diy solar

diy solar

Sunny island 6048

Well, to repeat:


But no there more nuance to the different modes with the SIs besides just a wider tolerance in measured parameters. For example in grid mode frequency variations are considered a sign of issues with the grid and cause the SMAs to trip into off grid mode as precaution. In generator mode the SMAs consider that signs of an over/under loaded generator and adjust power draw in an attempt to "regulate" the generator in a sense. They will also more aggressively attempt to balance load between the legs by shifting power between each other through the DC bus, as the detriment of an unbalanced load is more extreme to a generator but not to a grid. You'd be wasting energy in inversion loses if it was left in generator mode and trying to balance the load all the time.

Anyway that's most of the behavioral difference i know of / have observed.
That's interesting.
My solar is normally grid tied and my off grid system with SI's is totally isolated when the grid is working.. Since we only use it for power outages, I probably don't have nearly the amount of run time as you.
That said, I've had to use the generator twice because I keep my lithium bats at only 1/2 charge when in storage.

I set the SI's to draw 4000 watts from the 5500 watt portable generator and when the gen is running, the SI's draw a steady 4000 watts from them regardless of what I'm doing in the house.. or at least it seems that way because I never hear any change in the generator's noise output.

In the old days before the solar, when the house was running on the generator, the well pump would cause a big dip in generator noise as the pump kicked on.. generator rpms would dip a bit, cough, and then the generator's governor would apply more power and it would pick up again.
That behavior disappeared when we ran the generator through the SI's.
 
I set the SI's to draw 4000 watts from the 5500 watt portable generator and when the gen is running, the SI's draw a steady 4000 watts from them regardless of what I'm doing in the house.. or at least it seems that way because I never hear any change in the generator's noise output.

In the old days before the solar, when the house was running on the generator, the well pump would cause a big dip in generator noise as the pump kicked on.. generator rpms would dip a bit, cough, and then the generator's governor would apply more power and it would pick up again.
That behavior disappeared when we ran the generator through the SI's.
Yep you are seeing the SMAs act as "buffer" between load and generator, either sinking or sourcing power to always keep the generator loaded the same.

It's the same premise as a Hybrid vehicle (except we don't need a motor to draw down or boost the engine), engines become much more fuel efficient if the load is steady and at the point of peak efficiency. Sometime that peak point is almost double the efficiency than at a lower or higher engine workload. So for you and I that doesn't matter much, but when you have 24 of these SMAs all in a room acting as the power grid for a small island, then the difference in fuel costs adds up pretty quick which is why these will regulate to a user-set power point then best they can, and try to keep the electrical end balanced between the phases as well.

I don't think many people appreciate (or really just notice) how vastly over-engineered these bad boys are vs the much cheaper competition.
 
Hmm... Being new to Sunny Islands and doing a new large install with 4 of them right now.. I much welcome any suggestions on setup..

My plan was an off grid system with 2nd AC input in permanent generator mode but connected to the grid for backup.. I don't have nor plan to buy a gas generator. Also can't back-feed into the grid so figured putting in generator mode would ensure that... Or it sounds I should setup as grid input and set something else to ensure no back-feed?

When I have excess PV power available I'd like to dynamically power up loads... I know the SI will change frequency to lower PV output, but rather it just fire up more loads.. Otherwise I hope that telemetry can tell me the average daily amount of excess power then I will just plug in a permanent load that uses about the same amount..

madsci1016, as for battery management I just bought everything to implement your code as it does look like the way to go in the end.. but sounds like I will have to say they are lead acid until its working.. I could monitor/reverse the 485 communications to the GYLL batteries, however with 4 of them and 4 others from another company it seems the best way is just a smartshunt on the entire battery system..
BTW noticed another company that is doing similar control between their BMS and the SI. https://discoverbattery.com/products/lithium-batteries/lynk-gateway

Telemetry is another big question I need to figure out, maybe using Victron VenusOS is the way to go..but I only own the Smartshunt currently.. Otherwise SMA sunny portal or??? Should I get a SMA ethernet adapter for the SI's for telemetry since all my Sunny Boys have it? If I use the VenusOS, it can see the SI but not the Sunny Boys... Not sure best way to go with telemetry... also bought a Square D Wiser Energy Smart Home Monitor with solar, thinking it might be good.. or not..err.
 
Hmm... Being new to Sunny Islands and doing a new large install with 4 of them right now.. I much welcome any suggestions on setup..

My plan was an off grid system with 2nd AC input in permanent generator mode but connected to the grid for backup.. I don't have nor plan to buy a gas generator. Also can't back-feed into the grid so figured putting in generator mode would ensure that... Or it sounds I should setup as grid input and set something else to ensure no back-feed?
There's a parameter for that! 232.08. Per the manual:
If GdCharge is set, no energy is fed into the grid. If GridFeed (Default) is set, energy is fed into the grid.
And they mean it. If even by accident more than a few watts leaks back into the grid, the SMAs will sever grid connection and go off-grid for a 20 minute safety period, regardless of any other setting or command. Don't ask me how i know that...

So yes, set it as grid and change that parameter to GdCharge is all you need. In general i recommend anyone read the manual front to back a few times before diving in too deep with these guys.

When I have excess PV power available I'd like to dynamically power up loads... I know the SI will change frequency to lower PV output, but rather it just fire up more loads.. Otherwise I hope that telemetry can tell me the average daily amount of excess power then I will just plug in a permanent load that uses about the same amount..

There are 2 aux relays per SMA, and they all can be used to close contactors (aka turn on loads) based on battery SoC. That functionality is actually built into the SMA already. You just have to set it up. Yo can set SoC levels per relay.

the best way is just a smartshunt on the entire battery system..
I'd agree.
madsci1016, as for battery management I just bought everything to implement your code as it does look like the way to go in the end..
Hopefully you are already a teeny bit comfortable with basic Linux computer technologies like SSH, root, python, and github. If not I'd google those terms a bit first before trying my code, just so you have a solid start. It's not beginner friendly yet.

Telemetry is another big question I need to figure out, maybe using Victron VenusOS is the way to go..but I only own the Smartshunt currently.. Otherwise SMA sunny portal or??? Should I get a SMA ethernet adapter for the SI's for telemetry since all my Sunny Boys have it? If I use the VenusOS, it can see the SI but not the Sunny Boys... Not sure best way to go with telemetry... also bought a Square D Wiser Energy Smart Home Monitor with solar, thinking it might be good.. or not..err.
There are many (some third party) written drivers for solar inverters like (I believe) SunnyBoys if you google. Like this.
I assume that requires the SunnyBoys to be networked in some way. Worse case VenusOS supports buying a certain power meter and putting it on any solarPV inverter to track it's output, which may be easiest.

You aren't the guy that painted all 4 of his SMAs black recently by any chance are you? Saw pics of that recently.
 
The Sunny Island 8.0H manual says that the DC bus voltage limit is 63V, corresponding to a fully charged 15S lithium-ion. The REC SI BMS can connect up to 16S lithium-ion for 67.2V. So what is the result of doing this? Presumably you either end up with under charged cells, or it will cause an error in the BMS or the SI or both. Anyone tried it?
 
The Sunny Island 8.0H manual says that the DC bus voltage limit is 63V, corresponding to a fully charged 15S lithium-ion. The REC SI BMS can connect up to 16S lithium-ion for 67.2V. So what is the result of doing this? Presumably you either end up with under charged cells, or it will cause an error in the BMS or the SI or both. Anyone tried it?
Damage to the Sunny Island? Over-voltage alarm maybe?

I wouldn't do it..

However, if you connect up such a system and it is only half charged, I would suppose the Sunny Island would refuse to go over 63 volts.. probably produce some kind of error code.. but I'm not sure.
 
Is the 16s REC BMS meant for LiFePO4, which would be within the SI limits.
 
Is the 16s REC BMS meant for LiFePO4, which would be within the SI limits.

REC BMS is fully programmable and can work with any chemistry.. It can handle up to 16 cells in series..
 
The Sunny Island 8.0H manual says that the DC bus voltage limit is 63V, corresponding to a fully charged 15S lithium-ion. The REC SI BMS can connect up to 16S lithium-ion for 67.2V. So what is the result of doing this? Presumably you either end up with under charged cells, or it will cause an error in the BMS or the SI or both. Anyone tried it?

63V / 16 = 3.94V/cell.
No reason to charge LiFePO4 that high.
Undercharge doesn't harm it. Just need to get high enough for balancing to work (past the knee). People normally top-balance to about 3.6V, operate lower.
 
Finally the sma temp sensor and the 10k resistor arrived. But the sensor is 1.981k ohms. And the 10k resistor still work?
 
Is the 16s REC BMS meant for LiFePO4, which would be within the SI limits.

If you use the REC, you don’t need a temp sensor or current shunt for the SI. The battery parameters are solely controlled by the BMS when the SI is in lithium mode.
 
Finally the sma temp sensor and the 10k resistor arrived. But the sensor is 1.981k ohms. And the 10k resistor still work?

Guess I guessed wrong on resistor value. Never got around to digging through my stored supplies and measuring one.

You can install the temp sensor, but if temperature changes it will adjust voltage to your lithium batteries. Set temperature coefficient of voltage to zero.

Do you have several resistors? Five of them in parallel would be 2k ohms. Measure voltage Sunny Island floats at, and you might tweak voltage settings to get what you want. or set temperature coefficient to zero (which seems simplest)

I think with the temp sensor and zero for voltage coefficients, you'll have the convenience of temperature reading accessible from SI display or any remote interface you connect. You can also program a fan to cool the batteries at some temperature. Or enable a heater below some temperature.
 
Guess I guessed wrong on resistor value. Never got around to digging through my stored supplies and measuring one.

You can install the temp sensor, but if temperature changes it will adjust voltage to your lithium batteries. Set temperature coefficient of voltage to zero.

Do you have several resistors? Five of them in parallel would be 2k ohms. Measure voltage Sunny Island floats at, and you might tweak voltage settings to get what you want. or set temperature coefficient to zero (which seems simplest)

I think with the temp sensor and zero for voltage coefficients, you'll have the convenience of temperature reading accessible from SI display or any remote interface you connect. You can also program a fan to cool the batteries at some temperature. Or enable a heater below some temperature.

How do you know its an RTD and not a thermocouple?
 
How do you know its an RTD and not a thermocouple?

I don't.
I generally see thermistors in products and thermocouples in (some) instruments.
I guess he could measure voltage, then warm and cool and see what happens. Measure resistance at each temperature too.

Now that he's got the SMA temperature sensor, I would say use that and program zero temperature coefficient so it doesn't change voltages.
If heat could be an issue, maybe program a relay to turn on a fan. But I'm not sure lithium ever self-heats enough that a fan would matter like it can for lead-acid during charging. Could set an over-temperature shutdown.
 
I don't.
I generally see thermistors in products and thermocouples in (some) instruments.
I guess he could measure voltage, then warm and cool and see what happens. Measure resistance at each temperature too.

Now that he's got the SMA temperature sensor, I would say use that and program zero temperature coefficient so it doesn't change voltages.
If heat could be an issue, maybe program a relay to turn on a fan. But I'm not sure lithium ever self-heats enough that a fan would matter like it can for lead-acid during charging. Could set an over-temperature shutdown.
I was wondering about that.. If its a thermocouple, he could have just jumper'd the terminals inside the inverter and it should have registered room temperature all the time.
 
I bought a Jbd bms with canbus. Will that work or it need to be some special protocols built in ?
 
I don't know whether they use the same protocol or a different one.
Here's a post (on a different forum) which mentions use of JBD BMS with Sunny Island. I think he programmed a controller to talk to each of them.

 
Well, to repeat:


But no there more nuance to the different modes with the SIs besides just a wider tolerance in measured parameters. For example in grid mode frequency variations are considered a sign of issues with the grid and cause the SMAs to trip into off grid mode as precaution. In generator mode the SMAs consider that signs of an over/under loaded generator and adjust power draw in an attempt to "regulate" the generator in a sense. They will also more aggressively attempt to balance load between the legs by shifting power between each other through the DC bus, as the detriment of an unbalanced load is more extreme to a generator but not to a grid. You'd be wasting energy in inversion loses if it was left in generator mode and trying to balance the load all the time.

Anyway that's most of the behavioral difference i know of / have observed.
This gives me an idea to try next summer, if I still have issues with stability of the AC1 side from SunnyIslands, when all my micros are cranking and mostly running the old PSC motor A/C unit. The theory is that the microinverters anti-islanding detection is too much or too hard for the SIs to maintain a fixed frequency, and it oscillates +- a couple Hz around 60Hz. I have not tried telling it to be in generator mode and close the relay (and separately disconnect the grid from there, manually. danger will robinson), maybe then it would be stable. Thanks again for another good idea madsci. However, I'm currently replacing that old A/C with an inverter heatpump, which I hope will also prevent the issue from occuring.
 
Back
Top