diy solar

diy solar

Sunny Island SI-6048 BMS communication

I should hope there would be change left over.

It isn't in the same class at all.
Have any motors you need to start?
If I was you I'd want to figure out how to make the Sunny Island work.

I'd be quite happy to pick up 1x SI 6048US and 2x Classic 250 for the price of a DIY.

Cheap, but low or no stock.




The DIY line was just a relabel.
Midnight's "The One", 10kW AIO, may be more capable.
But if you want an Art Deco replacement for German Engineering, Rosie is the one to consider.

I would hope to have enough left over from selling my (2) SI6048, and (2) Classic 250 to buy another 200 Ah of batteries. However, I'd already come to the conclusion that the DIY wasn't up to task. The Rosie is outta my price range, but I love Midnite's tech support - they even speak English.
 
The SMA works with canbus set to pylontech, deye or goodwe protocol's based on my working setup, it's your battery that is not meeting these protocol requirements. So any solution must not expect to use canbus from the battery unless you have confirmation from another owner that it works with that combination.
Aolithium claims to speak those languages, and 4 others...
 
The claim is wrong then and the root of your problem, if you want to the replace the SMA so the 1he Aolithium speaks canbus to the replacement inverter you need to be 100% sure it will work. Based on the evidence so far it will not and you will be back to open loop with the new inverter,
 
The only lower cost battery packs that claims to communicate with SMA is EG4 through their add on coms box. No idea if the bugs alre all worked out.

Other than that, it’s high end SMA “partners.”

SOK almost cracked the code but basically gave up (something to do with SOC under 30%).

Most folks on here use REC or (I believe) Batrium BMS units for closed loop DIY batteries.

A few use VLRA and open loop.

The ability to broadcast in a certain language or protocol doesn’t meant the right info is being passed back and forth or if the BMS is smart enough to make the correct decisions. Sunny Islands in lithium mode hand almost all of the battery decisions to the BMS. That BMS needs the ability to make those decisions on behalf of the SI, and generic BMS units aren’t programmed with that ability in closed loop.

You could crack the batteries and install new BMS units but that defeats the purpose of buying off the shelf to a large degree.

This has been hashed out within this forum a bunch of times. I researched the piss out of it before buying REC BMS units (still had a few problems that were mostly sorted).

I would rather have the REC units though rather than a generic BMS that’s built into most lower cost premade batteries - with or without considering SMA issues. The big ass contactors that they use to turn off batteries in the event of a short gives me a nice warm and fuzzy - especially after the recent fire threads.

Most of the cheaper pre made battery packs are put together with bubble gum and alibi babba BMS units.
 
Last edited:
@Trukinbear

setup a new battery, and charge it. Do not discharge it at night. next day, let it charge again. after a while the soc on the sunny island should go to 100%
i think discharging at night on the first night is causing a lot of issues
 
The only lower cost battery packs that claims to communicate with SMA is EG4 through their add on coms box. No idea if the bugs alre all worked out.

Other than that, it’s high end SMA “partners.”

SOK almost cracked the code but basically gave up (something to do with SOC under 30%).

Most folks on here use REC or (I believe) Batrium BMS units for closed loop DIY batteries.

A few use VLRA and open loop.

The ability to broadcast in a certain language or protocol doesn’t meant the right info is being passed back and forth or if the BMS is smart enough to make the correct decisions. Sunny Islands in lithium mode hand almost all of the battery decisions to the BMS. That BMS needs the ability to make those decisions on behalf of the SI, and generic BMS units aren’t programmed with that ability in closed loop.

You could crack the batteries and install new BMS units but that defeats the purpose of buying off the shelf to a large degree.

This has been hashed out within this forum a bunch of times. I researched the piss out of it before buying REC BMS units (still had a few problems that were mostly sorted).

I would rather have the REC units though rather than a generic BMS that’s built into most lower cost premade batteries - with or without considering SMA issues. The big ass contactors that they use to turn off batteries in the event of a short gives me a nice warm and fuzzy - especially after the recent fire threads.

Most of the cheaper pre made battery packs are put together with bubble gum and alibi babba BMS units.
The latest JK BMS on my DIY batteries work with SMA SI5048 with SI6048 firmware once you work out JK got the Can_H and Can_L pins identified wrongly. I can pick on the JK Pylontech, GoodWe or Deye and the SI just does its thing and is fed the SOC and as the battery hits 55.2V (my set target V) the SB's are throttled.

Details here https://diysolarforum.com/threads/diy-280ah-battery-jk-bms-jk-pb2a16s20p-canbus-to-sma-5048.82330
 
Last edited:
@Trukinbear

setup a new battery, and charge it. Do not discharge it at night. next day, let it charge again. after a while the soc on the sunny island should go to 100%
i think discharging at night on the first night is causing a lot of issues
Yeah, I might have a problem with Mom's respirator, my BiPap, and my son's CPAP - we are totally off the grid. It IS catching up, now SI says 76% when the batteries are at 100%
 
The only lower cost battery packs that claims to communicate with SMA is EG4 through their add on coms box. No idea if the bugs alre all worked out.

Other than that, it’s high end SMA “partners.”

SOK almost cracked the code but basically gave up (something to do with SOC under 30%).

Most folks on here use REC or (I believe) Batrium BMS units for closed loop DIY batteries.

A few use VLRA and open loop.

The ability to broadcast in a certain language or protocol doesn’t meant the right info is being passed back and forth or if the BMS is smart enough to make the correct decisions. Sunny Islands in lithium mode hand almost all of the battery decisions to the BMS. That BMS needs the ability to make those decisions on behalf of the SI, and generic BMS units aren’t programmed with that ability in closed loop.

You could crack the batteries and install new BMS units but that defeats the purpose of buying off the shelf to a large degree.

This has been hashed out within this forum a bunch of times. I researched the piss out of it before buying REC BMS units (still had a few problems that were mostly sorted).

I would rather have the REC units though rather than a generic BMS that’s built into most lower cost premade batteries - with or without considering SMA issues. The big ass contactors that they use to turn off batteries in the event of a short gives me a nice warm and fuzzy - especially after the recent fire threads.

Most of the cheaper pre made battery packs are put together with bubble gum and alibi babba BMS units.
I wouldn't exactly call Aolithium cheap, or rather they aren't inexpensive. I opened one up to see if it was serviceable and it was full of flawless appearing Lazer welds, likey robotic. I think it's odd that the SI fails to even attempt to start when I choose the external BMS option. I still wonder about this inverter, SMA replaced my one year old new unit, with a well used example that appeared (to me) to have been dropped (same reason I cracked open the battery). I've still got a bad taste in my mouth from my experience with SMA tech support, and their well used warranty replacement - instead of diagnostic support. To top it all off recently (4 years later) SMA says they never received my return (they picked it up!) and my warranty and tech support are suspended (on my spare unit I presume) until I pay them thirty-six hundred some-odd dollars.
 
Wow! I wish I'd have gone LiFePo4 years ago - I thought I couldn't afford it. I was wrong because now I see that 50,ish % of my power went to heating up battery acid - I don't appear to actually use nearly as much power as I thought. Wow!
 
LiFePO4 became highly competitive over the past few years.
Several years ago, lead-acid price was more competitive.

For capacity to handle 3 days without sun, the difference in lifespan may be nil (vs. a system cycled 100% every night.)
Murdered forklift batteries negating that. LiFePO4 can be murdered too, but BMS supposed to prevent that. If FLA was invented today it might get similar protection, but dates from where that was a human job.
 
LiFePO4 became highly competitive over the past few years.
Several years ago, lead-acid price was more competitive.

For capacity to handle 3 days without sun, the difference in lifespan may be nil (vs. a system cycled 100% every night.)
Murdered forklift batteries negating that. LiFePO4 can be murdered too, but BMS supposed to prevent that. If FLA was invented today it might get similar protection, but dates from where that was a human job.
Looking back at my records that brand new picked up at the manufacturer forklift battery appears to have eaten over half my power from day one.
 
Putting the batteries in over a couple of months to spread the costs out - I thought I needed 600ah minimum, but what I'm seeing with the first 200Ah makes me believe that all my calculations based on my FLA system are WAY off, and 400Ah might be plenty
 
Cloudy and rainy today. Very overcast. Normally I'd be making only about 50% of the power I need, but I'm actually making about 20% more than I need
 
I went from a 17kwh gel battery to 14kwh LFP, the LFP gives me double the performance of the gel despite the lower number of kwh. Another 15kwh LFP on route.
 
I understood FLA consumes 1/3 of the power delivered, you get back 2/3. Put another way, what it consumes is 1/2 of what you get back from it.
Only from reading, I've never lived with FLA beyond small size.

I do use AGM, and I'm not logging its performance. Way over-paneled and grid-backup so most of their life spent basking in Float.

You probably have efficient off-grid house. I simply installed excess PV, back when I thought rebates made it a good deal. Was break-even with utility at best over 17 years. Today, PV + LiFePO4 is much cheaper (but if using lead-acid, lower cycle life may make cost same as grid.)
 
I understood FLA consumes 1/3 of the power delivered, you get back 2/3. Put another way, what it consumes is 1/2 of what you get back from it.
Only from reading, I've never lived with FLA beyond small size.

I do use AGM, and I'm not logging its performance. Way over-paneled and grid-backup so most of their life spent basking in Float.

You probably have efficient off-grid house. I simply installed excess PV, back when I thought rebates made it a good deal. Was break-even with utility at best over 17 years. Today, PV + LiFePO4 is much cheaper (but if using lead-acid, lower cycle life may make cost same as grid.)
We're pretty efficient, I designed this place knowing id likely remain off the grid, plus our climate is very mild. That 2200# forklift battery, who knows how many gallons (barrels) of battery acid I had to heat and keep warm. It was in the back of my mind, yes but in more of an anecdotal way, now looking empirically it's simply staggering. My original calculations weren't off, I just didn't adjust for 50% battery efficiency
 
Only takes energy to heat, not to keep warm. As you discovered when you insulated it. Maybe needed insulated enclosure and thermostatic vent.

Apparently insulation works for LiFePO4, just need a bit of electric heat, so even simpler.
I would consider if starting from scratch, but maybe only top name BMS like REC would be reliable? Even then, issues configuring.
I have strings of AGM for all.
 
Back
Top