DEYE Inverter UL Listed available in US

[the_colorist]

Now THATS a battery! Is that 269KW?!!
What do you have charging that if I may ask?

Honestly I don't have the specs on that yet, will soon though. This is a retrofit. Lead out, Lithium in. We're about twice the capacity of the preexisting lead. What I do know is it's a stack of Classics that I'll be triggering on/off with the AUX 2 input rather kill the entire system when a cell gets close on the high end. We'll still have a main kill-all contactor but it will have higher limits than the on-off/rest trigger for the Classics.

The inverters are a stack of 4x 8KW Radians.

Not an ideal system for lithium by any means but it works.

 

[solardad]

We're about to stack 5x of the 5.5A 16S active balancers in conjunction with a single Orion JR 2 operating 150 (15S) 280AH REPT cells. There will be 2x banks (150 cells each) with the Orions BMS kits (each with 5x balancers) operating in a Master+Slave configuration over CAN bus and an Orion WiFi Datalogger. The system this powers an island resort.

@the_colorist please do a build thread on this. I would love to follow your process...!

 

[robby]

Does sma sell an inverter that has all the capabilities of a Deye/SolArk?

No, but you can get almost the same thing using several different SMA products together and spending a lot more money.

 

[the_colorist]

@the_colorist please do a build thread on this. I would love to follow your process...!

LOL I will if I get time. Pictures as well if I can.

 

[Hedges]

The Sunny Islands have no computer hookup, no wifi, no ethernet (not really), no blue tooth.. and no options to hook any of these up. There is no way to remotely monitor them at all.. You can't even install an IP Camera and point it at the dot matrix display because the darn thing goes dark if you don't press a button every now and then.

I'm now using Sunny Web Box, which talks to both Sunny Island and Sunny Boy via RS-485. That (an obsolete product with known hacking vulnerability) provides remote monitoring and configuration, suitable for an air-gapped network.

Previously, I used Sunny Data Control with its own dinky display and pushbuttons to configure some Sunny Boys. It didn't recognize Sunny Island or the Sunny Boy 10000TLUS.

I configured the TLUS with Sunny Explorer and a Speedwire (Ethernet) piggyback, then reverted to RS-485 piggyback and tried to use them for grid-backup. I discovered the bug that feature wasn't correctly implemented in firmware, so SMA told me to use off-grid and they updated their compatibility list.

Colorist - you think Speedwire might work on older Sunny Boys for which it isn't listed?
We discussed RS-485 for power output throttling. Considering it is run at 1200 baud, 120 characters/second, I don't think it could do fully real-time control, but would be usable to average zero (or other amount) export over the one minute time frame for a multiple-inverter system.
I would consider a system with 48 kW SB, 24 kW SI, 10kW export limit. Catch is, need to limit power from SB feeding through SI (supplying other loads this side of meter) to about 24kW due to relay current limit. I don't think 1200 baud would be fast enough.

 

[houseofancients]

I just looked at the Orion BMS. It's over $1000. The Orion Jr is over $500, and then there are the wifi modules for $235, the SOC Meter, CAN adapter, cables....

look at the seplos bms

 

[the_colorist]

look at the seplos bms

I've installed, used and sell these. The 200A is nice and the physical quality is top notch. Best Chinese BMS I have ever seen or used and I simply would't use FET-based BMS units until I found this one. And I won't push it over 125A. It does have a great set of parameters though and does a good job of at least dropping the charge current to 10A during the last bit of charging. That said, I would never use less than 2 of those with a stack of 2x SI's. They will kill it. Really I wouldn't use a Seplos with SMA (small Victron system are great) but in a pinch it's okay. A single, bank with a single BMS is best where possible.

That said, the 200A I had in operation recently with a 5K Victron MPII just up and died. No reason. Replaced it with an Orion. Seplos said if we can't figure it out they will just send me a replacement.

 

[the_colorist]

Colorist - you think Speedwire might work on older Sunny Boys for which it isn't listed?
We discussed RS-485 for power output throttling. Considering it is run at 1200 baud, 120 characters/second, I don't think it could do fully real-time control, but would be usable to average zero (or other amount) export over the one minute time frame for a multiple-inverter system.
I would consider a system with 48 kW SB, 24 kW SI, 10kW export limit. Catch is, need to limit power from SB feeding through SI (supplying other loads this side of meter) to about 24kW due to relay current limit. I don't think 1200 baud would be fast enough.

You bring up excellent points as always.

I know there is a minimum firmware version for the Speedwire piggyback so assuming it had the correct version or newer, you should be fine. It's specified in the manuals floating round online (PDF) for the piggyback.

I had forgotten about the baud rate. I haven't tried this, but check out this tech note:

Service Tip: Setting up Sunny Boy Storage to control export of SMA PV inverters - Sunny. SMA Corporate Blog

AC-coupled battery storage allows any new or existing PV system to flexibly add battery storage. The new Sunny Boy Storage...

www.sma-sunny.com

When it talks about using Sunny Explorer to configure inverters with Webconnect only, that's inverters with a Speedwire Webconnect module installed (instead of an RS485 piggyback) and each inverter directly connected to a local ethernet network.

So ignoring the new direct RS485 functionality baked into the latest Data Manager M firmware, previously you would use the SMA COM gateway and that would give you direct access and control of any device on the RS45 bus and it's my understanding, you could do active limitation but I don't have that confirmed. It could be that the Webconnect via ethernet provides a much faster control interface than the RS485 piggyback's do hence no mention of them in the tech note above.

With the latest firmware, it should be that the Data Manager M maintains direct control of the PV inverters but per your note, perhaps full active control is only possible with webconnect-equipped inverters.

 

[houseofancients]

I've installed, used and sell these. The 200A is nice and the physical quality is top notch. Best Chinese BMS I have ever seen or used and I simply would't use FET-based BMS units until I found this one. And I won't push it over 125A. It does have a great set of parameters though and does a good job of at least dropping the charge current to 10A during the last bit of charging. That said, I would never use less than 2 of those with a stack of 2x SI's. They will kill it. Really I wouldn't use a Seplos with SMA (small Victron system are great) but in a pinch it's okay. A single, bank with a single BMS is best where possible.

That said, the 200A I had in operation recently with a 5K Victron MPII just up and died. No reason. Replaced it with an Orion. Seplos said if we can't figure it out they will just send me a replacement.

been using the 200A version for a while now, pushing it to 150A max, and has been working flawless so far...
I find Seplos is really responsive to support questions....

there as another user that had one fail, but to the best of my knowledge they helped him getting it fixed pretty soon

my only "complaint" so far is that they only support CANBUS for inverter communication, so i am having to use a raspberry to translate CANBUS to RS485 for my inverters..
My contact said however they are working on a firmware update so the BMS can do both CANBUS and RS485 , as the BMS has both natively onboard.
Offered to be their Beta tester if need be

 

[the_colorist]

been using the 200A version for a while now, pushing it to 150A max, and has been working flawless so far...
I find Seplos is really responsive to support questions....

Good to know! I'm sure others will appreciate hearing that as well.

Back to the Deye inverters for a moment, we're working on getting access to a 9000W continuous (9800W peak) version of the Deye inverters. Deye still hasn't released the 12K but this "10KW" version is accessible. The only issue is we don't have a UL1741 certificate for it yet. So the 8KW version is still appropriate for the installations that require a certificate. The 10KW does comply technically, it's just not certified.

Feel free to PM for details.

 

[the_colorist]

Other specs include 12,400W of PV input across 2x MPPTs as opposed to 10,400W on the 8KW.

 

[animatt]

I have a question about over paneling the deye inverter.

How sensitive is the integrated charge controller to over paneling?

It looks like max watt input is listed at 10,400w. Or 1.3x the 8kw rating.
I know over voltage can kill the inverter.
(Solark 8k has 11,000w limit in manual. And 6000k per mppt input)

I plan to install an array that would be at suboptimal orientation. Maybe around 13kw in size. So most of the time potential power input would be much lower. But if freak conditions align and system does have say 13kw of potential how will the charge controller handle it. Just clip safely at 8kw?

Obviously a charge controller needs to be able to tolerate having excess power when batteries are full. Any idea of exactly where the hard limit is?

I guess i could follow the manual but that would limit vast majority of the time to less solar collection.

Thanks

 

[solardad]

How sensitive is the integrated charge controller to over paneling?

I believe the MPPT controller just 'clips' current coming in. NOTE - The Open Circuit Voltage per MPPT needs to stay below 500v.

 

[Roqm]

I have a question about over paneling the deye inverter.

How sensitive is the integrated charge controller to over paneling?

It looks like max watt input is listed at 10,400w. Or 1.3x the 8kw rating.
I know over voltage can kill the inverter.
(Solark 8k has 11,000w limit in manual. And 6000k per mppt input)

I plan to install an array that would be at suboptimal orientation. Maybe around 13kw in size. So most of the time potential power input would be much lower. But if freak conditions align and system does have say 13kw of potential how will the charge controller handle it. Just clip safely at 8kw?

Obviously a charge controller needs to be able to tolerate having excess power when batteries are full. Any idea of exactly where the hard limit is?

I guess i could follow the manual but that would limit vast majority of the time to less solar collection.

Thanks

Me personally i wouldnt go over the rated max...they have that max for longevity i am pretty sure you can go over how much for how long? No one has tested yet(that i have seen)

 

[robby]

Me personally i wouldnt go over the rated max...they have that max for longevity i am pretty sure you can go over how much for how long? No one has tested yet(that i have seen)

I could show you a stack of 25KW SMA inverters at a business about 3 miles from my house.
The electrician who did the integration wiring is the same guy who is doing my house. Long story short is that he said they have over 1500 panels on the roof and some "Solar Engineer" advised them that they could get a lot more power out of the system if they followed a plan he had devised. The guy says that they were running something like 600V into the MPPT and the guys plan called for adding 2 more panels on each string which he said was still way below the maximum. Well it worked for a few months and then he said one inverter failed, then another a few weeks latter and then another. I think he said they have six of them in the store room waiting to be shipped back to SMA.

My advice is don't go even close to the limit.

 

[the_colorist]

Definitely stay below the voltage limit but if you can keep your VMP close to 370V-400V, you'll see some better efficiency.

The max ISC on the MPPT is 22A each. Let's lay this out:

370V x 22A per channel = 8,140W per controller

So you can exceed the 5,200W throughput listed on the datasheet but I would definitely stay below 8KW per input. It is very common for MPPT controllers to have a much higher ISC limit than their total current throughput to the low side of the controller but you just have to utilize it smartly. Look at the "Max. PV short circuit current" on the average Victron controller. It allows for a much larger PV array than it can technically use but exceeding the "Max. PV short circuit current" value does put the controller at risk.

Bear in mind, in the Deye inverters, the MPPT achieves 99% eff because the output bus is ~400V. That bus then feeds a ~400V to 40-60V 190A bi-directional DC-DC for the batteries and a ~400V to 120/240V 8KW inverter. You shouldn't have any issue exceeding the 10,400W limit some, particularly in suboptimal orientation. The MPPT's will clip at either the hardware limit or the software limit if it's lower. Try to avoid hitting the hardware limit (22A) if you can.

Also bear in mind the heat. These will heat up fast when there is 8KW+ of PV, 6KW+ of loads and charging the battery through the DC-DC at a few KW. So if you're in a hot environment, I would manually limit the PV input to 8KW-10KW, depending on what is appropriate for the loads vs environment. The battery will charge slower if the inverter is under a full load but it will keep the unit from derating due to overheating.

 

[SolarPrep]

been using the 200A version for a while now, pushing it to 150A max, and has been working flawless so far...
I find Seplos is really responsive to support questions....

there as another user that had one fail, but to the best of my knowledge they helped him getting it fixed pretty soon

my only "complaint" so far is that they only support CANBUS for inverter communication, so i am having to use a raspberry to translate CANBUS to RS485 for my inverters..
My contact said however they are working on a firmware update so the BMS can do both CANBUS and RS485 , as the BMS has both natively onboard.
Offered to be their Beta tester if need be

Do the Seplos work with NMC, and other lithium chemistries? I looked at importing some of their batteries last year, but it seemed they were mostly (if not totally) foucsed on LiPO. Thanks!

 

[the_colorist]

Do the Seplos work with NMC, and other lithium chemistries? I looked at importing some of their batteries last year, but it seemed they were mostly (if not totally) foucsed on LiPO. Thanks!

Yes, they do work with NMC.

 

[the_colorist]

Yes, they do work with NMC.

We're now a dealer for Seplos in the US.

 

[Roqm]

I could show you a stack of 25KW SMA inverters at a business about 3 miles from my house.
The electrician who did the integration wiring is the same guy who is doing my house. Long story short is that he said they have over 1500 panels on the roof and some "Solar Engineer" advised them that they could get a lot more power out of the system if they followed a plan he had devised. The guy says that they were running something like 600V into the MPPT and the guys plan called for adding 2 more panels on each string which he said was still way below the maximum. Well it worked for a few months and then he said one inverter failed, then another a few weeks latter and then another. I think he said they have six of them in the store room waiting to be shipped back to SMA.

My advice is don't go even close to the limit.

Thats kind of my feeling as well, although my panels would potentially be 480vdc@ 13a on each mppt(10 panels)! I know i wont see this constantly so it was a comfortable range for me.