠ SPOILER ALERT ⚠ Short answer: around 7w for the envoy & q-relay and near 0w real power for the micro inverters. Long answer: there is also a capacitive load of about 1µF per inverter which will cause some power dissipation in the wires.

This video is much more accurate than the first go... thanks for re-doing it. I've got 36 IQ7+, and your first vid had me seriously considering a contactor disconnect.

I worked in the tester development & deployment of a autoline for these Enphase Invertors. I follow you from long time, and it's proud to see you using a product tested by the testers developed by my team. :)

I have bought a very expensive but very optimized Unique 12/24V fridge. It uses about 600 Wh per day. It must be smaller than your family's fridge though. It works very well, but a bit slow, you must not put hot items in it, but now I have ice cream off grid! When I started to live off solar panels (and a little wind), I was worried about the night consumption of the charge controller screen and monitoring. But after a few years, I realize this power consumption is neglectable, I boil water all the time, cook and get coffee, I mean, one coffee in the morning must use the same amount of energy as the charge controller over one year! If I get a decent water heater with an element inside instead of my current oven element with a pot, I will save thousand times more energy that trying to optimize the generation system. There are other more important things to do, such as not forgetting the computer powerred on while doing something else.

The couple of watts drawn at night seems to match what I see at home. Eight IQ7 inverters and an Envoy S Metered, which I think is on the solar panel side of the production CT, and right now (11:39pm) it's showing a production power of -7 watts. When it comes to battery storage take a look at the dedicated units for the task such as the Sofar ME3000SP. Hook it up to a 48 volt battery bank and it'll draw up to 3kW of excess power (that would go to the grid) or supply up to 3.3kW on demand. Basically it tries its best to stop your meter turning, in either direction.

Suggestion to really show apparent vs. real power: Put it on a scope. Differential probe on channel one (safety first) and current clamp probe on channel two. Open a math channel as the product of one and two. The cycle mean will be the real power. The phase shift between channels one and two will let you calculate P.F. You could compare the vrms and irms with P.F. to the math channel result. Would make a nice video…

I also had my system set to top off a battery bank before switching on the enphase system. I installed a DC distribution system that ran, Co detector, alarm clock, wifi router, cable modem, network hub, cell phone, laptop, ecig, and basically everything with a DC wall wort. In the event of power outage 600ah battery could power an inverter run some of my more important electronics.

Thanks for the thorough coverage! 😄 To go off on a tangent about PFC, I find it really annoying that for example when it comes to power supplies with PFC, manufacturers crow from the rooftops, telling the amazing PFC value their PSU has. Except, they only give you one value at an optimal (high) load. However, PFC can be very poor at low loads. In typical applications the typical load may be only 50% of maximum PSU load capacity (or even less), where PFC specs are a LOT worse. They really should be be publishing graphs showing what happens across the entire load range. Same goes for efficiency, although it's more common to publish efficency curves, at least.

a very important subject on the factory floor and now at home when using inverters to keep clean grid electricity, which can be a big issue with VFD AC motor drives and 5th, 7th, etc. harmonics. those inverters usually need additional inductors and capacitors as per installation requires so not to pay for wasted reactive power "foam on the beer" 😎 thanks, great video

10:03 Apparent Power S is NOT the (linear) sum of Active Power P and Reactive Power Q, but instead S²=P²+Q²

Thx Dave, just decided to buy 14 Sunpower Maxeon panels installation with integrated IQ7A micro inverters. 🇳🇱

I would say the relays disconnect if they see a line voltage with the panels feeding in of over 253VAC for a sustained period, probably 5 minutes, and then disconnect to drop line voltage for the grid in your area. Disconnect will be at the furtherest points of the feed from each transformer first, simply because of the line resistance between the houses and the transformers, as that will have the largest swing in voltage, especially as the entire feed line to the transformer has solar panels feeding in, so the furtherest panels have to raise their generation voltage higher to feed in power. Then likely they wait for 5 minutes after the voltage has dropped below a hysteresis point, probably 243VAC, before they connect the panels again. 5 minutes to give a chance for grid company on line tap changers to do a tap change at the MV power stations, and drop to a lower tapping point, to reduce the voltage they see reflected back to them from all the residential transformers they feed, accepting more power in at the lower mains voltage, to feed back to the system to the large loads that are still using it. Otherwise if you do not have tap changing and no residential disconnects your nominal 230VAC will go out of the upper bound of 253VAC, and the rise has been known to get to 270VAC in cases, so the relays are definitely needed. Yes trip points, delays and hysteresis is likely all remotely programmable, and they probably have a very wide range as well, along with giving good data to the overall health of the system, so the data is available, but the setpoints are likely not exposed to be adjusted, only for the manufacturer to set, not even the installer is likely to have access to them. But no doubt sniffable by simple traffic analysis of the data over the power line, and then you can get the protocol, and the commands that are sent, to get a probable command set to try, though there is good chance of scrambling the device during this process.

At my company we have the hole factory coverd with solar and with that the power faktor is nicely held at around 0.95 with the power factor correction in the NSHV turned off. Also these Shut off relais are intresting here in Germanny we get a device installed from the grid provider that also talks over the powerline that gives a signal that has to shut off the solar/generator probably similar communication technique

The newer IQ8H does list night time power consumption as 60mW, which seems reasonable.

Given that the average home these days has many electronics and appliances with line filtering in them, I wonder how many uF of capacitive reactance a typical home presents to the mains.

Many modern installations don't use micro-inverters, instead they use DC Optimisers. They work in a different way, you still have a main inverter, but achieve the higher efficiency and eliminate the single point failure / shading problems of a string system. DC Optimisers keep the panel wiring at high voltage (lower current) so not heavy cables. Both solutions allow you to monitor the output of each panel.

Dave Jones, flying spaghetti monster amongst mere mortals. Science bless you 🙌

I have Chinese micro inverters on my roof, Deye brand. They seem to shut down entirely when panel voltage drops below a certain level. I can't even access their management interface at night. I suspect the microcontroller that provides the Wi-Fi connection is fed from the DC bus, as the management interface is still accessible when I disconnect the circuit breakers that connect to the inverter.

Some bigger solar installations (2MW - 10 MW) get ratings from the net provider for maximum apparent power and maximum ratings for anything. It's more for the bigger systems which must be able to disconnect from the grid operated by the net provider. They don't have these relays but big 30kV/110kV switch cells and you need a special license to operate these switches (hot sticks ... ppe; 240V/380V is low voltage system; 30kV is medium voltage system and 110kV and above is high voltage system). If there are to much of this systems in small homes they might get commands from the grid control center too, it may be manages by enphase (more control of the grid if they control all enphase inverters remotely).

Generally, they n my expertise, grid scale inverters can deliver any combination of real and reactive power as long as their apparent power limit isn’t exceeded so they can provide reactive power compensation while exporting power. The more advanced versions can adjust reactive power based on grid conditions or a command from the utility. Also Dave didn’t mention it but the Addition of real and reactive power to get apparent power is a vector addition not a scalar one so 1W+1VAR doesn’t equal 2VA

Just use the street lights to make power, ha! I had 9.25kw envoy system at my old house. It was pretty neat to see all the data. Now I just have an OFF-GRID setup 11.25kw with midnite solar and magnasine command center.

I often wondered why I see the clamp meter run to about 0.1 amp per inverter when I start up the Enphase kits prior to the inverters starting up properly.

Can you put in a second inverter but plug it in the power point? Or straight into the power box?

At 32:55, where you're talking about lower inverter efficiency for lower amp draw... a single Enphase battery uses multiple microinverters to invert power. At low power draws, where a large monolithic inverter would suffer a huge efficiency penalty, Enphase only need turn on a minimum number of microinverters ensuring they are always closer to the sweet spot. I also suspect that because the individual microinverters have relatively small ranges, the sweet spot is a lot wider than a single huge monolithic inverter.

17:10 BigClive found out that some are not even real capacitors. Some are failed runs, not filled with electrolyte.

working on an inverter some years ago, wanted to get standby power under 1 watt, but wifi modules are surprisingly power hungry. We shut off every non essential rail until the sun comes out.

for the power billing, some countries actually bill residents for VA and not W, which smart meters measure in both VA and W, and can be changed at any time. not relevant to solar, but surely to the majority.

I did some power factor testing on my off-grid inverter system a few years back and saw between 10% and 30% DC power reduction with "improved" power factor on a freezer and a drill press motor: kzfaq.info/get/bejne/p8yPja96vrqUgqM.html This was on a low frequency Tripplite sine wave inverter/charger, so that may make a difference.

And I have analog single stator meters on both of my panel sets, and also use Enphase microinverters on my grid tied solar... Also a net meter that was installed by the power company. The interesting thing is, on a brightly moonlit night, like the ones you can see colors on... I have seen some, not a lot, but some generation. Lunar and Solar power... what a trip huh?

Anyone wanting a real in depth read of how micro inverters work can have a read of the Microchip app note AN1444. It's a little dated, but still relevant.

From my research Enphase is the way to go. But what do I know 🎈

How comparable are "power optimizers" with a central inverter to the Microinverters? Am getting a 10KW array installed next week (Wollongong, weather permitting), so will be able to start being a data nerd on the data soon after.

There also used to be a lot of data that you could pull from the inverters when the system first came out and for some reason they disabled a bunch of the things that you could graph and chart it used to input voltage, input current, track temperature of the inverter along with some other parameters and then one day they just did a software update and all of a sudden half of the things that it could graph disappeared.

Dave, there's a DC linkage cap at the input (solar panel side) to your CCT box. They use the freewheel diodes to rectify the AC into DC to maintain the charge. These inverters act as a real load (at night) to adjust the phase angle of the current. I also agree. The 1A would come from the grid, even in a solar+storage configuration. The battery will supply that current but the grid will charge the battery. Therefore, the current comes from the grid.

BTW I've found having panels on east and west to outperform south facing panels. Grid-Tied, and Off-Grid.

Dave, does the sunnyboy has standby power? And I wonder if you could measure the production current and voltage (power) before the conversion on a panel. And then also the output power on AC side. Give you (and us) a view in power used by the converter in production mode.

get some inverter fridges and run them off DC, works like a charm :)

My enphase system has been running for 11 months now, and at night the whole system draws 7 watts at night. (Envoy and 20 micro inverters.)

could you use these inverters with a battery pack of the same nominal voltage as the panels and charge them during the daytime, then, at night use a relay to switch from panels to batteries and use the inverters?

I assume the battery system you're looking at will be grid-tie only due to the rest of your system's existing hardware limitations so there won't be a practical demo of "following the VA"?

I have a grid tie system with 4 350 watt panels and 4 enphase m215 inverters, I turn the power off to the cottage in the winter and back feed to the net metering program to build credits, but unfortunately I'm finding they consume more power at night then what they produce, my last bill showed I used 35kw and only produced 2 kw, so I may connect a relay to power them down at night.

So I have a question... Could you theoretically eliminate all of the Alternating current and change everything over to a + - direct current system? Would appliances still work?

It is Big Power Grid that makes the solar inverters to make shure there is no net payback for injecting solar power to the grid ;)

Italy residential we pay apparent power, so this will sux here

Awesome video Dave! Have you done any reviews of different systems. Have been thinking about get a solar system!

Yeah outside of youtube videos on the topic, I've never heard of apparent power. Residential power in the US (220VAC) isn't apparent power based.

I'm not an expert and may be wrong, but don't the new smart electricity metres in the UK measure apparent power?

someone can help me ? i´ve a 10kw solis inverter and by the night he is consuming 30W, i measure that with a Fluke 435 serie II analyzer, i dont now the cause of it.

Could you place a large enough Inductor/transformer on the inverter output could you balance it out?

Pretty jealous. I'm only making 30kwh on good winter days in Tassie off 10kw of enphase gear... Still well paying for its self over the year but its amazing how much difference the latitude makes.

There are people who get charged for power factor and these would be large users so the micro inverters may not be so good for them.

I will never understand power analysis like Dave

Of course you end up paying for VARs. Maybe not directly, but indirectly the power company has to pass that on somewhere. It's a bit like saying you get roads for free, forgetting you're paying g tax.

Here we were/are billed apparent power, they upgraded to digital and actually still do because the meters can record it.

Hi Dave, could you do a video on how solar inverters interact with the common split-phase 120V/120V here in the US. The Enphase inverters only put out 240V, not 120V. If the loads in my house are not perfectly balanced between the phases, I'm puzzled how exactly this works. I assume the 240V is exported through the meter and the street transformers, due to the center tap, can send the same energy back as 120V? If 1 A was flowing across the 240V lines (240W), and my load was drawing 2A at 120V (still 240W) on one leg, what does the flow look like to the meter? There is significant legislature debate here in California (and other places) to change the net metering plans such that the energy you export is not nearly as valuable as the energy you import, so it might be nothing like "net" metering anymore. The power companies are arguing why should they pay you $0.35/KWN for power to the grid when they can buy power wholesale for $0.06/KWH. They leave out the detail that the KWH from their wholesale generator 100 miles away going to my neighbor's home has much higher transmission costs than the KWH from my my solar panels going 100 feet to my neighbors home. In one proposal, you get credit at the wholesale rate for power you export, like $0.06/KWH, but pay $0.35/KWN for power you import. Load smoothing batteries could help a lot for a day or two, but don't at all help if I want excess energy exports in the summer to build up a net balance in an account to run a heat pump in the winter. If this much worse net metering becomes reality, I don't want to get paid $0.06KWH for the power I send to the street transformer to be turned into 120V, only to buy it back for $0.35KWH. I read there are potentially devices that could do the split-phase conversion on my side of the meter. Thanks!

Hi. I've seen some of the micro-inverters and their electronics. Found that these micro-inverters are not pure sine wave. Just wondering if all are designed that way. would you be willing to do a tear down of one of your units :)

can you please make a short video explaining your browser addons and how and why are you using them?

Would you be able to show the guts of a micro-inverter? I'm curious as how the circuit works.

My DaveCad license is running out. Are there any special offers announced for the coming months?

I want a microinverter system. But it MUST have an on-premises web interface for local management. I refuse to be dependent on a cloud-based system.

I think this calls for a teardown (please buy a cheap [not working unit] from ebay or something)... even if it is potted. ;-)

What happens when everyone has such a thing on the house, Dave. What will be the ramifications for the grid? You can't use it on business buildings right? Otherwise I would have to turn on my electric motor over night.

Fascinating. I remember my EE prof talking about how it’s called imaginary power, but it can kill you! Our power systems courses were terrifying. Safety talks, several wooden rods so you can separate a power source from the victim. It’s an interesting discussion as to whether consumers can power their own houses independently or if they have to connect it to the grid. During peak generation do they disconnect your solar array and your house is only powered from the grid, or is your house powered solely from your solar array?

survival power!

I wonder if the relays disconnect from the grid in a power out?

I would go with a string inverter personally dont see how those are going to handle large inductive loads

Common Dave, ideal capacitor. 10VA!=10VAr You still need bigger transformers and hence you have more losses in the grid! And yes, you pay indirect for the grid infrastructure and losses.

...mybe in the future these devices could find the best yield with auto switching parallel to series vice versa

Can anyone from Spain and EEVblog suggest me a solar panel system? io live in Alicante where the sun never goes... roof orientation east -west with south inclination.

My PV inverter shows it uses about 70w at night, it was installed 2011

If I understand correctly, No inverter real power, but real source and connector power required

Me too almost exactly. Had 2400 watt PV using Sunny Boy and Fronius inverters. And then I got 12 more panels with Enphase IQ7s. And I too saw the 1 Amp at night and went HUH? Thank you thank you thank for this. No worries whatsoever while I'm grid tied. But. I've got two Tesla PowerWalls too. If we get a Texas in Utah and the grid goes down for two weeks, a lost KWattt-hr or so from the PowerWalls that I could have tossed into my all electric heat pump furnace could be missed. Not a major worry, but I would like to know how much the reactive power of not only the IQ7s but all the other reactive loads in my house degrade the storage capacity of my PowerWalls. Would it pay to install an adaptive power factor correction gizmo, not a phony baloney one, but a real one?

7:33 Again. I as stupid electronic noob, would put a transistor on the output side that is triggered by having power on the input side of the converter ⇒ converter uses no power if none comes in.

And the answer is: Wait for it. No. Storage capacity of the Tesla PowerWall is NOT degraded by reactive load. In fact, their interface seems to have buillt-in power factor correction. You know? I think I'll buy a Tesla car. At Julian day 287 06:00, our irrigation pump turns on. It's a 120 Volt pump, so it presents a 1.5 kW load on leg 2 of our household. Notice the reactive power of that pump around 1 kW. But notice the load that is presented to the PowerWall. The load is evenly distributed between leg 1 and leg 2 and the reactive power is flat zero. I've put the data logger charts on my Facebook page.

Isn't it beneficial to go to DC in your home when you have a battery system independed of the grid? There are so many losses in the conversion, seems like a waste

Excellent video. Love it. But let me get this straight. Say it's winter in Utah and dark for 16 hours per day. My IQ7s draw 1 Amp at night purely reactive power. I've verified that. We're not talking "Wattful" power but power that nonetheless is sapping the storage capacity of my PowerWalls. By how much? Is it simply reactive power X time? 1 Amp X 0.240 KVAC = 0.24 KWatt for 16 hours = 3.84 KWatt-hr per day = 26.88 KWatt-hr per week. Ouch. I would like to have thrown that 26.88 KWatt-hr into our furnace if there's a massive grid outage like during the winter of '21 in Texas.

Enphase related, Dave I wonder if you have heard about Enphase breaking local API access on the latest envoy firmwares. I've firewalled my Envoy away from the internet now so it can't update or reach enphase servers.

My sunny boy 2.5 uses 1.5w overnight measured. i.e insignificant. I'd be willing to wager if you actually measured the watts flowing back out up on your roof AC wire you'd be suprised (in a bad way).

The answer is yes.

When they first came out I purchased a few of the Enphase m190 inverters and had a very small system I was concerned enough about night time power draw that I used a couple of solid state relays in my solar panel circuit breaker box and set it up so if the panels weren't producing at least two or three volts that cut the AC to the grid tie inverters and envoy. I know I measured how much power they were taking at the time but it was close to 15 years ago and honestly I don't remember I just know it was enough that I was not wanting to waste the power. I think my envoy took about 5w assuming 12 hour average night that's 21kw a year it would have taken me about 10 days of generation to cover. I also planed ahead for degradation of the panels and used 250 Watts worth of panels for the 190 W inverter.

in 848, that cap was 5uF.

Jeez, I would be tossing those fridges and freezers if they were turning on and off that many times in an hour

Your main spec is wrong it isn't +/- 10 it's +10 - 6 it use to be +/- 6 but they increased the + when they reduced our official voltage from 240v to 230v to keep all existing systems within spec

With the cost of electricity doubling (and more come October) here in the UK I am looking at battery systems (again!) and Dave's idea of fridges/freezers etc only has me thinking!......and not to mention also measuring the standby power on my Bosch solar inverter itself etc. Hmmmm!

Seems like everyone wants to get into the battery game. There is even a standard for off-grid and grid-disconnect situations to allow battery inverters to tell micro-inverters and grid-tie inverters to reduce the amount of power they send to the micro-grid/home-grid. The battery inverter shifts the frequency up from 60hz to 62hz or so (something like that) to regulate the amount of power the solar inverters inject onto the line. Without that the solar inverters would push the voltage all the way up to their maximum (typically 130VAC or 140VAC in the U.S.) and then probably either shut-down or start hicup'ing (turning completely off, then on again), when what you really want instead is regulated load-support for whatever is running in the home plus whatever the battery system can take (if anything). And because the standard is fairly new, various inverters either don't implement it or implement it with different frequency ranges and various brands tolerate different voltage limits before they trip-off. So it's a bit of a mess right now. -Matt

What happens as failures occur?

hello, nice video ! will be great to have a efficiency test bench to get a curve, because they don't give one... they said 98% efficiency, max in power 450W and max out 250W ... IQ7+

What happened to the last video on this?

What About A Clear Night. Doesn't the moon light make for activity

Why do we make the assumption that if we have a grid connected system with a battery, the apparent power draw at night will almost certainly come from the grid?

Dave, check out the channel "Ben's Solar and Battery". He opens up a I7+ and depots it.

Hola a tod@s, Me gustaría poder recibir la señal de un mando con un HCS301 (KeeLoq de MICROCHIP) en el emisor y con un PICxxx o un ATMELxxx en el receptor. Hay una librería para ARDUINO sobre recibir la señal de un HCS301 y otra para emitir la señal de un HCS301, pero no lo encuentro... sigo a la búsqueda. Un saludo a todos.

Can I license a copy of DaveCAD, or find it at the play store? 🤣

Disconnect the it over night, there is NO benefit connecting it during night time!

Enphase sounds "like" Inphase and htats cool!!!!

Question: Do you segment your panel groups with diodes? For instance, when sun shines on half of the group and not on the other half, the groups in the shade are a drain on the rest that are in the sun. Diodes can thwart this but, the current drop across a diode separating groups of PVs is a deficiency in and of itself. I built a 13kV solar tracking system back in 2009 and discovered, at least in that case, the diodes helped more than harmed. System "turned" on and made power much earlier and later in the day than the same system w/o diodes. I.e. the voltage got high enough for IGBTs to output a useful voltage earlier and later. How much did I lose in terms of power because of the diodes? I don't really know. Overall experience building that solar tracker: Not really worth the cost. Although it is still working today and has collected data since 2009. I haven't looked at the data for a long time. One sort of loses interest after a while. I do know that the electric meter is still running 'net backwards' for the building the thing is hooked up to.

I have a very similar setup to you Dave. An older set of panels with 2 Sunnyboy Inverters and a new system with 15 micro inverters. The Enphase/Envoy monitor box is on a dedicated circuit in a new SPAN (very cool breaker box) box. The monitoring software for the SPAN box shows a 24/7 draw of 4 watts. The new system was self installed and my son did a great video on it kzfaq.info/get/bejne/jduila5zxMqzoX0.html (the SPAN box was installed recently so it does not appear in the video).

This looks like a bit back pedalling video ... you wore much more woohoo lots of power now ... eeee.. anyway still good video.