EMONTX Continuous Firmware, Grid Power repeatedly drops to zero

Do you know what your module’s temperature coefficent is?
What’s their rated output?

Nominal Operating Temperature 45 Degrees C
Temperature Coefficient of Pmax -0.40% / Degree C
Pmax 250W @ STC (AM1.5, Irradiance 1000W/m2, Cell Temperature 25 Degrees C) Tolerance + - 5W
Year 1 down to 97.5% then -0.73% per year
They’re 5 years old
On Power depreciation that puts my 5 year old 250W Panels at 236W, the orignal M215 / 50Hz Micros have an EN50530 (EU) efficiency of 95.4% so that takes my 236W down to 225W, so if the sun was overhead they are way down.
Assuming I’m reading this right the maths for 6 years are 223W @ 45C Panel Temp and then less 0.892W per degree.
This ignores the fact that over here the sun is usually stronger than STC, so most panels produce more than their rated power.

Gotta admit, I wouldn’t be too happy with those numbers myself.

So, I looked at today’s production for two of my Siliken 255s since they’ve nearly the same rating
as your modules.

First module I looked at:
However, the VI product is not 174 - as the M-215 indicates - but 209.09, assuming
(and we know what happens there) the voltage and current measurments are correct.
Either way, something isn’t right.

Second module:

and, like the example above, the indicated power is not correct. 26.32V x 8.03A yields 211.25W.

I ran the numbers for several others as well and found the indcated power level anywhere from 5 to 25
watts lower than the VI product. Again, assuming the V and I quantities are correct.

TBH, I’ve never placed much trust in the data from the micros given their spec’d accuracy of 5%

You didn’t specify the source of your power “numbers,” but if your results are anything like these two examples, you may be in for a surprise.

Very interesting!
My source is a calibrated EMONTX, and lately I have a script that scrapes the Envoy for the individual power outputs. The 2 broadly agree with cloudless skies although I’ve yet to realise what relationship they have. Since swapping 4 of the non performing modules for new 330W Panels the inverters flatline when the sun is out at 229W for the M215’s and 259W for the M250’s IIRC
Looking back it seems the only time the old modules perform properly is when the sun is directly overhead in late december, although I wonder if my EMONTX was being hopeful. Between now and the end of the year I’ll see if I can get the 2 monitoring systems to agree and will then see if the old panels are performing. The problem will be the radiation in December will be above the test conditions and the performance then isn’t documented or guaranteed for 25 years.

That’s what mine do when conditions are good.
Sure sounds like your module performance has gone down the drain.