Replacing a Sunny Island on short backplate above OG E-panel

[Trukinbear]

Hi, looks like I've got to replace my SMA SI6048 Sunny Island.
It is mounted to the short backplate, above an OG E-panel, with a twin Classic mount to the side.

The Sunny Island was the first thing that went up on the backplate after it was mounted to the wall, followed by the E-panel and the two Classic 250's. Can I remove and replace the Sunny Island without taking this all apart - or do I have to do it in the reverse order of installation, removing the E-panel and Charge controllers first? Neither is fun, one is much less fun than the other.

As always any advice, insight, or comments are very appreciated.
Thanks,
Bear

[FNG]

It lifts up to come off so other than getting to the screws on the bottom (If they were installed) it should not be that bad. What I have done in the past is have a piece of 1x6 handy and rocked the inverter one way and had some one slide it in between the E-Panel and inverter and lifted it while they push it through. Or another option is an engine style hoist to lift the inverter.

It is curius to me SMA thinks the inverter was damaged by listening to the shunt. I cant see how that would be possible?

[Trukinbear]

This is my THEORY - It was severely short cycling my battery (at times). Treating it as though it was fully charged but falling flat on face when attempting to handle a surge. Would be putting 70-100A into the battery with my system humming along with a very light 8-12A load - I'd kick on the coffee pot, or the microwave, and it would attempt to surge, fall flat on face when the charge it was seeing was not really there (superficial?) throw a W309 overload on L1 at the transfer relay, abruptly shut off both the generator and itself leaving us in the dark. Luckily I have a very durable battery and it appears not to have suffered serious damage.  The Classics, WhizBang Jr, and the Sunny Island all thought the batteries were (in varying degrees) charged, but the Hydrovolt unit proved they were nothing of the sort. Now, it seems to be working - everybody agrees with the Hydrovolt that we're fully charged - but the relay still sounds like a klaxon when the generator runs so SMA wants to send it to Germany to find out why.

[Trukinbear]

The PV parts just came fully online about 7 or 8 months ago - we'd nearly gone a year with the SI cranking the generator at 30% SOC, and stopping it at 80% SOC 6-days a week, with a full charge on the 7th, and an equalization charge on the 30th. Once the PV came online it became obvious something was amiss. When the Hydrovolt came the Classics/Whizbang Jr said we were 140Ah beyond fully charged, and the Sunny Island said we were at 89% SOC... the Hydrovolt read between 1,100, and 1,120 on all 24 cells of the battery. Was still lots of sun but we couldn't get any of it to put a charge in the battery as the classics had gone into Float. Now, after running the generator for 11hrs straight we've done an equalization charge and was fully charged - and this morning it took 2 hours to replace the 90Ah we'd used since the sun went down yesterday. Hydrovolt reads 1,270 - 1,275 - the Classics say we're at 100% and the Sunny Island says we're at 98% - never have they read so closely.

[Trukinbear]

Hmm, pretty sure I put every screw in there was.
Tried the engine lift on the installation, wouldn't get close enough to the wall - but with only 140# or so it would be safe for me to extend the boom (somehow) to the length of the legs I'd think... that might get me close enough. Bottom screws? I guess I've got to refresh myself with the instructions I was following at the time - I'd hate to have to dismantle the whole thing, will make it really hard to use the generator while the Inverter is down. I'm sure I can wire up something, I always do.

[Trukinbear]

Continuation of this completely autodidactic layperson's (though I was raised by a TV broadcast pioneer and got enough RF radiation from playing inside a transmitter to battle cancer all my life) theory. I believe there was a conflict between the SI and the Classics, the SI would see the voltage of the Classics and think it was battery voltage, but when it put an actual load on the battery it was apparent it was a superficial voltage, the actual voltage of the battery would drop below 45V, the SI would not be able to provide the power necessary and would attempt to pass thru additional power from the generator kicking the W309 code and disconnecting the L1 relay, still unable to cover the load it shuts off which immediately shuts off the generator too. We're in the dark. Simular occurrence with solar being controlled by the SI, but the SI not fully understanding the readings it gets from the shunt thinks the battery is pretty well charged from the voltage it sees and stops the Classics from providing an additional charge. IDK probably all wrong, but that is what it looks like has been happening to me.

[Trukinbear]

And/or it saw the current from the classics twice and counted twice as many Ah in.

[Trukinbear]

And/or it saw the current from the classics twice and counted twice as many Ah in.

Yes, THIS one. It was told of the current from the individual Classics via the MODBUS/Canbus deal, and also saw current (the sum of the two classics) passing thru the shunt.

[Trukinbear]

Well, this looks like a major finger amputation hazard with a 142# unit residing 1-2mm above the knife-edge sides of the E-panel. Tell me what your idea was with the 2x6, please?

[Trukinbear]

Well, thanks for the encouragement and great ideas. I've removed every obstacle I can, have fashioned a lifting bar to thread thru the top, and have 3 big guys showing up here at 3 tomorrow. It doesn't appear that servicing this down the road was a design criterion. Looks good though. I'll let you know how it goes.

[Trukinbear]

The final countdown has begun.

[Trukinbear]

Whew, finally over and just in time as darkness falls. For the system shutdown, we used this opportunity to fully test the Rapid Shutdown System - it disconnected the solar panels, turned off the generator, turned off the Inverter/Charger, and disconnected everything from the 986Ah 48V battery well within the 20 seconds required. After I turned off the voice announcements it made noises that sounded like a zombie gasping for air every few minutes until everything leveled out - kinda disconcerting until you figure out where it's coming from.

Piece of cake, the 3/4" bar that someone in another forum suggested was the ticket.