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)