Investigated inverter's panel meter communications

Found that all 8 wires (4 twisted pairs) had a signal.

And just probing with an ungrounded probe showed that these were 120v peak-to-peak:

Confirmed by switched to a 10x probe:

So my conclusion is that:
1) They are using a differential carrier on all 4 twisted pairs
2) I need to use a comparator to get a digital signal
3) My readings will be distorted by 50Hz 230v mains hum.
4) The timebase was  2 mSec/Div so period was 12 mSec / 80Hz

Interestingly the meter is packed with electronics:

And the 20A charge controller uses the base as a metal heatsink:

I'm hoping I may be able to upgrade a 10A controller to a 20A one myself but it doesn't look that straightforward.

2nd panel produces 7.2 A

Wired up second solar panel to a charge controller and 24V batteries.

Interestingly, this charge controller is rated at 20A (rather than 10A). Could this be why it reached 7.2A compared with 5.2A the first panel achieved yesterday through the 10A charge controller?

I must investigate before I buy the remaining three charge controllers.

Panel supplying 5.2A@36V (187W) and went off grid

Went off grid using 600W 24V pure sine wave inverter.

It can supply the extension (except for electric shower, washing machine and dryer).

And the main house (except cooker, kettle and fridge).

So I had it supplying lights, TV, satellite receiver, broadband router, etc for 2.5 hours (from 10 pm) from a 40 Ah 12v car battery and 90 Ah 12V leisure battery in series.

It was also supplying the extension from 12 noon to 8 pm.

It turns its fan on when it heats up. All worked better than expected.

Wired up Rasp Pi to monitor charge controller meter

I'm going to use the Raspberry Pi to replace the charge controller meter. (I can then monitor 5 charge controllers at once an dpost the result son the web).

But first I need to find out how the communications takes place over the RJ45 cable.

So I will monitor it with the Rasp Pi

The pin-out for the GPIO header and for the RJ45 I found on the web

I believe the Rasp Pi uses 3v3 voltage levels so I will need to buffer its inputs and if the RJ45 is using current mode rather than a voltage swing I may need to get a bit more crafty.

I'm also assuming a asynchronous protocol say 1 start, 8 data and 1 stop bit at 19K2 baud. But if they transfer in encrypted blocks it will be harder to decode.