Envoy-S - Distance from PV Panel

[bryankloos]

Hey Guys,

I'm curious if I'd be better off installing my Envoy-S in the attic by the dedicated PV panel or 85' down the homerun towards the MSP?
My router is near the MSP, but I'm happy to run cat6 from the attic to the router.
What is the school of thought with distance from the microinverters and the Envoy-S?

Thanks,

Bryan

[sdold]

Mine is about that distance from the panels (about 80-90 feet). I had communication problems initially, even on an outlet right at the panel. I then moved the Envoy to an outlet on the sub-panel going to the panels. I still had no communication, so I put an RF choke in each hot lead between the sub-panel and the main panel, which effectively isolated the Panels/Envoy combination from the main panel (and whatever was causing the comm problem).

So in your case I'd first try the Envoy where you'd like to have it the most (wall of the garage, etc.). If it doesn't work there, and you still want it there, try removing any other devices in the house that use power line communication. If still no communication, you could try moving it to the attic if that would get it closer electrically to the panels. But being physically close to the panels probably won't help unless it shortens the wire route the communication signal takes to get to the Envoy, for example if the plug for the Envoy is fed from the sub panel. That would let you filter in the same way I did if the need arises. What matters is whether or not it is attenuated, or interfered with by something else in the house.

Edit: For some reason I could never get the carrier-current power line ethernet devices to work from my Envoy to my router. I had to install Cat-5 to make that part work.

Good Luck Bryan, let us know what you find, these things can be finicky. The good news is that in a couple of months the fun of checking it goes away and you won't need it anymore

[idnominal]

The enphase microinverter plc system is finicky in part because enphase does an extremely poor job in installation guidance with regard to the power line communication features. For example, there is even one enphase installer document that suggests direct connection of the PV strings directly to the main panel (without a dedicated PV panel). That means that if there are any signal problems, you are left filtering several to many of the main home branches.

IMHO, a separate PV panel should be a minimum enphase requirement, if for no other reason than, it provides the connection between the PV panel and the main panel where some filtering can be done, both to isolate the microinverter/Envoy system from the home as well as to substantially remove the 144 kHz plc from the home main panel.

In my install (2 14 S280 branches), using the enphase recommended version with two double 20A breakers in the lower two positions of a main 200A panel, I had almost all microinverters reporting five grey boxes (Zero signal strength!), and occasionally one or two inverters with one red block (signal strength 1, poor). The system still reported, but I was unsatisfied with operating at the lower edge of comms performance.

Now, I am running a PV panel under 50 ft from the roof junction box (just the way it ended up, nothing special about 50 ft). I chose the Square D (home line series) 12-Space (six doubles) 125-Amp Main Lug Convertible Load Center HOM1224L125PGCVP, $41 at Lowes. Standard U.S. single phase 240VAC with L1 / L2 / N / Gnd. That way, I could reuse the double 20 breakers for the enphase branches which were originally in my main panel. Also, it comes with two free single 20A breakers that I can use later in my main panel someday for future applications.

The PV panel is about 25 ft from the main panel. For the combined PV power, I exit the PV panel bus with a double 40A breaker into two sections of a 50A EPCOS / TDK filter (B84143A50R105) (about $150 from Mouser Electronics or Digikey) mounted against the back wall inside the top upper section of the interior of the PV panel (just fits partly under the front lip, to clear the liner for the breakers on the cover). I used two sections of a three phase filter, slightly overrated for margin at 50A (my max combined enphase S280 branch circuits output is about 35A). The line side of the EPCOS filter is connected to the PV S280 branches via the double 40A breaker in the PV panel, the load side of the filter feeding a double 40A breaker in the other main panel. Note that the white neutral wire does not go through the filter and is connected directly to the PV panel neutral bus bar. Only the L1 black / L2 red lines go through two of the filter sections. If the neutral wire was connected through a filter section, it would reduce the plc signal.

The EPCOS / TDK filter B84143A50R105 filter data sheet is easy to access from digikey as a pdf. The insertion loss graphs for this filter series show good alpha values at 144 kHz.

The envoy is mounted on a plywood board next to the PV panel and fed by a double 15A breaker in the PV panel. (The Envoy-S cannot go in a metal box, probably not related to the 144 kHz PLC, but because of WiFi.)

At power up with the separate filter isolated PV panel, I get almost all 28 S280s at five green boxes, occasionally one to four drop to four green boxes.

Also, about to change #12 lines from the roof to the PV panel to #8 (#10 would be fine according to the enphase white paper, TECHNICAL BRIEF, Calculating AC Line Voltage Drop for S280 Microinverters with Engage Cables, but I have an inexpensive source for #8). The #12 cables were running warm (and beyond the length recommended by enphase for a 1% drop beyond 38'), no safety or code issue, but I hate to waste even a single watt more than I have to!

(Later note, #10 cable from the enphase branch circuits to the PV panel -for my installation lengths- would have been ideal by the enphase white paper, going to #8 does not add any practical advantage, and certainly does not justify the extra work required from difficulty of cable pull, dressing the larger cable in, to terminating the larger wire.)

[ktran1]

My envoy-s was installed inside my house, that circuit belong to sub-panel and it monitor 21 panels fine

[idnominal]

Clearly the enphase system can "work" wired directly into the main panel, even with signal strength zero. In some installations, there might be good signal strength.

My opinion is that enphase should go for a more robust system design and installation plan (dedicated PV panel with a filter), where on power up, there is a solidly designed and installed system with good signal strength (3 - 5 green bars, all micros).

The new solar install should be reasonably independent of the existing building or home electrical system. Beyond good quality transmission of power to the grid, the systems should not interact. That means the home does not substantially effect the enphase power line communication (plc) and visa-versa. It is not that difficult or costly to achieve, especially when the project budget is tens of thousands of dollars.

Suppose a new car dealer said, okay we realize you are getting choppy acceleration. Just move your foot around on the gas pedal until you find a good place where the new car runs a little better? You did get to work this morning, right? they say.

Clearly enphase has talented engineers and designers. The system appears to be very elegant, the microinverters of reasonably high quality, and the reporting of the enphase manager (next step up from enlighten manager) is excellent (e.g. near real-time access to the engineering performance details of each microinverter). So, why not go for a more robust installation plan / policy for overall excellance?