Hello, I will soon have a Solis 4G 6kW inverter installed at home as part of a Solar PV setup and am looking at buying the emoncms kit. When I stumbled across this curious article from Owl (who make a Solar PV monitoring device): https://theowl.zendesk.com/hc/en-gb/articles/205589735-Solis-Inverter-Compatibility In short they state Solis is not compatible with CT clamp measurements.
Anyone got any expert opinions on this? Any success or problem stories with emoncms and Solis please?
Thanks!
*"In our experience, the Solis and Samil ranges of inverters operate in a different manner to other inverters on the market. *
During the evening when the unit stops actively providing a positive generation output, it goes dormant. However, it then draws a high current (approximately 2 Amps as seen in the output statistics) with a very low voltage by adjusting the inverter’s power factor.
Clamp based monitors (all of our units being these) monitor current only - and so this will result in an artificially high reading in this dormant period; producing erroneous readings throughout the non generating evening period.
Unfortunately we are unable to absolutely overcome or mask the unique characteristics of these inverter models, and as such these will produce readings during the night with a higher generation value than that actually being generated within the property.
As noted earlier, all clamp-based monitoring systems will suffer from this issue - not only our own; therefore the only absolute solutions we are aware of would be either the installation of a wired-in solution (which often carries a hefty installation expense) or a pulse-counter system."
Since the Owl devices do not sense voltage, instead they use a preset fixed value e.g. 240 volts for the UK, 2 amps x 240 volts = 480 watts which is not correct for low power factor loads (or anything other than unity PF 1).
The same would be true of OEM devices IF an a.c. adapter isn’t used, but the emonPi and emonTx’s all have the ability to measure real power by monitoring voltage amplitude and phase relative to the current measurement. It is strongly advised, to the point of being “the norm” to sample voltage as well as current.
So long story, short. No this doesn’t ordinarily apply to OEM devices, a correct reading should be possible with CT’s.
Not specifically answering your question, but a few months ago I had a Solis 4G Mini inverter installed, connected to a secondary set of (2) panels on a barn roof. The inverter works fine, as did a CT sensor coupled to an emonTx.
However, having bought a wi-fi dongle for direct access to it, I could NEVER get it to connect properly to the truly dire Ginlong iOS app. I tried every which way, but never succeeded, so gave up which was a shame. If you have better luck than me, please do let us know.
(I used to use Owl on my main pv system, but moved across to Open Energy Monitor when I bought an EV car and a storage battery - it all became to complex for Owl to handle really.)
All that sounds like misleading marketing hype from Owl. In effect they’re admitting that as their monitor doesn’t measure true real power, they are also claiming that nobody else’s will work either. As Paul points out, we beg to differ.
I have a Solis 4G PV inverter (one of these: these
I have been monitoring it just fine with a CT on an emonTX (one of the older V2 ones). I have never noticed any odd behaviour overnight - the power typically reads zero or very close to it.
The inverter contains relays which I believe disconnect it totally from the AC grid when it goes dormant overnight.
These inverters basically work fine - although they don’t support much if anything in the way of access to their logging data except using the GinLong monitoring website. But as I say - you can measure their AC power perfectly well with an emonTx and a CT.
As I wrote above, they (or maybe just one person who is, shall we say, “lacking awareness”) are claiming everyone is in the same boat, when the reality is different. In truth, exactly the same situation arises with most if not all appliances that feature a timer - the supply to that will most likely be through a capacitive dropper, and the resulting current, even though it’s milliamps, will have a most lousy power factor and show as a few watts in Owl and others that operate under the same assumptions, when the real power is only a tiny fraction of what is shown.
Typically, with the emonTx or emonPi, an inverter will show a small consumption of a few watts overnight - I term this its “stay alive” power that keeps the display, data logging and data transmission (if it does that) going.