Then I would test using the above diagram what are the last two pins for? (the ones outside the 485 pins). Assuming the pins are in some sort of order but you should have a positive electric signal on the first 2 pins and a negative signal of the next two pins either between each other or between each set of pins and Ground. You should be able to test with a multimeter. The signals on the pins may be too short to register much but should at least show up. from what i have read seems like in is either 5V or12V.
I measured with a multimeter there in the 1.3v region, there is neither 5v nor 12 for sure. Very strange, the manufacturer himself does not seem to know what he was producing. I bought another garbage ...
There's a (small) chance you may have an RS485, maybe in some form. What producer says I think is correct BUT if they expose the interface there must be a reason. I never used an RS485, I used RS232 for years, so I'm assuming stuff.
RS232 used two wires and a GND, it was 2,3 (RX and TX pins) and 5 (GND) on a DB9 connector.
RS485 from what I see (photo attached) only TX and RX pins, so two pins could be ok; but I see other connection schemas which also have GND so you might have to discover which GND is, and the test @floydR says should help you find it.
I found you a manual of the model similar to yours, it says the names of the software you could search for, here:
So, on a Friday night with ice cold beer I would try connecting that green port (first one and then the other, I see two on the photo) using a generic adapter (e.g. see attached photo, could be USB or a serial connector like the photo), to my computer. And on the computer I would run a terminal program trying to see if I receive anything.
I have used the software, I have used different variations many times and they are all without result. The only thing that the rightmost pin is GND. Before contacting the forum, I ate many times to solve the problem on my own, but this did not bring any results.
If you are brave you might open the charge controller up and take photo's might be that it isn't hooked up to anything. It obviously has knockouts for USB and RJ45 The only Rs-485 I have is a RJ45. perhaps the circuitry is there but not hooked up?
So, uhm, let me think... I would do this, you "2^3 combinations of order 2" (If I remember the correct definition!).
So, Pin4 is GND. Now, scientific approach to find RX & TX... although you make me think you may have already done this is to connect pins as follows, testing each case.
Unfortunately I don't know RS485 protocol, so I'll pretend it's RS232: I'd connect, one by one, the four combinations to my serial port on the PC, trying, brute forcing, all well known configurations for transmission. Be sure the MPPT charger is doing something and set the terminal on the PC:
and so on until I receive some bytes.
And before starting this test I'd be sure that: the three cables are good; the connector you're using works; the terminal software you use on the PC always works in the same way when you change the protocol parameters (for e.g. you could find a software which just blocks and stops responding when you play with the parameters, even if you set them right again).
Obviously if you know something more about the protocol first would be read good (I don't).
I had a quick look at RS485 specifications. You have 4 pins so it could be that your charger uses a "full-duplex" communication, so all 4 pins are used (RX- TX- / RX+ TX+). But by using that table I posted with the four combinations you should be able to receive at least one packet on the PC (the charger sends to the PC).
On the manual it says that default parameters are: 9600,8N1 so that could be a good start.
So how did you connect the wires from the charger to the "USB to RS485"?
I would try to use a RS485 Terminal program, set it to 9600,8N1, so you could quickly see if any message arrives (even if it's a messed up text):