diy solar

diy solar

EG4 new AIO rated 12K output and 18kPV aka "EG4-18Kpv-12LV"

I mentioned it because I have it, admittedly with cheap and nasty builder grade lamps and cheap Amazon dimmers. However, the dimmers work fine on good lamps from HD.
I also had flickering/buzzing with my Sol-Ark, but after replacing with higher quality lights, it has completely went away. Even with some cheaper dimmers.
 
Saw Will's initial review of the inverter today. Really encouraging. Hoping to see some testing results which can confirm if this inverter will have NGB concerns or if its all configured correctly.
 
Saw Will's initial review of the inverter today. Really encouraging. Hoping to see some testing results which can confirm if this inverter will have NGB concerns or if its all configured correctly.
It's a standard common neutral setup.
 
I did not notice till I did the screenshot, but Will has a white wire on the L1 of the load. Electrically that is fine, but that is not to code.... white is defined as the Grounded conductor (Neutral). (It is one of the few defined colors in the NEC). I am sure Will knows that. He probably didn't have any red or black (or any other color) of the correct gauge.
 
Screenshot from wills video:

View attachment 147542
All the neutrals go to one busbar - Clearly common-neutral. I have said for a long time that common-neutral is the best way to go for stationary systems.
I do wonder about a system like mine where the N-G bond is located 90 feet away from the inverters and load panels. There is some resistance to the bond that far away. I went to a common neutral and the light flicker became worse. I do see voltage on the neutral to ground at the loads panel less than 3V. When my lights are flickering under inverter power, I've measured 7V. I don't know the reason why I'm seeing the increase during the flicker. I went thru and checked the bond at the meter pole and ensured neutral and ground are tight in the house panels.
 
I do wonder about a system like mine where the N-G bond is located 90 feet away from the inverters and load panels. There is some resistance to the bond that far away. I went to a common neutral and the light flicker became worse. I do see voltage on the neutral to ground at the loads panel less than 3V. When my lights are flickering under inverter power, I've measured 7V. I don't know the reason why I'm seeing the increase during the flicker. I went thru and checked the bond at the meter pole and ensured neutral and ground are tight in the house panels.
Have you put a clamp meter on the ground conductor?
You shouldn't measure any voltage drop, if there's no load on the conductor.
 
Have you put a clamp meter on the ground conductor?
You shouldn't measure any voltage drop, if there's no load on the conductor.
Voltage drop tests should always be done with a load in place. Otherwise it would be a waste of time.

My clamp meter in the house doesn't measure that low.
 
Voltage drop tests should always be done with a load in place. Otherwise it would be a waste of time.

My clamp meter in the house doesn't measure that low.
If you have the length. Wrapping the conductor through the clamp multiple times, multiplies the reading. Or you can add a piece, just for testing.
 
I did not notice till I did the screenshot, but Will has a white wire on the L1 of the load. Electrically that is fine, but that is not to code.... white is defined as the Grounded conductor (Neutral). (It is one of the few defined colors in the NEC). I am sure Will knows that. He probably didn't have any red or black (or any other color) of the correct gauge.
Yes I ran out actually. I did label it with black tape. I didn't even have any red tape to label it.
 
If you have the length. Wrapping the conductor through the clamp multiple times, multiplies the reading. Or you can add a piece, just for testing.
(y) Yep, I remember that trick now. 10 turns= take reading and divide by 10.

I'll see what I find before I take these out, hopefully I can get a reading with light flicker too.
 
I do wonder about a system like mine where the N-G bond is located 90 feet away from the inverters and load panels.
That is a potential red flag.
I know why and where NEC want's NG bonds but it sometimes works against us.
 
I noticed you are supposed to register at
https://monitor.eg4electronics.com/ to get an account to view your data online but noticed that is still going through Alibaba:
11 eqix-sv8.alibaba.com (198.32.176.180) 63.834 ms 58.616 ms *
the registration is global, the server is north america
 
the registration is global, the server is north america
Please fix. Us should register in US. I don’t like ANY data, especially registration user/pass anywhere but in the US

EDIT: Looks like it may not be out of the US, but I hope EG4 considers this connection equivalent to the keys to a car.
 
Last edited:
Please fix. Us should register in US. I don’t like ANY data, especially registration user/pass anywhere but in the US
I'm not clear where exactly data is being stored/sent but that IP# I listed is in California. Here was the trace route (removed the first 5 which were from my internal network and local ISP):
6 100ge-3-2-mpls-mn.stellarllc.net (162.211.40.205) 14.309 ms 14.572 ms 13.319 ms
7 e0-7.core2.msp1.he.net (216.66.74.225) 14.105 ms 13.887 ms 13.411 ms
8 * * *
9 * 100ge0-54.core3.sjc2.he.net (184.105.64.69) 55.254 ms 55.129 ms
10 * port-channel7.core2.pao1.he.net (184.104.198.254) 59.570 ms *
11 eqix-sv8.alibaba.com (198.32.176.180) 63.834 ms 58.616 ms *

The 184.* and 198.* is in California.
 
Okay.. There is a DNS A record for monitor.eg4electronics.com that points it directly to 47.254.33.206 which is a server in the USA. An A record allows you to point even a subdomain to a specific ip#. So this looks okay.
 
Some of the geo ip lookups have it wrong. The "eqix-sv8" is the Equinix SV8 datacenter located at 529 Bryant Street, Palo Alto, CA.
Equinix could also be a connection pathway to other US or global endpoints, like AWS Direct Connect.
 

diy solar

diy solar
Back
Top