New SolArk 12K, X2, issues/glitches..

[Shawn Hayes]

Edit: This is an off grid setup. The issues described below are with the two new units. Old units are in the box on the shelf currently.

SolArk is aware of the two issues I am posting about. I’ve been in communication with Tech Support. They have suggested trying one solution to see what happens and so far, no luck.
Therefore, I wanted to post this here and look for thoughts and suggestions? They’re a lot of knowledgeable and experienced folks on this site. Some people may have had these issues before and understand the problem/solution.
That is why I wanted to reach out to the forum as well. The two new 12K units are replacing two original first gen (2019/20) models .
System: Dual SolArk 12K, 16s4p diy LiFePo4 (63kwh), Batrium BMS w/ Closed loop comms.

Issue #1: When the cooling fans for unit 2 kick on, solar production and battery charging drop to zero for a second. They slowly recover to about one quarter of what it was before the fans kicked on. So, when this happens, unit number one ramps up solar production, and battery charging production to compensate for the lagging number two unit. It takes anywhere from 20 to 45 minutes for unit 2 to get back prior production and equal unit 1. Yes, I’ve timed this!!!
Sol-Ark suggested reconfiguring to make current unit 2, new number 1, and vice versa. So I did.
Now in new configuration, when the cooling fans kick on the new unit number one (old number two) solar production and battery charging production, drop to zero on both units. Typically your lead unit, number 1, controls unit 2.
However, it seems to recover quicker in this configuration, but with the old number 2 unit still lagging in recovery over old unit 1.
As a note: i’ve noticed that the cooling fans on the original number one unit seem to turn on 6-8 times a the day. However, the cooling fans on the old number two unit, the one that seems to be giving the issue, may turn on once, or twice a day.? Thoughts?

Issue #2: At 35-36 amp battery discharge per inverter, I’m getting a yellow batteries warning symbol. 70-72 total amps. Not sure why when discharge is set to 100 amps per inverter.
A few pictures attached so you can see if I’m missing something?
FWIW: SolArk so far hasn’t offered thoughts on this issue yet.

Also I have videos of the inverter drop out when fans kick on but don’t know how to or best way to share here?

Edit: These units were installed Tuesday 10/17, today Saturday 10/21 for reference. So these were brand new.

Photos are of yellow battery warning and amps, my settings and setup.

I

 

[Quattrohead]

1. Do you have all your working settings from the old units on the new units.
2. Why did you replace 3 year old units ?

 

[Shawn Hayes]

Yes. Old settings basically the same. I screenshot all screens before the swap.
However, gen charge is now on grid input, recommend by SA in an off grid setup.

Replaced because I had early models with “green” boards as I was told. Certain small electric items (hand held router, electric drills, band saw, heat gun, Keurig coffee maker, hair dryers etc) that shouldn’t pull enough power to do so, would trip my inverters off!! So, for example, I would have to buy three or four different (coffee makers, hair dryers) bring them home and try them out to see which one would not trip my system.
Yes, this was a thing with early models. And by the time that we realized that was an issue, Sol-Ark would not help or replace in any way.

So far, these have not tripped off with Keurig, hair dryers, heat gun or band saw.

 

[Shawn Hayes]

Here’s a video of the glitch when cooling fans kick on. Watch solar drop out and battery charge turns to discharge for a bit. It recovers but shouldn’t do this when fans kick on.
With this unit being the current number one, it draws down number two. When the other unit was number one, and this glitch happened to number 2, only number 2 would drop out, then it took 20-45 minutes for number 2 to recover.

 

[Shawn Hayes]

@HighTechLab have you seen this before or have thoughts on this?

 

[HighTechLab]

I'm confused. Is the fan issue with the new units or the old units?

Based on the detailed info screen, the yellow battery warning has something to do with how many amps are flowing. You may try running without the BMS communications briefly to see if it's something caused by that. Also make sure you are on the latest firmware and check the time of use settings.

 

[Shawn Hayes]

I'm confused. Is the fan issue with the new units or the old units?

Based on the detailed info screen, the yellow battery warning has something to do with how many amps are flowing. You may try running without the BMS communications briefly to see if it's something caused by that. Also make sure you are on the latest firmware and check the time of use settings.

Issues are with one of the brand new units. (Installed Tuesday 10/17)
Old units are in the box on a shelf currently.
Latest firmware was done the day after they were installed. Both new units are on same firmware.

 

[Quattrohead]

6.2.2.2 ? Was the latest for the 12k when I last looked. It has not been updated in a while now.

 

[Shawn Hayes]

Both are:
Comm: 143f-1724
Mcu Ver: 6222
Update was done Wednesday.
I saw another post here where SolArk leaves out 12K on the settings screen. Ditto for mine.

 

[Oldphile]

Both are:
Comm: 143f-1724
Mcu Ver: 6222
Update was done Wednesday.
I saw another post here where SolArk leaves out 12K on the settings screen. Ditto for mine.

View attachment 173637

That's just a typo. The 15k is 7222 or something that starts with 7, I forget.

 

[MaikaiLife]

Regarding the yellow battery shading, from my understanding that typically means that the currenltly flowing amps are near the programmed limit. Is it possible that since they're in a Master/Slave combination, the 100 amps you have set is actually the aggregate? That would kinda make sense because 73 amps is close(ish) to your configured 100 limit.

Regarding the issue with the fan turning on, have you tried measuring with another tool to see if the display is just a glitch or if that's what actually is happening? The reason I ask is because in my studies on my 15K the Sol-Ark tends to be a little slower at displaying the changes in current than what my meter or my Victron shunt show. In my testing, my Victron shunt will show a charge or discharge at least a second or two before the Sol-Ark display seems to register. It's a small amount of time, but I indicates that there is a tiny bit of delay, and that delay could be exacerbated in a master/slave config.

 

[Shawn Hayes]

Regarding the yellow battery shading, from my understanding that typically means that the currenltly flowing amps are near the programmed limit. Is it possible that since they're in a Master/Slave combination, the 100 amps you have set is actually the aggregate? That would kinda make sense because 73 amps is close(ish) to your configured 100 limit.

It’s my understanding from SolArk that the 100 amp limit is per inverter. (200 amps total) Wednesday morning, when talking to tech support, the tech confirmed this, but, it may be worth a try to raise it and see.
I did pull Batrium comms, with SolArk on its own, I still got the yellow warning.

Regarding the issue with the fan turning on, have you tried measuring with another tool to see if the display is just a glitch or if that's what actually is happening? The reason I ask is because in my studies on my 15K the Sol-Ark tends to be a little slower at displaying the changes in current than what my meter or my Victron shunt show. In my testing, my Victron shunt will show a charge or discharge at least a second or two before the Sol-Ark display seems to register. It's a small amount of time, but I indicates that there is a tiny bit of delay, and that delay could be exacerbated in a master/slave config.

I have not tried any sort of test, quite frankly not sure how to or what to test? However, my Batrium screen/data confirms the actions described above.
Current unit number one has the issue, not current number 2. In its current configuration, number one drags both units down.
In reverse configuration only, unit number 2 (currently number 1) only shows the dip, while number one (currently number 2) actually ramps up to compensate.
At this point, I’m not sure what to measure or check? SolArk mentioned a factory reset, however I’m awaiting the next response from SolArk.

 

[Quattrohead]

Sounds like you going to have to split them up and test them individually, one seems like it might be faulty.

 

[400bird]

If you think it's related to the fans kicking on, have you tried testing to verify this? As a test, I'd put a fan under it blowing air up through the unit

 

[Shawn Hayes]

Yes, however why would fans coming on cause PV to drop? The unit should work as designed. The other unit works even when fans come on.

Update: Just performed a factory reset to see if helps and it did not!

 

[Shawn Hayes]

Yes, however why would fans coming on cause PV to drop? The unit should work as designed. The other unit works even when fans come on.

Update: Just performed a factory reset to see if helps and it did not.

 

[400bird]

Yes, however why would fans coming on cause PV to drop? The unit should work as designed. The other unit works even when fans come on.

I 100% agree the unit shouldn't be doing this. My external fan suggestion is just to prove that is related to the internal fans.

 

[MaikaiLife]

As @400bird mentioned in his suggestion, sometimes the actual cause of an issue is the most unlikely guess. Testing them individually seems like a good starting point, if they behave differently then you have a baseline to show that the two units are not equal in their behavior. These are complicated devices both physically and programmatically, you could be the first person to discover this as a fault in the hardware or even a glitch in the firmware.

I would focus on testing each issue separately so you don't get too strung out in multiple directions. For the yellow highlight issue, raise the maximum current, then also try them individually and see what happens.

For the fan issue, try what @400bird suggested keeping the unit cool might prevent the fan from turning on and further help confirm the PV drop is related to a thermal event.

I have confidence in the Sol-Ark product and I'll do my best to keep thinking on what could be the issue, I'll even stare at my unit for a little while and see if I can come up with some further ideas.

Keep us posted on your efforts so we can collectively help you find a solve. Cheers!

 

[Shawn Hayes]

I 100% agree the unit shouldn't be doing this. My external fan suggestion is just to prove that is related to the internal fans.

As @400bird mentioned in his suggestion, sometimes the actual cause of an issue is the most unlikely guess. Testing them individually seems like a good starting point, if they behave differently then you have a baseline to show that the two units are not equal in their behavior. These are complicated devices both physically and programmatically, you could be the first person to discover this as a fault in the hardware or even a glitch in the firmware.

I would focus on testing each issue separately so you don't get too strung out in multiple directions. For the yellow highlight issue, raise the maximum current, then also try them individually and see what happens.

For the fan issue, try what @400bird suggested keeping the unit cool might prevent the fan from turning on and further help confirm the PV drop is related to a thermal event.

I have confidence in the Sol-Ark product and I'll do my best to keep thinking on what could be the issue, I'll even stare at my unit for a little while and see if I can come up with some further ideas.

Keep us posted on your efforts so we can collectively help you find a solve. Cheers!

Thanks the good info and suggestions. That’s why I enjoy reading or posting here.
Currently getting latest firmware/comms/mcu pushed again even though they’re on latest version. Since the factory reset did not help, we’ll see what this does. If issues persist, then see what SolArk suggests on the fan issue. Perhaps it’ll require more testing on my end?
Also have taken Batrium comms away now too see if things change. Just from personal history and with the other inverter behaving “normally “ I don’t think it will affect things.

I’ll try raising my discharge limits as well to see if that helps the battery warning

 

[Shawn Hayes]

Just fyi:
Firmware update just finished and now says MCU Vers: 6224.
Last update was Wednesday, (5-days ago) and it was 6222.

 

[SCmountainman]

A couple of questions/ brainstorm points…

Why do you have battery charge setting of 40A on batt set screen? Don’t you have tons of PV?

Why are you running a 16s battery at such low voltage settings? (Have you tried 54.4 float/55.8charge/equalize off)

Can you/have you tried having BMS not communicating to SolArk? Let BMS on DIY pack be stand alone?

Have you tried having SolArk use batt V instead of %?

Limited discharge to 100Amp? Total of 2 units x 100A = 200Ax53v only 10kW? Why 2 units? Just for extra PV handling capacity?

Maybe a single 15k unit would have handled loads and split PV arrays up with stand alone charge controllers to not max out a single unit. You’ve got a huge battery bank and looking at settings you’re barely touching it.

Just some thoughts

 

[Shawn Hayes]

A couple of questions/ brainstorm points…

Why do you have battery charge setting of 40A on batt set screen? Don’t you have tons of PV?

Why are you running a 16s battery at such low voltage settings? (Have you tried 54.4 float/55.8charge/equalize off)

Can you/have you tried having BMS not communicating to SolArk? Let BMS on DIY pack be stand alone?

Have you tried having SolArk use batt V instead of %?

Limited discharge to 100Amp? Total of 2 units x 100A = 200Ax53v only 10kW? Why 2 units? Just for extra PV handling capacity?

Maybe a single 15k unit would have handled loads and split PV arrays up with stand alone charge controllers to not max out a single unit. You’ve got a huge battery bank and looking at settings you’re barely touching it.

Just some thoughts

Good to hear from you. I text you a few times with no replies. You outta come visit and check things out here, should come visit. Reach out if you want.
So far 40 amps X 2 = 80 amps, has been sufficient to charge back to 3.35/6 v per cell each day. 3.35/6 Vpc, 53.9/54 volts is about 90%, that’s where I want to charge, based on the reading/research I’ve seen on this forum. So far my balance has been good and not using balancer at all. I’m using Batrium as my BMS.
Due to current issues, troubleshooting with Sol-Ark, I’m currently out of closed loop comms. Sol-Ark is controlling the batteries currently.
Currently using Batt V charged currently. Percentage is when Batrium is in closed loop only.
100Ax2=200 amp discharge, so far according to Batrium data, my max discharge has been 137 amps total on the batteries.
16,830 watts of pv the reason for dual units. Plus the Preppers rule: (one is none, two is one)

PM me.

 

[niktak11]

Good to hear from you. I text you a few times with no replies. You outta come visit and check things out here, should come visit. Reach out if you want.
So far 40 amps X 2 = 80 amps, has been sufficient to charge back to 3.35/6 v per cell each day. 3.35/6 Vpc, 53.9/54 volts is about 90%, that’s where I want to charge, based on the reading/research I’ve seen on this forum. So far my balance has been good and not using balancer at all. I’m using Batrium as my BMS.
Due to current issues, troubleshooting with Sol-Ark, I’m currently out of closed loop comms. Sol-Ark is controlling the batteries currently.
Currently using Batt V charged currently. Percentage is when Batrium is in closed loop only.
100Ax2=200 amp discharge, so far according to Batrium data, my max discharge has been 137 amps total on the batteries.
16,830 watts of pv the reason for dual units. Plus the Preppers rule: (one is none, two is one)

PM me.

3.35V is not close to full. Bump that up to 3.45V/cell at least.

 

[MaikaiLife]

are you still having the fan issue and drop in solar production or did the firmware resolve it?

 

[Shawn Hayes]

Nothing has helped. Even swapped out new MCU board.