OpenEVSE charge rate problem on solar divert

Integrations OpenEVSE
1

I have a problem, which is almost certainly of my own making, but I’m hoping that somebody more familiar with the inner workings of OpenEVSE might be able to put their finger on it.

I’m running quite old firmware which has it’s issues, but overall has been working fine. I don’t want to introduce another unknown variable by updating firmware until I’ve sorted this issue.

Recently charging became erratic and then stopped altogether, reporting “EV not connected”. Sometimes I could get it to charge by waggling the cable. So I isolated it and checked the cable with a meter, sure enough the pilot wire was disconnected, somewhere in the cable. I checked in the car plug and at the wallbox, but the fault was somewhere within the cable. So I ordered a replacement cable and fitted it.

I was pretty careful to feed the right wires through - live/neutral through the first ferrite (the one with the 4 orange wires) and live only through the second ferrite, with earth and pilot bypassing both. (Does the orientation of either ferrite matter? I think I got them the same way round as before, but cannot be certain). Does that sound correct?

The behaviour I have is as follows:

  • Reports EV connected correctly now with the new cable
  • Manual charging fine. I can set the max current anywhere between 6A and 32A, click start and it charges at the correct current. So presumably the pilot is connected ok and OpenEVSE is able to send the signal correctly to tell the car how much current to take.
  • Overnight timed charging works fine.
  • If I put it into ECO PV divert mode, it is picking up the available current from my EmonPi correctly, and working out the smoothed current correctly.
  • It starts charging when more than about 5A is available, as it should
  • But instead of limiting the charge current to the available amount, as it should, it always charges at the full 32A (or whatever I set the max current to) instead of regulating it to the available solar.
  • I haven’t changed any settings from before I changed the cable, and it was correctly regulating the charge current on PV divert before.

Can anyone suggest what I might have messed up that allows everything else to work correctly but just this one fault of always charging at full current when PV divert charging starts?

Any help greatly appreciated!

Thanks,

Pete

2

Can you share a screenshot of your self-production page e.g

unnamed
unnamed758×461 67.1 KB

3

Hi Glyn,

Thanks for looking at this for me.

As I mentioned I am on a pretty old version of the firmware, so it looks pretty different to your example. I should update it sometime, but right now, as it was working before I replaced the cable, I think that would just introduce additional variables.

Here is how it looks right now, although the sun has gone in so it is not trying to PV Divert right now:

image
image819×1057 42 KB

4

That looks good as far as I can see. Maybe try posting some more screenshots when PV is generating and your trying to use divert

5

As you suggested, here are some more screenshots as the sun is climbing in the sky this morning.

Firstly, at about 7:40am, the PV is now making a bit more than the house is using:

image
image960×683 52.6 KB

At this time, EVSE looks good, correctly picking up the solar from my EMONPi and working out available and smoothed current, and not charging the car yet as there is not enough:

image
image677×886 32.8 KB

I’ll post the next screenshots as soon as the sun is enough to charge the car.

6

So now we see the problem. The sun is now strong enough to generate spare PV to charge the car.

image
image682×828 30.4 KB

As you can see, it is correctly working out the available solar with just over 6A available, and it starts charging the car. But instead of charging at about the avaialble current (previously it would have charged at about 5.5A or so) it charges at the full 32A, resulting in drawing over 5KW from the grid.

The amount being drawn from the grid is shown correctly, but it is charging at the wrong rate. If I manually modify the current limit from 32A, it then does charge at the rate I have manually set.

image
image883×833 58 KB

You can see the other two 8KW spikes when the sun came out briefly earlier.

So something must have changed when I replaced the cable - either I have not fitted it correctly in some way, which I can’t work out because all the of the other functions that require control of the charging current work ok, or have I somehow disturbed something else?

7

This screenshot is showing +5.6kW which means 5.6kW if import. A positive value is import and a negative value is export.

I would recommend instead of using Grid Import/Export to change solar PV divert to be based on solar PV generation. You’ll need to change the MQTT topic for solar PV generation.

8

Hi Glyn, yes, 5.6KW is what is actually happening, it is showing that value correctly, but it is failing to limit the charge current to the 7A that is shown as available.

It is working out all the figures correctly using the settings and MQTT topic, which was originally set up as per the instructions. It is getting the correct import/export figures, and working out the correct available current (7A in the example above) but then failing to limit that current in the charger.

So if you are saying that the MQTT topic needs changing, why was it working perfectly until I changed the physical charge cable? I havn’t changed any settings or topic etc.

If I change the topic so it looks at solar generation alone, how would it know how much of that generation is available to charge the car and how much is being used by the rest of the house and not available?

9

An update on his one. Some good news.

I did try your suggestion, Glyn, of changing the topic that EVSE looks at to the solar generation instead of excess solar, but as I expected this made things worse. Then it started the charger when generation reached 6A, regardless of whether the house was already using that 6A (which often it was), and it sill did not regulate the charge to available current. So I put that back how it was (and always has been, back when it was working).

So for a few weeks I have been managing by manually setting the limit current to 6A during the day and 32A during the night - the former for solar charging and the latter for overnight timed charging.

This works, as the charger starts whenever there is 6A or more of excess solar available, and limits to 6A. However, it does not of course make use of any spare solar above 6A unless I happen to be monitoring and manually change the limit current. It also means that when I forget to change the limit back to 6A in the morning, the first time it gets more than 6A available, it starts charging at 32A, pulling up to 26A from the grid, until I remember to change it.

Then a couple of days ago, I had forgotten to change it back to 6A as happens from time to time, and lo and behold, it magically started working properly again, automatically limiting the charge current to the available excess solar.

I had not changed any setting, so I still have no idea why it wasn’t working properly and now it is. I wonder if some NV variable in the firmware got corrupted, and it has now managed to fix that as some confluence of conditions caused it to rewrite the corrupted variable?

Anyway, fingers crossed the problem does not recur, and it is happily working properly again now.

After a while of reliable running, I may bite the bullet and update to the latest stable firmware.