Solar power setup advice needed

I have a iotawatt+local emoncms service on RPi3 up and running. Now we are about to install solar panels on two separate roofs. There will be two separate DC inputs to one inverter, and then a connection to the 3phase 230VAC cabinet. I do not currently know how this will be connected, maybe someone here could explain?

If I want to do this, what hardware would be required, and how should I set it up?

  • (Preferably) measure power input separately from the two panels (I want to know the production from each, since they will be placed quite differently),
  • or, if not possible, measure combined power input to my grid
  • Measure ambient temperature
  • Locate thet hardware outside (the inverter will be outside), but dry under roof. May be exposed to temperatures -25 to +25 degC.

The IoTaWatt forum may be a better place to ask a question that is specifically about the operation and use of IoTaWatt. That is managed by Bob Lemaire, its creator, and he and the other users there are more likely to have the specific knowledge to help you.

Hi, sorry for the iotawatt confusion. Please disregard that. What I am asking is how to set up a separate system to measure power output from a new dual-solar panel setup + ambient temp.

Measuring DC is different to measuring AC, the OEM kit (and Iotawatt) are specifically geared up for AC, so no, you would not be able to measure the individual arrays without additional hardware of some form.

Could you provide more info about the inverter gtrid connection? a photo might help, as it is not clear if it’s a single or 3ph inverter. Both the OEM and Iotawatt kit have ability to measure 1ph and also 3ph (perhaps with additional setup or hardware depending on approach).

OEM kit can definitely do temp readings with the addition of a temp sensor (up to 6 on most devices)

As for installing outside under cover, the damp (dew/humity etc) may still pose a problem despite being under cover, the +25C would be fine, I’m not convinced -25C is ok though.

I see that the simplest solution here is to measure AC on the 230VAC fuse. But if there is some hardware I can add to measure the DC power input I would be interested to know. It seems like the emonpi comes in a solar bundle version, but it says AC sensors (lik the ones I have on my iotawatt), so I guess something else is needed…

About 1/3phase inverter connection - the main fuse in the house is 3phase, but the individual loops are of course 1phase. Currently I do not know what type of connection the inverter will have. I will need to check with the electrician.

This is the area of difficulty. Measuring the current at d.c. is easy enough - a Hall effect transducer will give you the necessary isolation. However, measuring the voltage is a lot harder, and there is no easy and inexpensive way to obtain the necessary isolation with reasonable accuracy.

If you can absolutely guarantee that no-one will ever make a direct metal-to-metal connection with the d.c. side, then powering the sensors and a radio transmitter from the d.c. panels themselves will provide all the isolation you need, but you would have to be very, very careful when testing and calibrating to ensure you didn’t create the conditions for a fault.

But you write:

The obvious question is, does the inverter know the separate powers from each panel? And are you able to extract that information if it does? This is something to ask the supplier.

Or take a look in the consumer box, the inverter breaker will be either a single or triple breaker.

See PZEM-017 DC Energy Monitor, i cannot speak for it’s quality or safety, I just noticed it and noted it here on the forum. It would involve interupting the DC lines to add a shunt and you would require 2, 1 for each array. But you should speak to an electrician if you are not 110% sure of what you are doing, the DC can be notably more dangerous than AC on domestic PV inverters.

I can help you with talking to this device, but the DC side is not typically DIY.


As Robert states it might be easier (and safer) to communicate with the inverter, what inverter make and model is it?


Just realised you are not in the UK, I’m not fluent in Norwegian electrical standards so there may be other options, a photo would probably let us work it out,

I have still not got the make and specs for the inverter. Plan A is to measure on the AC side, but I will check once I get specs on how/if I can communicate with the inverter and get some useful data directly.

Indeed it can. But, that particular issue lies with string inverters. It’s not uncommon for a string inverter DC input to be at 300 to 600 or even as much as 1000 Volts. For the benefit of others reading
this thread at some later date, I thought I’d mention that the DC voltage associated with microinverters
is the voltage of a single PV module, typically 36 to 48 Volts. A bit more with 96 cell modules, although
they aren’t as common as 60 and 72-cell modules.

I’m not saying it doesn’t exist, but the only scenario that I know of where an inverter “knows” the
output of an individual PV module uses microinverters. (one inverter per module)

No Bill, whilst your comments about voltage levels are correct, the context of my comment was mainly to do with the fact that the DC side tends not to be protected or have isolation at the source end.

Isolating the (AC) grid connection, the consumer units and/or the inverters will not isolate the DC side and it’s easy to think it is isolated when every possible switch is OFF. Unfortunately PV is rarely isolated at source and the cabling to the inverter(s) from the panel(s) is unprotected.

In the event of a short etc, you have to wait for something to burnout or for the sun to go down for the potential to reduce. Whilst there obviously a greater threat with higher voltages and currents, even a low voltage microinverter could be an issue. Even cars at 12v are an issue when unprotected circuits short, they easily cause fires.

I know a guy that caught a DC cable (from panel to inverter) that was run inside a double skinned timber wall with an alligator saw, the whole saw body was too hot to touch to pull out of the wall, they had to try and smash it out with a scaffold pole, by the time they had done that a sizeable fire had got a hold and the building was largely lost.

Albeit that was indeed a stringed PV setup, so high voltages were involved. But the same can happen even at lower voltages. I have repaired too many 12/24v car/truck wiring looms to underestimate even low voltage DC, no, it is unlikely to electrocute you, but it’s biggest problem is being underestimated because it’s low voltage.

It is easily over-looked that the PV panels are also a power source when thinking about isolation, just because the inverter(s) are isolated the panels may still be attached and live, that’s where the danger creeps in.

That’s perhaps why it’s an obvious question!

(You quote me, but I was quoting Robert). I think we both (Robert and I) understood the OP meant panel arrays, it is unlikely (but not impossible I guess) that he has just one panel attached to each of just 2 inverters. If you can communicate with the inverter(s) you may be able to report values for each string array, it is not likely that the output of each panel can be reported without additional equipment.

Not likely to happen with a microinverter and single PV module. You can short the output leads together on a PV module and it’ll run all day long in full sun with no problems. (during normal operation, a PV module runs at 90%+ of its rated Isc and 70 to 85% of its Voc)

Ref: power supply - Solar panel short circuit - Electrical Engineering Stack Exchange

PV module DC wiring (at least the wiring sold in the US) is rated to carry significantly more
current than the output of a single PV module.

No argument there. What with a current source capable of supplying 400+ Amps, that’s definitely
“a fire looking for a place to happen.”

To add just a little to the dangers of d.c: An a.c. arc tends to extinguish every half cycle, a d.c. one won’t, it will just keep going until, as you say, something else interrupts the current or the gap becomes big enough. D.C. circuit breakers usually have powerful magnets in the arc chutes to draw the arc out and so lengthen it.

Really great to see this discussion. Most of it is above my tech level, I admit, but still learning.

To clarify: We are currently installing two separate PV arrays (8 + 7 panels on two separate roofs). One DC cable from each array into one inverter. I have asked for the specs and if the DC circuits are independent or not, so far without any luck.

I fail to see how you can perceive a current source that will not falter as “no problem”!

This is precisely why I posted the warning about DC, it’s a perfect example of why things can go seriously wrong.

So we have a single PV panel that can happily kick out ~250watts of power into a dead short all day long, the fact that most PV panels will be most likely connected with suitably (over?) sized cable means the weak point in the circuit is quiet possibly the point of short, but even that’s not definite, the short could be able to carry the current and then the heat becomes the problem.

Based on your assurances of low voltages, low currents and micro inverters being “no problem” a user could disconnect their microinverter from the AC side and stuff their hand somewhere it shouldn’t be, if for example their wedding ring (for example) shorted out the DC, this would become a 250w heater element wrapped around their ring finger. Have you seen the damage a 30w soldering iron does to flesh in a few seconds, let alone 250W until the sun goes down.

Going back to my example involving the alligator saw, the panels were fine pumping power into a short and the cables were ok with transferring that energy, even the saw blade managed to withstand the current, but the whole tool got so hot, so fast, the guy couldn’t remove the tool, the blade got so hot it ignited the wooden walls around it, had he not smashed the saw out with a scaffold pole, the saw blade might well of hung in there until sundown. That IS the problem with an unprotected power source.

The act of installing protection like a fuse or circuit breaker is to build in a known weak point so that it becomes the point that burns out or trips, not the poor sole (or equipment) hanging off the the other end of the cables.

My original comment was purely to say that extra precaution needs to be observed with inverters as the DC side is more troublesome BECAUSE it is an uninterruptible power source with no protection or isolation (at source) AND because it is often misunderstood that low voltage/current is “safe”, It would only take a few watts of power delivered to the wrong place to destroy a RPi (or even an expensive laptop whilst commissioning a connected monitoring device, for example), it depends on what your concept of safe is I guess, I wasn’t concerned about damaging the PV panels or the wiring from them. I was concerned more about what else might be completing the circuit, even a quick but unexpected bang or flash could make someone fall off a ladder or smash their head against something.

I once had to go and inspect/assess the damage done to a brand new Saab that had a 10A fuse fitted (in error) to a 3A circuit when a radio was fitted and the wiring was crushed when the radio was pushed into the dash. The fuse didn’t blow, but the wiring rated for 3A got so hot over night due to the partial short that it set light to the floor mats and slowly smouldered away until the staff found a smoke filled burning car the following morning,

It doesn’t need to be a large current source or high voltage to be a problem, that’s the point.

Even the fact it’s power is derived from the sun cann be a problem, disconnect the DC cables and put them to one side at night when they are dead, when the sun comes up the following morning when your not about the DC cables become live “unexpectedly” (unplanned might be a better word) and start a fire whilst your out or in bed etc


I know what you’re trying to say, but there are some details here that might prove misleading to someone not well versed in electrical theory.

I think Bill was talking about what we call a “bolted short”. That is, one with a negligible resistance. He also said that the cables were rated above the short-circuit current of the panels. In that case, the power is dissipated partly in the cables and the remainder in the panels themselves. Because the cables are adequately rated, they won’t reach a dangerous temperature.

The real danger comes, as you say, when there is a resistance associated with the fault. In that case, there will definitely be power dissipated in the fault and heat generated, and that is what happened with the saw incident.

In the case of the car, the wire was rated suitably and would have been protected by the specified fuse, but wasn’t rated for the current required to rupture the 10 A fuse, and didn’t rupture it because the current was limited principally by the resistance of the wire itself. Here, the wire did reach a dangerous temperature.

I’ve got to emphasize that PV panels ARE dangerous, except possibly on a dark night - and even then, they are still potentially dangerous because the sun will come up.

Robert, I do understand the theory and why/how the examples I provided came about.

I’m not sure why I’m getting corrected here. 4 weeks ago i posted a piece of general advice, namely

Which I stand by, possibly even more so now.

For 4 weeks that comment has gone unchallenged until yesterday when Bill posted

Which is misleading in suggesting that these warnings about DC do not apply to lower voltage single PV panel microinverter arrangements which is not true for the reasons I have given.

If that is the case, not only is it totally out of context in this discussion but that should be explained and not left for “someone not well versed in electrical theory” to believe

So for the benefit of the less well versed in electrical theory,

If you are going to accidentally cause a short on a PV panel’s DC output without being aware of the danger, you MUST ensure it is a “bolted short” to avoid causing damage, injury or fires!



Without any caveats about microinverters not being dangerous that could cause false security and possibly lead to a dangerous situation.

All I was trying to say, especially for those not well versed in electrical theory was be careful and be aware that the DC side can hold surprises, making it potentially more dangerous. Which I firmly stand by as good advice! The rest is just defending that advice as it applies to ALL PV installs until you KNOW otherwise for sure.

I’d have said “to lessen the risk of damage, injury or fires”.

And that “advice”, if you can call it that, is generally true for any type of electricity, on the very important assumption that everything else had been done correctly to the proper standard and the protective device or devices are in place and working correctly.

I was actually being a little facetious there, how would someone not well versed in electrical theory know what a bolted short was? And even if they did know, how could anyone (well versed or otherwise) ensure an unplanned accidental short was of the bolted variety?

The original general advice of leave it alone and get an electrician in unless you know exactly what you are doing is far more suited to someone that is less than totally sure of what they are doing.

Many users are “ok” with AC as they understand they must disconnect it before touching, it is so easy for a novice diyer to think the inverter “is isolated” and therefore safe, overlooking the DC generation side of things, plus it is often assumed that low (DC) voltage/current is safe and that’s when it bites ya.

While that would be fairly easy to do with a string inverter (think DC combiner box here) it’d be difficult to do with a microinverter. (given as Robert mentioned above, the installation was done correctly. We know there are going to be installs that aren’t done to electrical code.) The DC side has no uninsulated connections, and MC4 connectors are of the locking type, and require a tool to unlock / disconnect them. And even with the tool, they’re a %(@!& to get apart. That’s not to say the wires can’t be pinched, cut, etc, but the leads on a PV module are not long, i.e. less than a metre. At the junction box where they leave the module, they are about 3 inches apart. So save for the point where they enter the microinverter, it’d take a deliberate act to get across both of them simultaneously. The insulation is much thicker than that found on wires used for non-PV purposes. Much tougher too.
Not saying it can’t be done, just that it’s much more difficult to do on a “micro” system than a string system. (I’ve had 16 micros on my roof and 26 in my back yard since 2013)

I’m not saying the DC side doesn’t present safety issues. It does. But with considerably less danger (as far as electrocution is concerned) than the voltage levels involved with string microinverters. Put another way, the odds of getting killed by the DC side of a microinverter at 48 Volts are considerably less than getting killed by the 300, to as much as 1000, Volts feeding a string invereter.

Again, that’s not saying the potential for fire is any less. Both are very capable of starting a fire.

Speaking of rings on fingers and the results, yes I have seen the results of bridging an aircraft 28 Volt
bus (rated at 400+ Amps) with a metal wristwatch band. The guy has a silhouette burned all the way
around his wrist. Not pretty. I can only imagine what it looked like the day it happened.
Even worse with the ring, but not related electrically, was a picture we were shown when I was in technical training with the USAF. It was one of a guy who jumped from an aircraft door and got his ring hung on a corner of the doorframe. The distance to the ground was farther than he estimated as well as greater than the span of his raised-hand-to-bottom-of-feet. That one was really bad.

Again, given a proper installation - in the US, both sides of the inverter are required to have disconnects. (isolators) located not more than 6 feet from the inverter. And again, we know there are going to be installations that aren’t up to par, with all the attendant safety issues. And of course electrical safety requirements vary wildly across the globe.

I have been following this with particular interest. I have 2 separate 1.5kw PV installations. They are both on shed roofs. Self installed using SMA Sunny Boy grid tie inverters. I followed the SMA instructions to the letter. The Sunny Boys have a built in DC isolator. The AC goes to a Consumer Unit which in both cases is within a metre of the inverter. The AC isolator is a circuit breaker in the Consumer Unit as indicated in the inverter installation manual and is of the type recommended by SMA. Everything is clearly labelled as required by the DNO complete with circuit diagrams. I did wonder at the time of installation, should I have fitted a separate AC isolator adjacent to the inverter. SMA say not necessary, their idea being that their inverters are easy and quick to fit.

The other thing that bothered me at the time was should I fit a ground/earth connection to the PV mounting rails and if so what should I connect it to at the ground end? I checked this with the PV installer who fitted our industrial unit with a 3 phase 10kw system and he said the earth is not a requirement. Even if it is not a requirement would it be safer to add ground connections?