Featur request: `Total PV Power` about sungrow-shx-inverter-modbus-home-assistant

[Louisbertelsmann]

The problem with this solution would be, that the MPPT currents aren‘t really accurate. They don‘t get reported by the inverter correctly. Iit also outputs as accurate as 100mA. For example, in an extreme case the current of MPPT2 could be 0.55A, that would be converted to 0.60A. That‘s an extremely big difference. That would bring your „Total PV Power“ down to 631W. It‘s still a big loss, but not as vig. Maybe there is some undocumented sensor that puts out the real total Power, because MPPT1 Power and MPPT2 Power are calculated using the corresponding voltage and current. A friend of mine has got a SMA inverter and that actually outputs power of each MPPT. And as far as I know Sungrow and SMA are somewhat similar.

[Louisbertelsmann]

For example, my PV is starting to generate. Total DC Power is about 500W, the voltage of the two MPPTs is there, but the current isn‘t getting reported. It‘s always about half an hour late.

[elektrinis]

You are partly right. It turns out MPPT current is reported in 0.1A steps, which ends up in power steps of 50W (if MPPT voltage is around 500V), which is a lot. I double checked it in protocol documentation and yes, 0.1A.

However it is updating as usual, with every data packet (10sec in my case). See below.

For now I just created a helper to report the sum of MPPT1+MPPT2 power, and it is working fine.
Will report back later, but leave this open for now, in case someone else is interested as well.

[elektrinis]

Just did a quick comparison of total DC power with my helper that I named solar roof power.
I wonder why it differs so much. Any ideas?

[elektrinis]

To make my case stronger, just adding a screenshot from my another issue. In this case my other inverter is generating power (connected between smart meter and SH10RT). In this case my solar roof is covered with snow and therefore not generating any power, however another inverter (solar wall) is generating something. And this somehow ends up in SH10RT's DC power....
From this it is obvious that Total DC Power is just a derivative from something, just not sure what.

[mkaiser]

this is all very strange, indeed.

I am still very confused, which "power value" would be the correct power value.

Measuring DC is more accurate than AC. So The "Total DC" power could be in fact "AC generated" power.

[dylan09]

I think you have connected your HA instance to the master inverter? And you have configured master / slave information and settings correctly in iSolarCloud?

Probably your problem is related to how the two inverters are configured in iSolarcloud. There are different ways how to setup how the values for master and slave will be calculated correctly.
But if I remember correctly, there is no configuration which will be give you valid values in all cases.

[elektrinis]

No, I only have a single Sungrow SH10RT and, to be precise, 14 pcs of per-PV_module micro inverters. Those are in no way integrated with the Sungrow and they output AC directly, just as any other household device.

So SH10RT is seeing a mismatch between smart meter readings and what the inverter measures itself, and it panics.
It's an obvious bug, as even Sungrow marketing material shows one can connect 3rd party inverters in this way:

[dylan09]

I understand. Thought you have two Sungrow inverter.
Maybe a question for photovoltaikforum.de. Maybe there is someone with a similar configuration.

[Gnarfoz]

Indeed, this kind of setup requires configuring the Sungrow hybrid inverter in "retrofit mode". Probably too complicated to handle here. A Sungrow-specialized PV installer should be able to do this, it's described in Sungrow's training materials.
If you can read German, you can try this forum thread on this particular topic: https://www.photovoltaikforum.com/thread/211268-sungrow-sammelthread-kombination-hybrid-normaler-wr-retrofit-l%C3%B6sung-shxxrt-fremd/?pageNo=1
(Or maybe using a translation service.)

Also, going back to your initial post:

There's no way 125W are lost to heat in inverter?!

That's actually not an unrealistic number.