AC Coupled SMA Sunny Boy & Solis Hybrid - Power Oscillation

[MrPablo]

A friend has just installed a Solis RHI-3.6 alongside a battery bank using cells from Colin.
Both are up and running fine, but there's a bit of an issue when the existing SMA Sunny Boy comes online each day.

At the moment, both are AC coupled (the Sunny Boy is part of a FIT system elsewhere in the house) and each morning when the Sunny Boy starts generating a power oscillation seems to start up.
Basically, the Solis sees the export, starts charging the battery and within 1-2s the output from the Sunny Boy drops, so the battery stops charging. Export ramps up again and the cycle carries on unless the charge rate of the battery is capped to 3A or so, then slowly ramped up manually.

Has anyone on this forum got any ideas how this could be resolved?
I've mentioned getting the FIT meter changed to a bi-drectional and running the PV into the Solis hybrid inverter, but if there's an easier fix...

[nowty]

I not sure if your quite describing it correctly as the Sunny Boy output should not drop as its a dumb grid tie inverter, whatever is happening downstream of the sunny boy should be irrelevant to it.

Its the Solis which should be doing the power limiting once export (to the grid) goes over 3.68kW and that includes from the Sunny Boy so it sounds like works as designed. As the Sunny Boy starts up it will cause the Solis to do something to compensate, charge batteries or reduce solar PV output from itself.

There is a possibility that the Sunny Boy is disconnecting and reconnecting if the voltage rises too much which is a risk the more generating capacity you add. If thats the issue, increase the cable size.

[Marcus]

If the sunny boy is dropping out it should put some sort of error up: a.c. undervoltage or grid disturbance or something?

[MrPablo]

Nowty, you're quite right when it comes to the description of the issue, I've been able to get further info and the Sunny Boy is just sitting there doing its thing based on the owner's Sunny Beam display.
It's a bit of an awkward one in that I'm unlikely to get access to the Sunny Boy in order to check directly, so I'm working on second hand information from the owner at this point.

I've managed to set up remote monitoring of the battery BMS, and should have the same for the Solis later this week, hopefully allowing for some additional detail.

The Solis is currently running battery only, PV expansion is planned for later this year.
The oscillation is starting even when there's just 50-150w of PV being generated, so I'm thinking it's unlikely to be a voltage rise issue with such low current.

I'm starting to lean towards a seesaw effect due to the Solis using an Acrel meter that updates every 1s.
In the example of 100w export:
- Acrel meter records 100w export
- Solis starts battery charging, circa 100w
- Sunny Boy output drops slightly (to 50w for example) due to cloud, etc at the same time charging starts
- Acrel meter records 50w import, instead of 0w as hoped for.
- Solis stops battery charging, starts discharge of 50w to compensate
- Acrel meter records 100w export (50w Solis battery, 50w Sunny Boy)
- Cycle repeats

A manual slow ramp up of charge current 'fixes' the problem so I think it supports the above theory.

I could do some control logic with the microcontrollers that will be attached to the BMS and Solis to automatically soft start charging when export reaches a set threshold (200w for example), but that's extra work I was hoping to avoid.

[Colin Deng]

If it the setting of the BMS not correct ?
Or if there is any loose on the connection of the cells ?

If there is the quality problem , please let us know
Cheers
Colin

[MrPablo]

Hi Colin, no concerns with the battery cells or BMS, that side is working well.

I've found that the Solis hybrid inverter can actually have two meters attached, one for grid, one for external PV, as long as they're both Eastron meters.
They've been ordered and as soon as they're wired in, the problem should be solved. I'll update here either way in case a future visitor has the same issue.

[MrPablo]

A quick update on this, managed to resolve it by replacing the Acrel meter that came with the Solis.
The system now has two Eastron meters:
- SDM120CTM, used to measure grid energy using a CT clamp on the main live tail out of the meter.
- SDM230M, with the existing SMA live and neutral AC wired through it.

The two meters are Modbus, so to get it working with the Solis inverter, I had to:
- Wire the SDM120CTM to the inverter as per this diagram - https://s3.amazonaws.com/cdn.freshdesk. ... 1658490752
That means Brown and Green going to terminal A, Brown/White and Green/White going to B on the SDM120CTM.
- Connect the two meters together via Modbus, so terminal A to A, B to B.
- Set the Modbus address on the SDM230M to 2.
- Set the meter type to Eastron 1PH on the inverter.
- Set meter location to Grid + PV meter on the inverter.

As soon as that was done, the oscillation effect has disappeared and it's running as if the PV input is direct into the Solis.
Good to get it running, especially as there's 2 MPPT inputs remaining on the Solis for system expansion.

Useful info in case anyone else suffers the same issue.

[Fintray]

Good to hear that you solved the problem.