Unreliable Grid CT readings

[pdw]

I've had a H1-6.0-E running happily for a couple of years, but in the last few weeks I've noticed some odd readings which seem be down to an issue with the Grid CT. An obvious example is here, during an overnight force charge window:

Initially grid consumption goes up to 6kW as expected, to match the charge power, but then drops to pretty much zero, leading to a corresponding bogus -6kW load power.

It seems to intermittently flip between a correct reading and zero, although I think it may be under-reading at times.

The system isn't logging any faults, although it does if I actually disconnect the CT completely.

Has anyone seen this type of fault?

[Dave Foster]

It does look very much like the CT clamp is mis-reporting, it tends to happen if the CT cable has been spliced at the connection point, the wires are often twisted together and over time that causes problems and it needs to be re-joined properly - can you see if there is a connection/splice in the meter box ?

[pdw]

Thanks, I'll investigate.

I believe that the CT clamp is actually inside in some trunking that's a bit of a pain to get to, but I'd be surprised if it had to be spliced as it's in the same room as the inverter.

[Dave Foster]

As a precaution it’s worth doing an inverter shutdown / restart to rule that out.
If the CT cable is direct the only other connection point is in the ct/meter connector underneath the inverter has anything need disturbed there?.

CT clamps can be affected by close proximity of other CT clamps so if there’s been any additions like an EV charger, having the clamps close together (less than 5cm) can cause problems.

[pdw]

I've finally had a chance to look at this properly, and can't find anything amiss with the Grid CT. The connections in the plug are fine, and nothing has changed where the CT is.

I'm increasingly thinking that this is a problem with the inverter not the CT clamp, as I'm struggling to see how a clamp could fail in this kind of intermittent way, or how it could fail without triggering an error on the inverter.

Interestingly the pattern I posted a graph of seems to be quite common - it gives a correct reading for a short time during the first part of the overnight charge window, but then seems to drop out and read zero for the rest of the day.