MPPT briefly enters fault state when PV input spikes

I have a SmartSolar 150/35 (FW v1.74) with 3x250W panels in parallel. During times of intermittent sun the MPPT sometimes enters a fault state when the sun first hits the panels. From what I understand I am within the MPPT specifications. I know I am above 130% over-panelling but these are old panels with heavy shading.

System voltage: 12V (12.8V Lifepo4 rated for 100A charge current)
Vmpp: 30.5V
Impp: 8.2A
Voc: 37.5V
Isc: 8.7A

(3 x 8.7A) x 1.25 buffer = 32.7A max PV current (MPPT specs say 40A short-circuit).
BUT… (3 x 250W)/12V nominal battery = 62.5A potential charger current when PV spikes?

I am monitoring the ve.direct port so I can see the current tracker mode, PV and charger current and error reason etc.

The MPPT tracker mode goes through the following states during a fault

  • Tracker initially in ‘mppt’ mode (when it is cloudy)
  • When bright sun hits the panels it briefly moves to ‘vi_limited’ (1.5 seconds) - I see a current spike on the charger of about 40-43A (charge limit is set to 35A)
  • Tracker enters fault state and turns off (9 seconds)
  • Tracker turns back on with ‘vi_limited’ (4 seconds), then ‘mppt’ (10 seconds), then happily remains in ‘vi_limited’ without issue

This does not happen every time. It seems like the charge controller cannot limit the charger current quickly enough, so it shuts the output down. There is no error code and the VictronConnect app changes too quickly for me to see or click on anything.

I should also add that the 150/35 MPPT is connected in a ve.smart network with a 100/30 MPPT (working fine) and smart shunt. At the end of the day when light is low, the 150/35 MPPT will turn on and off (off > starting_up > mppt > off) 50+ times until it finally shuts off, but the 100/30 MPPT doesn’t do this.

Any help is appreciated, thanks!

how a MPP limits the current? normaly by reducing the pulswith of the step down converter. But if there are shadows, there is another powerpeak. This new peak will be found by a scan and will be used. if the main peak recoveres because the cloud ist gone, the working point is located left of it. So the power increases by reducing the pulswith.