That fuse box is the epitome of "Good enough for Australia"

Over here in the Netherlands we solved this problem by not having any sun to begin with.

Aussie fuse boards look absolutely terrifying! I'm amazed the country isn't on fire more often!

As we say in italy: "The shoemaker has a hole in his shoes". That box is simply nasty!

That fuse panel, i'm speechless. Amazing that it's legal to have that mess.

That "switchbox" looks like a 1950's radio! Has Big Clive seen this video yet?

Wow, that fusebox made me wince. ...situated outdoors in a rickety box no less. Yikes!

This video was maddening to watch. It's obvious what the problem is and it took you 15 minutes to get to the point.

Forget onions! That distribution panel makes me cry quicker!

Short answer: Everything's upside-down in Australia

Dang that panel is a mess! Would love to see you rewiring it

Should point out that "whacking them in parallel" also only makes sense if the two current clamps have the same ratio (e.g., 1mA per A).

The Aussie fuse boards wiring looks hunky-dory to me. No DIN rails, no distribution boxes

Yikes, that box is calling for a few terminal blocks, rails and daily beatings for the electrician who installed it; until morale improves.

At "stop this video" I'm already tearing my hair out..figured it out as soon as you said two systems not aware of each other

There are some nice busbars for circuit breakers, it now makes more sense why they exist....

instead of giving each monitor seeing the sum of both productions, what about set each monitors consumption to house feed minus the others production. In parallel but one reversed. They now see their only production. consumption measured ( -3A out, + 2A )= -1, 2A production, 1A calculated consumption.

ref: 26:12 Has never seen a nest like that in Norway where I live. Even in houses build in the 60's they mounted meters and fuses on power-rails.

If your wires had been long enough and the clamps wide enough you should have been able to drag the Enphase power through the reference clamp of your old system and the old system's power through the Enphase reference. This way and given the correct direction you should have been able to see the difference of the currents with single clamps.

22:08 - Oh, boy! A spaghetti factory run wild! I retract my 'solution' below! ;)

WOW... that fusebox is a real mess ! That cant be up to code ?

I have a feeling a breaker box like that would be highly illegal in Canada.

Panels so powerful they give power to the sun.

Why not have two current clamps on every solar system output. If you wire one of them to the production input of the data logger of the corresponding system and the other one to the load input of the other systems consumption input in parallel with the actual load measuring clamp but having the wires reversed for the clamp measuring the production of the other system (or just reversing the clamps install direction) you would subtract the generated power of the other system from the measured "load" and get the real remaining load. EDIT: This is of course assuming that all current clamps have the same ratio, the output of the system is reasonably well synched to the grid (which it should be with the inverters tracking the grid frequency) and no massive amount of inductive or capacitive loads being downstream of the swichbox.

is this a legal powerboard? it looks like it will burn your house down every second now!

Installers should have known all of that and informed you and asked if you wanted the correction installed. Fail!

So, any chance negative consumption could be called positive production...that might explain relation to the Sun activity?

Haven't seen 1080p50 for a while, what a throwback.

My system doesn't read the consumption, the power meter does that. SMA sells a meter for that, but my installer didn't bother, I guess it's not that popular in Germany. So for now I made my own scripts for monitoring (added to temperature/humidity monitoring in the house) to feed a Grafana server with the solar production. One of these days I swear I'll get to plug the optical transceiver for the energy meter to read those values, both power consumption/feeding and 24h/7d/365d values. I guess if you had a system with equipment from the same manufacturer they could manage to get you something that does that out of the box. In my case, I can't be bothered with the SMA website other than checking the monthly trends and the web interface to the inverter is bad, so I made my own thing :-D So the alternative is to find the API manual for the Enphase and Solar Analytics inverters, merge the data and read the consumption separately, and merge again. We're nerds, aren't we? :-P

Wow that breaker box meets code anywhere on this planet? Scary looking rats nest! Is that made from peg board Dave?

Looks perfectly fine for me. However, for Australians it looks upside down. Installer must have forgotten that he needs to change polarity because everything in Australia is upside down.

Or you could just whack the wires coming from both systems into the same current clamp and still use that spare clamp to measure your ACs.

Paralleling 2 current transformers in order for them to double the current is absolutely contingent on them having the same turns ratio. Ideally being the same make and model. Otherwise if one had 1000:1 turns and the other had only 100:1 then you be adding 10 times less current from the second transformer.

The back of that panel looks like a right lash up no neatness no labels on the cables.

My (simplistic perhaps) solution: Pre-join the arrays and have one measuring system measure the joint power production.

Excellently clear as mud. One man's negative is another man's positive (especially in electron flow) I still don't understand the basics of how excess power can be fed back the grid, but there must be a vid somewhere.

Dave, It is permissible to place two large opening size current clamp after the load breakers. It will read the sum of all the loads. This is easiest when all the loads exit the same side of the box, such as the bottom. Now both current transformers will read only the total load. Both sources will properly record the production of both, and show the system load, but will show the total system load even though there are 3 sources of power. Each system will show the total load, and each unit production. A buss bar from the main AC breaker to the load breakers is common. Instead of metering the AC utility, simply sum the individual loads through one CT.

Totally followed the problem and solution, but what bothers me it what the panel wiring looks like. it looks like it was designed and constructed by a two year old, complete with a wooden back board. no wonder the toilets flush backwards.

To add to the confusion, both systems have peak production at different times of day due to roof angle.

@EEVblog have you heard about or checked out the Iotawatt? Fantastic little open source device that gives you graphs like these but with two major differences: First it can monitor up to 14 separate circuits with a single device so you can monitor every single circuit in the house independently (with 5 second resolution logged too!), and two it avoids all the "cloud" infrastructure - your data is stored on the Iotawatt only. I've had mine for over a year and the detail is incredible

I personally would have done it differently. I would have installed the second current clamp from the first meter on the output of the second system, just reversed that the current is subtracted. That way each system would still be measuring only the power produced from that particular system.

You can run more than 1 wire through a current transformer. The sensed current will be the *sum of currents* in each wire, if the currents are all in-phase with each other. The sensed current will be the *difference of currents* in each wire, if the currents are 180 degrees out-of-phase with each other. So for single-phase systems you can add and subtract at will, to prevent the kind of errors discussed in this video, and usually correct them without need of any additional current transformer(s). Also you can increase the sensitivity of a given sensing range by puting the same wire through the current transformer more than once, in the same direction. For example, if the wire goes through 3 times, just divide the indicated current by 3 to get the true current. This technique is very useful for measuring very small currents.

7:12 Do we do anything to prevent current flowing into the solar setup from the grid? Sorry if that's a stupid question!

not a panel board or din rail system, but a real panel board....... wow. learn something new every day.

positive video. no negative feedback from me. I have lost all my remote settings for solar monitoring but not worried the unit is working well.

So this mainly seems like an Australian issue due to the outdoor box o wires. Here in the US unless the wiring was done between 1880 and 1940, you would generally have a meter base on the outside of the house and then have a service entrance panel in the basement or sometimes attic with common busbars for all the circuits in the house. If you install a transfer switch or have multiple buildings on the property you would often have an outside service distribution panel or transfer switch between the meter base and the service entrance panel in each building on the property. While you may have subpanels after the main service entrance panel, they are all fed from the service entrance panel upstream so it is very easy to either add in a Service distribution panel to connect the solar to or have it on its own small breaker panel and wired into a commoning bar in the transfer switch so the load in the house is on separate wires with enough room for separate current clamps on both the load side and the solar generation side.

19:16 that so cool you can see each panel and the low ones are shaded

Great video on current topics delivered with a sunny disposition!

There's a better solution: instead of adding the production of both systems and giving that to the production input for both systems you should add the production of the other system to the power from the grid. So with Pa and Pb the production systems A and B and with G the grid input you should send Pa to power input and G+Pb to house consumption for monitoring of A And Pb ... G+Pa... monitoring of B

Australia, One Phase? I am used to Two Phase with a 200Amp Main Breaker. And that is a breaker box? Rats nest.

Would a good solution be to put a small box upstream of the whole house which is fed from the grid and feeds the original house box? That gives you a clean place to put the solar connections and a clean place to put the clamps for monitoring house consumption.

The fix of combining the 2 monitors in parallel means that you can no longer track how each set of panels is performing. If you can export the data from each system it would be simple to make a spreadsheet that would combine the data from both and get you the house consumption data you want. Add the generation numbers from both systems together and subtract the power sent to the grid from one system (both should be reporting the same current going out), that should give you the current draw of the house. Other way would be to take the positive generation value from one system and add it to the negative value of the other system, the difference there should be the current draw of the house.

Couldn’t you use the parallel transformer trick to subtract production A from the consumption reading made by monitoring system B? Rather than combining production readings?

Emon pi open source energy monitor has logic to combat this. It's a perfect solution. I have a similar logical set up as yours and I can determine all energy components using only two CT's

Another solution would be to run the AC wire from the opposite system through each grid side current clamp. Any current from the opposite system would be going through the current clamp in both directions on it's way to the grid, thus cancelling itself.

If you put the third clamp intentionally backwards, and wired it in series instead of parallel with the second clamp, would it end up canceling out the current of the opposite inverter? I can see where that would be beneficial to see only the current produced by each inverter.

That switchbox rats nest really bothers me. It hurts my electrical engineer eyes.

Yout switchbox is a mess compared to ours (Netherlands)

With my Fronius Smart Meter, it's between the input and the output, it has 2 inputs and 1 output, 1 input is from the inverter, the other input is from the street, all the circuits have no choice but to go through that smart meter, which allows me to monitor the street power being imported, the power being exported, the power being generated, and the power being used by all the connected devices in the house. But my house was also built in 2014, and Western Power tends to be a bit more anal about certain things :P

Apparently, electrical wiring standards are sure different between countries! It seems a lot simpler with a single 240 V phase. US 240 V split-phase panels are much more organized and modular. In Japan, modern 200 V split-phase panels are also organized and modular, but they don't really involve the concept of a safety ground. Most wiring is two wires only. Thanks for showing your electric service panel, I learned something, in addition to the subject of solar power monitoring.

If you can fit both of the conductors that you want to monitor through the same CT clamp jaws, that the easiest and most accurate way for a summed measurement. You can indeed parallel CT's when well matched (same number of turns and specs), but you only need to do these things if the measurements are physically somewhat far apart or can't fit multiple conductors in the CT. There are also instances where placing the CT's in series is a better option, if burden is affecting the measurement accuracy.

can you hook an oscilloscope or spectrum analyzer to the panel? those micro inverters are dreaded amongst amateur radio operators as many of them cause terrible interference

but you could to measure the tow systems independently let the current flow thru the second parallel coil backwards and subtract the current that is produced by the other system when you put it in parallel with the Grid coil ??? or is there a fault in my thinking ? would maybe not be the most accurate system^^

As of 6:50, thinking about this...two solar systems, with monitoring systems that don't know about each other, and both say solar consumption drops when the solar production goes up. Because the two monitoring systems don't know about each other, all they know is how much power is produced by the system each monitors and which way the power goes (into the grid or into the house). When total power provided from both systems exceeds household consumption, each monitoring system sees only that the power from the solar system it monitors isn't being fully used in the home and records that. To eliminate this monitoring anomaly, you need either one monitoring system for both solar arrays, or a means to cross-connect the monitoring systems so that they both know about production and consumption from the other array as well.

17:51 if you parallel them up you won't be able to know what each array is making on their own

Instead of paralleling the currently clamps, couldn't you connect the Enphase current clamp to channel 3 of the Solar Analytics? That way you could still see production of both systems instead of the aggregate production of both systems.

For the love of god can PLEASE somebody send him a german style fuse board??!!!1 He already uses all these breakers and stuff for the DIN-Rail but why do you have to keep putting them each in their own housing. My eyes are bleeding! Just get a german style fuseboard, big ones have over 100 slots for breakers and what not. (these slots are called "Teilungseinheiten (TE)" in Germany) - Greetings from Germany 👍

Could you put a "supply" wire from one inverter through sensing transformer of another in reverse?

18:50 Was it not possible by inverting the parallel connection of the input and the respective other generation (measure twice per generator) to subtract the respective other generation from the network measurement in the measurement?

Why not putting 2 wires through the same CT Clamp - either then measuring both solar systems or sustracting the additional generation of the other PV System from the false grid current Amazing how this entire house runs only a 1 Phase system. Meanwhile in europe with 3 Phase systems in nearly all households, we would need 3 ct clamps for each contact point

It's even simpler than that, simply run the wires from the two solar system, together, thru the two "probe transformers" , am I wrong?

What a lot of malarkey. As you pointed out Dave, what about clamps at the border of the house to the grid? Make me wonder what the installers are doing or am I missing the point?

As the unit has three inputs with a software change it could be made to measure both individually and combine them in software.

Yeah that fuse panel is one heck of a rats nest. You should replace it next opportunity. In the UK we use a common neutral DIN rail mount neat & tidy. Get a battery soon as you can - way better than pushing back into the grid. Just not Tesla. Many people build their own 18650 battery system. Also if you still want to keep an eye on your aircon, Shelly make a wifi transformer meter shellystore.co.uk/product/shelly-em-50a

Both current transformers should have same ratios to parallel them.

You don't even need parallel transformers! For the 3kw system, on the consumption sensor, run the 5kw source wire BACKWARDS through it. It will sense the forward current, and then the negative of the same current to zero it out. Repeat for the 5kw consumption sensor and the 3kw source wire!

GAAAAAAAAHHHHHHHH! oh my god, if I ever left a panel looking like that I'd be fired on the spot, also if uh no WHEN you have an electrical fire I'll know why. When I go on service calls for intermittent or circuits not working I seriously question my choice of profession when I open a panel like that.

Instead of feeding the grid, how about installing your own energy storage... batteries of some sort? Is the net cost of consuming less from the grid lower than amortizing a battery investment?

really surprised your breaker box wasn't updated when the solar was first installed that layout is awful

C’mon Dave, time to get your main switch board updated. You’re making the UK and US look good. I see an old analog electricity meter and even a ceramic re-wireable fuse used in the 1950’s and everything is arranged like a Picasso painting. Or are they some kind of museum pieces :) In QLD they replaced my two analog meters with a smart meter after they did the solar installations (in 2017) which freed up a heap of space. I’ve got a typical 5kw solar system and twice the number of cct breakers (all in ONE row) and still have tons of room open in my switch board if required for future add-ons.

That is one messy fuse box. What about busbars? DIN rails?

Well, those switching boards are a mess. Used to see and work in European din rail mounted switch boards

If the current clamp is measuring AC current, shouldn't the orientation back/front not matter? This would only be an issue for a DC current clamp no? ( *EDIT* just learned about measuring U and I at the same time and using the phase shift to determine net AC current flow direction, if both instantaneous values are multiplied with each other and both are positive= flow in one direction, if negative aka waveforms 180 deg out of phase of each other, current flow in the other direction!)

Could you have added the third current clamp in parallel (but backwards) with the clamp measuring current from the grid instead of the feed from the other array? Then for each monitoring system, production from the other array would automatically be subtracted from consumption instead of added into the production? That would allow you to still monitor the production of each array independently?

And what if join two clamp meters on the consuption side? One on the old place and one on the second solar system OUT. Production will be mesured separate for each system, and consumption will be "electricly" calculated like Dave has explaind.

Is that a Faber-Castell TR1 Calculator/Slide Rule? You should do a video on that. Great Channel by the way.

Or add a CT on each load breaker, then parallel them for total consumption? Then keep one on the mains input for net metering?

The measurements were not wrong, it was just not presented in a way you would like. A negative consumption is equal to putting energy into the grid. If you connect both systems on the same phase, things may interfere. I tell you a story, a young man and his girl friend are getting married. She asks him: "if we are married, how will our financial status be?" The he answered: "we will have one million and five Dollars." The she said: "And I have one million Dollars, so together we will have two million and five Dollars." The the husband to be said: "No, I have added it all up already!" The real solution to your problem is, let every solar power inverter only give results about what that single inverter is doing. Nothing more and nothing less. I know people with more then one inverter and both only report about their own results, the App will add it up. There is a saying about messy measurements and untrustworthy results: garbage in is garbage out.

To sum it up, Aussie electric panels are a hot mess so it's hard for Dave to get good readings. Even without the pic of the panel, just the whiteboard drawing gets you to the concluding it's a mess.

What if you run two wires through the current clamps? Like, run the input current through positive-wise and then run the generated current from the *other* solar panel backwards through the current clamp. You would be left with the difference.

What if you have a current transformer reversed over the output from the solar analytics parallel to the enphase output current so it wil subtract the current from the other system, and the same setup on the other system

Would be easy to solve - alternatively - if there was a second phase to split out the separate generation sources and loads. Then envoy and SA would only be dealing with their respective output and load.

Since it's AC, reverse reactive power seems to be a problem! From the graphs, rather than double the negative consumption during the day, it looks like the systems are fighting during the day and producing little net surplus, flattening the middle of the curve. But then your inverters would heat madly...or w/ a certain split transformer. Does it add up better than that?

Not an EE, was expecting a swapped grounding contact erroneously adding up the voltages or something like that. Independently/double-counting the consumption just sounded too dumb on the end-user scale.

They should have an offset output/input so you run a wire from 1 box to the other, so it can say "this is what I am putting out" and then the other box can calculate the difference. That way you could chain them as well because each system would be subtracting the offset of the ones wired before it.

Can't you just put a second current transformer in parallel with the load sensor, but reversed on purpose, on the solar output of the respective other system? That would subtract its output from the total going to/coming from the grid. Btw. I'd claim that a switch box looking like that will get an electrician fired on the same day here in Austria (not Australia). As if you hadn't enough fires already.

It's hard to get the clamps wrong, my first thought would be that the panels are from two sets. I like that you can see the panels separated. It's a first in first out problem.

You could put 2 wires (from solar panel inverters) through one current transformer to get sum current. And save second transformer for air conditioner.

I have the same issue as I recently added some more solar panels with a new inverter. Good to know u found a solution. Btw where can I get a current clamp?

so that box is forty years old, guess that's not to bad with all the upgrades, can you show us a modern panel board?