Our LEAF disrupts our PV system while charging at L2

[RegGuheert]

2 bars is low. Enphase recommends that you install the Envoy in a location where you receive 3 bars or higher. I initially had my Envoy installed in a location with 2 bars. Worked fine for months. But then started randomly dropping out. Relocated to where it gets 5 bars and have only had one instance where comms dropped out - coincidentally it was around 8am-10am one morning.

Great information! Thanks! Interestingly, after reading your post I just walked in there and did a communication check and it now shows three bars. That probably explains why it worked yesterday afternoon but not yesterday morning. (I would test charging it right now, but I let my son use the LEAF today since it was fully charged.)

I really doubt it's related to the amount of power the PV system is generating. In residential areas, the grid will be noisiest in the mornings and evenings when everyone is home and running their appliances. I think it's more likely that the additional noise generated by charging simply puts your system over the edge.

That actually makes a lot of sense to me! It is quite rural here and we are quite close to the end of a stub, so I'm sure the impedance of the grid is fairly high, meaning noisy loads should have some significant impact on voltage noise. Assuming the bars represent signal-to-noise ratio (which they should) rather than signal amplitude, my communications readings match your theory. (Of course, they also match the theory that the signal is now greater.)

If it were me, I'd do these 2 things:

1. Relocate the Envoy where it gets at least 3 bars during a comm check.
2. Install a ferrite bead on each line for the EVSE - the ones Herm linked to earlier cost $5 from Radio Shack.

I will do number 1 right now! That's super easy to do! Frankly, I had forgotten that the thing had a communications check until Wayne mentioned it since the Envoy has worked flawlessly for so long.

If I cannot find an existing outlet that gets me at least the three bars I need in the morning hours then I can certainly drop a dedicated circuit off the PV subpanel as mrradon suggested:

After trial and error, the only way I could get the Envoy-Inverter communication to work reliably was to install an outlet off my solar array sub panel and connect the Envoy directly to it, I had to run a CAT5 (internet) cable to the Envoy.

Try moving the location of the Envoy.
When I installed my sisters system, the Envoy didn't work either unless I plugged it into an outlet off the Sub-Panel.

 

[LKK]

Is it possible that all of your equipment is OK and the problem caused instead by the incoming grid voltage being out of tolerance for the inverters?

 

[RegGuheert]

Good news! I tried a few outlets in the same room where the Envoy was already located and I got 5 bars at the outlet right below the main breaker panel and four bars at the one next to it. Right now I have the Envoy plugged into the one with 4 bars and I will test that in the morning. Ultimately, the one with 5 bars could work out best. I will need to add an Ethernet plug nearby, but that should be pretty easy in that location.

Is it possible that all of your equipment is OK and the problem caused instead by the incoming grid voltage being out of tolerance for the inverters?

It's interesting that you bring this up, since I noticed on Wednesday when I installed the L6-20R outlet that the voltage was high in sunlight conditions at no load. Today after you posted I measured 252 VAC near the main panel with PV generation of about 7500 W. I had not noticed this before and I believe it may be the result of a recent upgrade the power company made to a 3-mile stretch of 1-phase distribution line to 3-phase. Anyway, I designed the PV system for a low drop from the inverters to the panels. I don't think it can get beyond about a 4-V drop for the worst-case inverter and condition. That means the inverters are all below about 256 VAC, which is below the high end of the datasheet range for the inverters which is 264 VAC. To date, I've never seen any inverters drop out.

But 252 VAC is above the upper voltage range of the charger in the LEAF. That could be an issue and I may need to get the power company to fix their wiring if I measure these high voltages when I am not producing PV electricity. (I do not know if I am within my rights as an authorized net metering customer to request the power company to fix the voltage for the case of full PV production. Anyone know?) 252 VAC is also above the upper end of the range for the EVSEUpgraded Panasonic unit, but I'm hoping there is some margin on the 250 VAC upper limit shown. Many universal power supplies are rated up to 259 VAC. I'm hopeful the one in the EVSEUpgrade is like that. In any case, it seems to be working so far. I'm not sure what the voltage range for the Envoy is, but if anything should be designed to work at a high input voltage, the Envoy is it!

At the end of the day, I cannot conclude that out-of-spec line voltage is my problem since the communication works properly in the middle of the day when the PV production is maximized (~8000 W) and it does not work when it is barely producing (~800 W).

But I do think there is one important point to consider here: I originally considered running the power for the EVSE from the PV subpanel. Given our voltage is already a couple of volts high at the main panel, it will be even a couple of volts beyond that at the PV subpanel. And the run to MY PV subpanel is overengineered. I expect some contractor-installed PV systems will have much higher drops than we are seeing here. The bottom line is that running wiring from a PV subpanel to an EVSE could result in a serious over voltage for the EVSE and for the charger in the EV. It's something to consider, anyway.

 

[Herm]

The output of the inverters is not adjustable?

 

[RegGuheert]

The output of the inverters is not adjustable?

Not by me. They have to adjust to whatever voltage they need to produce the peak power that the PV panels can put out. I just measured the voltage again with my house now consuming power and the voltage is still high at about 250.3 VAC. So it appears that the drop to my house including the 13 KV line, transformer, 240 V run and meter must look like about 65 mohm from my house since my 30 A production raises the voltage about 2 V.

I just feel like since I am legally allowed to produced up to 10 KW, I should be able to request that they get my no-load voltage below 248 VAC so that I will stay within specifications. They've been pretty reasonable to work with, so I'll give them a call and ask what they can do.

There are additional voltage rises within the house leading up to the solar array, so I need to take caution what I connect there, as well.

 

[Herm]

They can tap a different winding at the transformer.

 

[RegGuheert]

They can tap a different winding at the transformer.

Exactly. Either the transformer on the pole here or the new one they put in at the end of our road when they ran the new 3-phase line.

 

[DaveEV]

FWIW - each inverter logs the "grid" voltage at the inverter. You can compare these values for inverters in the same string in Enlighten (click on a single panel then use the graphs on the bottom of the page) to see what kind of voltage rise the last inverter in the string sees and also compare it to the voltage you are seeing in your main panel, too.

252V in the main panel is pretty high - the inverters are definitely seeing at least a couple volts more than that and may be getting pretty close to the point where they shut off due to over-voltage. If this does happen, you can call up Enphase and ask them to bump up the voltage limit - they can reprogram the inverters remotely for you.

It may be worth shutting measuring voltage in your main panel during peak production time and then shut your system off for a few minutes to compare.

 

[RegGuheert]

FWIW - each inverter logs the "grid" voltage at the inverter. You can compare these values for inverters in the same string in Enlighten (click on a single panel then use the graphs on the bottom of the page) to see what kind of voltage rise the last inverter in the string sees and also compare it to the voltage you are seeing in your main panel, too.

Cool! It's been a very long time since I have used the restricted site. I simply got tired of logging in. Anyway, as expected the end-of-run inverters stayed below 256 VAC all week except for at exactly noon on Tuesday four of them jumped up to 257.3 VAC. As you can see, we had scattered clouds on Tuesday and the power spiked up to 8 kW at noon. I suspect some edge-of-cloud effect was involved at that time.

252V in the main panel is pretty high - the inverters are definitely seeing at least a couple volts more than that and may be getting pretty close to the point where they shut off due to over-voltage. If this does happen, you can call up Enphase and ask them to bump up the voltage limit - they can reprogram the inverters remotely for you.

I believe the standard specification is 264 VAC and the extended one is 269 VAC. If I include the 2-V rise I see at the main panel, I would need to double the total peak drop to hit the standard limit, so I think I'm pretty safe. I wired the inverters as 6 strings of seven, with two of the pairs of strings each sharing a breaker and a 12 AWG run back to the subpanel. That's why the peak voltage is seen on four inverters instead of six, but is still relatively low. The 120-foot run back to the main panel is 2-2-2-4, so I expect to see less than 2-V rise there. According to Enphase's calcs, I would have seen an additional 3-V rise had I wired my panels in 3 strings of 14! That would have put me pretty close to the limits, particularly in the wintertime. (But I suspect many systems still get wired that way!).

BTW, Enphase's new system looks much nicer! Better wiring system and no clipping in wintertime! I loose about 3 kWh to clipping many days in the winter.

It may be worth shutting measuring voltage in your main panel during peak production time and then shut your system off for a few minutes to compare.

I may do that today.

Regarding the communication issue, the outlet that measured 4 bars yesterday afternoon now shows 3 bars this morning. That is consistent with the other outlet in that communication is worse in the morning than the afternoon. But alas, I have no LEAF to test with! My son did not return with it, yet.

 

[garygid]

269v RMS is a LOT for something designed for 208 to 240v, that will tolerate 250v.

Also, you might have voltage spikes at the tops and bottoms of the (normally) sine wave power.

Switching your PV system OFF for some testing is a good suggestion.

Measure the AC RMS voltage, and peak voltage (or waveform) if possible.

 

[RegGuheert]

269v RMS is a LOT for something designed for 208 to 240v, that will tolerate 250v.

It sure is, but the 269 VAC would exist only inside a microinverter designed for this voltage, and then only in a system which contained very large voltage rises. Nothing designed for 240 VAC would ever get exposed to 269 VAC. In our system, I am measuring 252 VAC near our main panel, so everything in our house is likely seeing this voltage. This is the concern, not any voltages at the inverters.

I have not yet connected anything to the subpanel near the solar array, but it is very tempting given its location right next to the garage. Unfortunately, that panel may see an additional 2 V beyond the main panel on a sunny day. I think I'll forego doing that until I can get my line voltage dropped down some by the power company.

Also, you might have voltage spikes at the tops and bottoms of the (normally) sine wave power.

That's certainly a possibility.

Switching your PV system OFF for some testing is a good suggestion.

Yes, but I cannot test the communication link with it off, since the communication that is broken is with the microinverters. Note that the PV system shuts itself down each night.

Measure the AC RMS voltage, and peak voltage (or waveform) if possible.

Unfortunately, I do not have an oscilloscope here. I've been measuring the voltage with a 20-year-old Fluke 77, which I do not believe is a true RMS meter. I just did a test at an unloaded outlet near my main panel comparing the Fluke 77 to a Brand Model 20-1850 and the Fluke gives 125.9 V while the Brand gives 124.8 V. Since nothing here has been calibrated in forever, it's a little hard to know what to believe. In any case, we are either right at or above the high end of the North American AC voltage range.

I guess I'm just glad we don't have a houseful of incandescent light bulbs!

 

[DaveEV]

269v RMS is a LOT for something designed for 208 to 240v, that will tolerate 250v.

The 240V Enphase microinverters are rated for 211-264V nominal - 206-269V extended. Quite bit more than the 256V RegGuheert's inverters are seeing, so no issues there. Not sure what kind of voltage spikes they can tolerate before being damaged.

 

[RegGuheert]

It's working! Our LEAF is currently charging while the solar array is generating power and all 42 microinverters are successfully communicating with the Envoy unit!

Here are some details of testing that I did this morning:

I started testing earlier this morning because I figure that right after the microinverters wake up is when communication is most difficult. I first tested with the Envoy plugged into the second-best outlet which had shown 4 bars in the middle do the day and three bars in the morning. Before starting the test around 6:30 AM, I confirmed that the Envoy was communicating with all 42 microinverters and the communication check showed 3 bars. I started charging the LEAF and the result was the same as previous testing in the morning: The Envoy lost communication with ALL 42 microinverters, even though the communication check continued to show 3 bars.

So I decided to try the outlet right below the main panel which has previously shown 4 bars in the morning and 5 bars during the middle of the day. I also decided to move the Envoy without removing the LEAF from the charger. After I plugged the Envoy into this outlet, it took a VERY long time to initialize. Once it finished booting, it displayed the last power level that it had measured at the other outlet and was not successful in communicating with any microinverters. So I started a communication check and it was not successful establishing communication with the microinverters even just for the test.

At this point I decided to take the LEAF off the charger and let the Envoy establish communications with the microinverters. After about 10 minutes it was communicating with all 42 microinverters both for power production and as part of the communication check. 5 bars were displayed. Please note that it was now about 7:15 AM and it was steadily getting brighter outside.

Then I plugged the LEAF back in and reinitiated charging. The communications check remained at 5 bars and 42 microinverters were shown in the test display. But the main display started gradually showing fewer microinverters: 41...39...38...34...33. But it did not go below 33 microinverters! It then gradually increased the number of microinverters back to 42. By about 7:45 AM all communication was restored even though about half of the PV panels were still in hard shadows from the house and many more were in partial shade. All of the PV panels were dew-covered at that time.

So, is the problem fully resolved? The situation is certainly much better! After 7:45 AM, all communication works fine. But will we loose any data before then if the LEAF is charging? How about during cloudy weather? I'm not sure. I guess I will compare the ramp-up of the array shown on Enlighten on similar mornings between charging and not charging.

I'm pretty sure I have some ferrite beads kicking around here somewhere. Now that I have some baseline data on what works and what doesn't I'll see if I can find them as well as some place to put them. Unfortunately the box in which I installed the L6-20 outlet for the EVSEupgrade is small and quit inaccessible, so I'm not sure it will be a good place to put them. I would prefer to put them by the charger in the car, as suggested, but I won't have time for such a project for a few weeks. We'll see what I can come up with...

One final note: For whatever reason, the voltage seen at the main panel in our house is lower this morning: 123.0 V was shown on the Fluke while the Brand showed 122.0 V. Perhaps there are fewer power plants running on Sunday morning...

 

[QueenBee]

Glad you got this working! I thought I'd just add my experiences to this thread.

The noisy LEAF charger that my car used to have didn't seem to have any negative impacts on any thing other than generating audible noise.

LEAF charging with and without the revised charger didn't seem to have any impact on my Envoys ability to communicate.

One issue I did have is with regards to AFCI breakers. I initially had an outlet wired near my main panel for the Envoy. The communications from the Envoy (Regardless of if the inverters were online or not) would cause the AFCI breakers in the main panel to nuisance trip. It turns out the issue was simply that the Envoy's signal was too strong being connected so close to the main panel. For example using the same outlet but with a 50 foot extension cord would resolve the Envoy issue. I ended up moving the Envoy directly to the PV sub panel and have not had a problem since.

 

[RegGuheert]

Well, it's doing it again. We went out shopping this morning and plugged in when we returned because I need to go out again this afternoon. After I plugged in I noticed some devices dropping out on the Enphase website. So I looked at the Envoy unit and it was showing only 2 devices. It has been going up and down between 1 and 4 devices over the last 30 minutes. I ran a communications check on the Envoy and it reports 5 bars and all 42 devices.

This is occurring at about noon on a cool day when the inverters are all near their maximum production, so I am really surprised to see this behavior.

It's not a big deal since I assume the data will not be lost, but it's too bad that the LEAF charger puts out so much garbage on the power lines.

 

[LEAFer]

My belief is that this is caused by the same problem reported in the Electrical line noise while charging at 240V thread, but I had understood that our LEAF was built after Nissan had already incorporated the fix for L2 conducted emissions.

I still agree with what you said in the OP

So, maybe the newer chargers in the LEAF are LESS noisy, but not good enough. Also use a Kill-A-Watt to check voltage being delivered by utility ... is there any serious fluctuation observed over a period 5-10 minutes ?

Invite a variety of (different) BEV friends and see what happens to your solar when they charge ?

 

[QueenBee]

My belief is that this is caused by the same problem reported in the Electrical line noise while charging at 240V thread, but I had understood that our LEAF was built after Nissan had already incorporated the fix for L2 conducted emissions.

I still agree with what you said in the OP

So, maybe the newer chargers in the LEAF are LESS noisy, but not good enough. Also use a Kill-A-Watt to check voltage being delivered by utility ... is there any serious fluctuation observed over a period 5-10 minutes ?

Invite a variety of (different) BEV friends and see what happens to your solar when they charge ?

Just as another datapoint I never had any issues with my inverters dropping offline before or after I got a new onboard charger.

RegGuheert, my car was suspected to be built be after the charger was updated as well but it still came with the old revision. Although considering that mine's 4000 cars earlier than yours I'd guess it would be a lot less likely that you'd not have the new revision.

 

[RegGuheert]

FWIW, I also see something weird going on with Tony William's Enphase system today. But I know he has a Rav4 EV now. Tony, were you charging your Rav4 EV until about 11:00 AM today? If so, I noticed that the website was reporting about 1kW of production at the same time the number at the top of the page was showing over 5W. I think the site goes back and corrects the power for each individual inverter and the totals for the day, but it does not seem to correct the power plot at the bottom of the page. So, no harm done for long-term totals but it does seem that EV charging is an issue for these Enphase systems.

One further thing I noticed about my system was that after I posted, the communication with the inverters was restored even though the charging continued at full L2 power. I now suspect that the issue is a function of the SOC of the battery and that the signature of the blocking signal from the LEAF changes in some way as the battery voltage changes.

 

[DaveEV]

FWIW, I also see something weird going on with Tony William's Enphase system today.

Looks to me that there's some sort of Enphase reporting issue. If you click on "Today's Energy" you can see that there are no drops in power.

My system has a few drops today, too. Never seen drops like that before due to clouds. Typically when there is a loss in PLC you see a flat-line for the affected time period as the data gets averaged out - not complete drops in power. Also, if you look at the logs on your envoy, if comms were dropping out you'd see error messages there. I don't see any for today.

 

[Weatherman]

Same here. Something screwy with Enphase's public systems web site graphs. I logged in to examine my own system more closely and the graph looks a lot better, but it was sputtering a bit, today.