SolarEdge inverter - optimizer communication

[dapiya]

How often does the inverter(SE5000A-US) talk to/coordinate with/give instructions to the optimizers(P300)s. I haven't been able to find the info on any SE publications. I'm curious how quickly the system will react to a shadow moving across different panels in my system. Thanks!

[ButchDeal]

How often does the inverter(SE5000A-US) talk to/coordinate with/give instructions to the optimizers(P300)s. I haven't been able to find the info on any SE publications. I'm curious how quickly the system will react to a shadow moving across different panels in my system. Thanks!

They react very quickly and indipendently to shadows. The logging is different and more slow (minutes).

http://www.solaredge.us/files/pdfs/moving_forward_to_module_level_power_optimization.pdf

[dapiya]

A solid response that quickly?!!? I've been trying to find this info for a while... Thanks Butch!!!

[inetdog]

Basically the inverter tells the optimizers what current they should try to produce, based on the overall output of the array at the time. In between instructions from the inverter each module will try to maintain that current, dropping its voltage as necessary if a shadow moves across its panel.
When the inverter runs its next command cycle it might then decide to tell all of the optimizers to reduce their current slightly to keep the string voltage at its most efficient operating point.
I do not know how frequently the inverter sends new commands to the modules.

Unfortunately the PDF that Butch linked deals with the overall results on the system output and not the details of the independent behavior of the modules and the way they are frequently, but not instantly, reconfigured by the inverter.

[sensij]

Basically the inverter tells the optimizers what current they should try to produce, based on the overall output of the array at the time. In between instructions from the inverter each module will try to maintain that current, dropping its voltage as necessary if a shadow moves across its panel.
When the inverter runs its next command cycle it might then decide to tell all of the optimizers to reduce their current slightly to keep the string voltage at its most efficient operating point.
I do not know how frequently the inverter sends new commands to the modules.

Unfortunately the PDF that Butch linked deals with the overall results on the system output and not the details of the independent behavior of the modules and the way they are frequently, but not instantly, reconfigured by the inverter.

With some certainty, I can say that the communication in which the optimizers report their status is one-way, and slow... they are not individually polled by the inverter. See patent US20080147335.

I do not think the inverter addresses individual optimizers. The inverter has no knowledge of the string architecture (number of strings, or number of optimizers in a string), or the power capability of each panel attached to the optimizer. There are other ways that the inverter could control the optimizers without discrete messaging, and it surely happens faster than the rate at which the optimizers are reporting. The output power displayed on the inverter's local display updates quickly... that is probably the fastest that the signal to the optimizers would be adjusted.

In another of their patents, it talks about a shunt regulator in the inverter to hold the DC input at the desired voltage. By monitoring the current through the shunt, it follows that the inverter could be controlling the current input to maintain the target voltage. Optimizer data from a system with two strings of unequal length would probably be revealing.