The blue line is the power consumption, and red-line is the production power from Enphase Gateway.

You can see some strange behavior, Power consumption looks line a reflection of production.

High production power stage it is showing much deeper

 

it can be happening when Consumtion CT connection is no good, or not climped well at L1/L2 line side.

 

 

I just fixed the connection of Consumption CT, and the power consumption become normal, it is matched with actual value around 500W.

 

This is the ferrite core 30A/65A

 

Fair Rite 0475176451 for 65A

Fair Rite 0475164181 for 30A

 

Those ferrites are not high current model, so, need put a gap

Enphase has applied 0.065mm Kapton Tape one 4 sides, and it makes 0.26mm gap.

 

This is the reference table

 

65A model can support upto 72A with 70% of Bs

30A model can support upto 53A with 70% of Bs

 

I have used 0.1mm x 2 = 0.2mm tape to reduce the gap, and reduced current as 56A.

and, 2 pass widing to increase the impedance 4x, and reduce the staturation current as 28A.

using 4 ferrites, I have set 19Ohm @ 110kHz

Today, I have received Power Line Filters and Ferrites

 

RP125-30-0-S It is single stage power line filter, and there is no Y capacitor, It can avoid a capacitor loss. and, much less power loss while production

RP230-30-4.7-S it is double stage power line filter with 4.7nF Y capacitor. I have RP230-30-10-S at the combiner box. maybe too much to try.

 

B64290L82X87 TDK 50mm ferrite core, I have used B64290L0082X065. But, I have decided to test with this ferrite becuase some reasons

1) X065 have 8~9A staturation current, but, this one have 21-22A

2) X065 has 10uH, and X087 has 4.460uH, so, I need more ferrites to get a similar result

 

Fair-Rite 0475164281 it is ferrite clamp, it is really hard to find a effective ferrite which can help to filter 100~120kHz noise. this is the lower frequency model, and will try on C-CT for isolate it also (even it is not make sense)

 

 

 

 

 

 

 

 

I am logging INVENTORY.JSON to check the status of Micro-inverters in my system.

And, I found some flags are 'false' but  inverter is producing power while some flags are fail.

I am not sure the flag is correct, or wrong flag due to the noise.

 

 

Recently, some micro-inverters lost communcation with gateway, and, the lost flags are happening more frequenctly, and, the flag error is happening before lost communcation.

 

the lost communication is happening when the production level is higher than 220W mostly.

the communication is recovering when the production level is going under 200~220W.

farest inverter from the gateway is under 100ft,

and, only 6 inverters in a group,

the grid voltage is around 250V while lost communcation.

 

I have checked without any extra filter (such as ferrites) on the grid side and check the status in cloudy day,

and, I found so many flag error (but no communcation loss when production power is under 200W)

 

 

I have install 3 ferrites on L1, 2 ferrites on L2 on grid side, and monitor the status while cloudy day

no more flag error, so, it is an evidence my gateway is affected by an external noise.

 

 

Just problem is, the noise level is not high enough at the situation (Enphase PLC noise detection graph)

 

at the noisy environment, it is easy to reduce the noise signifcantly,

But, it is really hard to reduce the noise when quiet (under 50dBuV) environment.

 

 

1. Will getting worse

2. Not related with GRID - no need to check grid or profile

3. Just power on reset - at first power on reset, it is showing 'Grid Instability Error' because no grid profile

 

https://support.enphase.com/s/question/0D52G00004ZVmwjSAD/anyone-else-having-issue-with-some-of-their-iq7-microinverters-randomly-dropping-out-around-the-same-time-every-day

 

https://support.enphase.com/s/question/0D53m00007P62ncCAB/grid-instability-issue-anyone-have-similar-issue-

 

 

I have requested a detail information abut LCF-064-1P,

     Enphase - datasheet only

     Manufacturer - can not provide it because customized item

 

No more solution with LCF, I have decided to return the filter on Enphase,

I am not sure the part is no good, or filtering is not enough. 

 

So, I start to find the solution on my way,

 

I have reinstalled RP230-30-10-S and GTX-2300

 

    BUT, it has too high line loss, so, I have disassembled & checked GTX-2300.

    the coil configuration is the two-in-hand winding, it is good for common mode filtering, But, it has lower magnetic 

leakage, so,not good for differential mode filtering. it is not much useful for me.

    so, I have removed it. and, reduce 0.5mA leakage current & 10mOhm Resistance Loss

 

    

 

        RP230-30-10-S have 60dB 50/50 insertion loss, but, can not block

the signal from next door also, I guess, my noise level at my home is too quiet and no one use PLC devices.

Yes, high SNR, so, just Envoy signal can be detected too easily.

 

      so, I have decided to add an extra differential mode filter using ferrite

 

 

Ferrite core is B64290L0082X065

 

and, look the wire direction, it is the coupled differential mode, and 2 turns.

over one day, it can isolated successfuly, the expected core loss is 200-300Wh per year. 

it is much less than 1~2kWh in GTX-2300-Y02. 

 

This configuration will be saturated under 10A. so, I have tried on single turn. 

at least, I need 20A because higher current will act as a noise, and will reduce the sensitivity of enovy.

and, I just need this inductor is working at the lower current (quite) zone.

 

 

This is final configuration.

It is working well, now, my system has been isolated completely.

and, no more cross-domain-traffic issue.

 

The cost of Ferrite & RP230 filter is just $80, and under $100 include tax & shipping.

It will working when system current is under 24A (around 20, IQ7+),

If want to apply on the bigger system, need bigger filter

 

 

 

 

 

I have installed shortest wiring, and no overlap to avoid a crosstalk, and just put one Ferrite core to filter a common mode for wire organizing and 10cm overcome earth wires. It may affect on the high frequency, but, it is not our concern because LCF is for 110kHz L-to-L PLC & Noise filtering now.
 

Let me check what will happening, 

 

Updated on NOV 29 2021

 

Long story is short, It is not working

 

This is an error what I do not want to get..

I have requested some advice and, Enphase said

 

Avoid a magnetic PLC coupling by

  - Own enclouser :: it is related with a distance, but, -.-a

  - 12" distance between conductor  :: point to point is not important, parallel run is the key

 

Anyway, I have installed as

 

 

If I add an enclouser for the filter, I have to get a permit & inspection from local office. it become so complicated issue. and, I feel, so strange, the engineer do not understand what is the real problem.

 

anyway, I am not sure what is the cause why it can not block the signal from my next door, 

i understand as it is a customized part for IQ. Maybe it is just a issue of my part. 

 

while low current, Enovy is detecting the other enovy at outside.

IT is happening because my house has very quite in electric noise.

if not cut-out completely, the filter is not enough because still high SNR. 

 

 

Very good video what is difference between Mico and String Inverters

 

 

Micro vs String Inverter

 

 

You have to watch it before install a solar system.

 

Enphase MI system is much expensive than String inverter,

even more than Optimizer system.

 

But, String inverter system need 6+ panels in series, it means there is a minimum number of panels under similar sun-light condition. but, MI system do not have the minimum requirments, technically 1 panel.

 

so, if have small zone & multi zones (more than 2), MI should be better

except this condition, There is no advantage via the extra cost

 

https://www.youtube.com/watch?v=UQ9Szhl1ceQ

 

Have to watch this video ~ Really ~ Really important

 

Most solar users are fooled by the micro inverter company,

Their advertizing is a fake ~

 

Typical FAKE INFORMATION

 

 

String Inverter

The pannel has connected in series. when some panel has shaded, it will affect on VOLTAGE not current. so, and, power will be by-passed on the the shade panel, total power output will be 100-100-100-50, the drawing for string inverter is 100% fake, or they just ignore the role of by-pass dioe, 

 

Micro Inverter

when partial shading, it will reduce the voltage not current, and there are 3 or 4 sections in a panel, and each section has by-passed using a diode. if the panel has 3 sections (60-72 cell), voltage will drop to 2/3 level. Even shade a SINGLE cell. if 36~39V system voltage drop to 2/3, the votlage output with the same current will be 24~26V. UNFOTUNATELY, most micro inverter need higher than 27V, effectively 29~30V. So, Even 1 CELL has shaded, MI will just output 5-10% power which has generated by the diffusion light.  so, 100-100-100-10 are the correct otuput in the drawing.

 

NEVER FOOLED, Practically, String inverter will ouptut more than 99% of MI system, even better long term.

 

If someone ask, when MI system is better ?

 

I want to say when have small sections. you need at least 6+ panels for start-up and 8~12 panels for the best efficiency in a section (similar light condition). and, the most string inverter has 2~4 MPPT line. if there the section size is less than 6, can not use the section in case of string inverter. and, if the section number is more than 4, need multiple inverters. example,

BUT, micro inverter system free from this requirments.

 

It is the real advantage of micro-inverter system,

Any other explanations are 99% BS

We knew Enphase Envoy and Inverters are communicating using Power Line Carrier,

 

The Envoy IQ is using Line-to-Line Singal instead Envoy-S's Line-to-N/G Signal.

 

It is a huge difference, because Envoy IQ is generating Differential mode signal

and, OLD Envoy is generating Common mode signal

 

To filter the signal from my neighborhood (he has installed Solar System last summer, unfotunately, he & I under the same utility tranformer). To solve the cross domain traffic issue, I need to filter the differential mode signal.

 

I have installed RP230-30-10-S between IQ combiner busbar And Wire_to_load_center.

It is suggested for Envoy-S (Line-to-N/G) system. but, Double stage filter has a good filtering on DM signal. Especially, this model shows 60dB insertion loss on DM. so, I have expected, it will filter the signal.

 

Insertion Loss (dB) = 20 * log10 ( Unfiltered / Filtered)

60dB means, 20 * 3 = 1000/1, the signal should be filtered by 1/1000 at 150kHz

Even consider some decay at 100kHz, the filtering performance should be 1/100 or better.

Leakage current of this filter is 1.8mA, it is pretty high loss

 

https://www4.enphase.com/sites/default/files/Power_Line_Filter_Single-Phase_TechBrief.pdf?utm_source=community&utm_medium=answer_links&utm_campaign=internal_reference

 

 

 

but, envoy is showing ‘Communications from other Envoy(s) have been detected’ early morning. And it is disappearing till evening, and showing again.

 

I have installed GTX-2300-Y02 between Wire_from_envoy and Breaker_Load_Center

 

Insersion loss is around 20dB in common mode. there is no differential value.

I am expecting around 10dB minimum, but, the leakage current is 0.5mA,

Too expensive to use it,

 

https://www.mouser.com/ProductDetail/KEMET/GTX-2300-Y02?qs=yqaQSyyJnNgbwvs18djAZg%3D%3D

 

 

 

Load center and combiner has located as below

 

 

 

MI - Combiner - RP230 - GTX2300 - Critical Load Center - XW 6848 - Main Load Center - Grid

 

Unfutunately, it can not solve the issue,

the message come again and again ~ like a die hard ....

 

Wait a minitue, just thinking again,

 

 

New Envoy is Line-to-Line communication, so, I have to remove/block a differential mode signal,

it means, the common mode filter is not working.

 

• Envoy-S    same as Common Mode Signal
• Envoy-IQ   same as Differential Signal

 

If install Ferrite Core on L1/L2, yes, it should not working -.-, that is just for old Enovy ONLY. !!!!

 

GTX-2300, it is a typical single stage common filter, it is good on common noise, and, have some filter for differential mode noise, i feel, that is not strong enough.

 

 

 

and, RP230-30-10, It is double stage filter, it have 60dB insertion loss on DM at 150kHz.

 

 

Unfotunately, those filters can not block the signal and, my envoy still detect the other Envoy(s). 

 

 

BUT, I feel so strange,

 

(1) it is detected when early morning  - when MIs start to work

(2) it is detected when evening - when MIs stop to work

(3) sometime at evening

 

so, I have tried to check the noise level and filtering level using SA

 

 

it makes me so disappointed, because I can not detect any signal from outside for couple of hours.

I have checked my device again, turn-on Enovy and check the signal

 

 

My equiment can detect a communication signal which has been generating from my Envoy.

I can find a signal, and confirm the filtering even I can not find the exact level of filtering.

 

[SO, THERE IS A BASIC QUESTION]

 

How my Envoy detect the other Envoy(s) signal ?

How my Envoy detect the other Envoy(s) signal after 2 filters ? there is 60dB insertion loss. Is it possible to be detected ?

 

ANYWAY, most power line filter is designed to minimize the power loss using small capacitor.

It means, not strong enogh on DM filtering, BUT, can not explain the the message while low power mode. Because filter will lost the performance while higher current mode,

 

I have ordered Q-LCF-064-1P-1 to make sure,

 

Because it said, IQ compatible,

 

75W maximum loss (via the maximum current), should be under 1/16 at my appliaction.

anyway, around ~19Wh is so high power consumption 

 

75W x 1/16 & 4.15h = 19Wh per day (or 7.1kWh per year)

Maybe need to try for the DM mode noise