Hi all.

This is something very weird that happens quite often on my system.

For the image below I've selected just the battery SoC and the PV input from the Panels for this morning.

You can see that about 6am PV starts to push in power, and battery begins to charge nicely. Then about 8am, the rate of PV generation reverses - slowing down till it's barely making 10% of Solar capacity by 10am.  But the battery has never reached 100%, in fact only just got to 80% before starting to discharge to carry load.

The sun has been going strong all day, and I have no shading at all.

Any ideas on what on earth is going on here?

 

 

And today I see a similar pattern, though it did manage to get the battery to 100%. But since the rate of generation slowed about 9:00 am, it just won't come back up again, and now the battery is being depleted while the sun is blazing.

I have 3.27 KW solar capacity installed. yet I'm only getting 1.2 KW in the first couple of hours of the day, and less than 400W most of the time.

Please HELP...!!! 

 

 

For comparison, look at the behaviour on 1 November, 2 days after the original installation. That day we had intermittent clouds, but look at how strong the generation was throughout the day... Often generating at over 2KW.

Have you changed any settings on the battery / inverter or received a firmware update?

Hi @mzezman. Thanks for responding.

I have been making adjustments as we constantly try to balance efficiency vs this crazy Load shedding situation.  But I don't believe any setting I've tweaked should cause this change.

I do not believe I've had a firmware update, unless it would do that automatically without my involvement. Current versions are: M 3.3.7.4 / S  1.5.1.5 / C E.4.2.4.  I know I verified that the version was up-to-date when the system was installed (30 October 2022), but I see there is a newer version available now.

Last night my installer stopped by and checked physical wiring etc, which he said looked to be all good.  He was concerned that the inverter felt very warm. We've put a fan to blow over it all day today and see if that helps.  FYI - here is my temperature graph from yesterday - are these peaks a problem (DC:78, AC:57) ? Could this explain the degradation of panel generation capacity?

 

You can see where we added the fan - just after 6pm last evening.

 

32 minutes ago, Patrick OReilly said:

Hi @mzezman. Thanks for responding.

I have been making adjustments as we constantly try to balance efficiency vs this crazy Load shedding situation.  But I don't believe any setting I've tweaked should cause this change.

I do not believe I've had a firmware update, unless it would do that automatically without my involvement. Current versions are: M 3.3.7.4 / S  1.5.1.5 / C E.4.2.4.  I know I verified that the version was up-to-date when the system was installed (30 October 2022), but I see there is a newer version available now.

Last night my installer stopped by and checked physical wiring etc, which he said looked to be all good.  He was concerned that the inverter felt very warm. We've put a fan to blow over it all day today and see if that helps.  FYI - here is my temperature graph from yesterday - are these peaks a problem (DC:78, AC:57) ? Could this explain the degradation of panel generation capacity?

 

You can see where we added the fan - just after 6pm last evening.

 

Heat could be a factor - see how it performs today with the fan on. I know the 5kw derates once DC temps go above 75

13 hours ago, Patrick OReilly said:

Hi @mzezman. Thanks for responding.

I have been making adjustments as we constantly try to balance efficiency vs this crazy Load shedding situation.  But I don't believe any setting I've tweaked should cause this change.

I do not believe I've had a firmware update, unless it would do that automatically without my involvement. Current versions are: M 3.3.7.4 / S  1.5.1.5 / C E.4.2.4.  I know I verified that the version was up-to-date when the system was installed (30 October 2022), but I see there is a newer version available now.

Last night my installer stopped by and checked physical wiring etc, which he said looked to be all good.  He was concerned that the inverter felt very warm. We've put a fan to blow over it all day today and see if that helps.  FYI - here is my temperature graph from yesterday - are these peaks a problem (DC:78, AC:57) ? Could this explain the degradation of panel generation capacity?

 

You can see where we added the fan - just after 6pm last evening.

 

How many panels do you have and what are there ratings?  How are they wired up, all on one MPPT or split across both?

Heat might be a factor, but if it derates the DC bus it will impact both battery and solar as both are DC so this does not fully explain the behavior 

On 2023/01/12 at 2:11 PM, Patrick OReilly said:

And today I see a similar pattern, though it did manage to get the battery to 100%. But since the rate of generation slowed about 9:00 am, it just won't come back up again, and now the battery is being depleted while the sun is blazing.

I have 3.27 KW solar capacity installed. yet I'm only getting 1.2 KW in the first couple of hours of the day, and less than 400W most of the time.

Please HELP...!!! 

 

 

For comparison, look at the behaviour on 1 November, 2 days after the original installation. That day we had intermittent clouds, but look at how strong the generation was throughout the day... Often generating at over 2KW.

It might be helpful to add the load to these graphs as that will influence the PV generation. If the battery is full and there is no load the PV will idle as there is nowhere for the power to go.

Sorry guys - I was traveling for a few days...

 

Quote

@mzezman: Heat could be a factor - see how it performs today with the fan on. I know the 5kw derates once DC temps go above 75.

We have run for several days with the fan blowing. Temps are consistently 15-20 deg C lower (good!), but I still have the same behavior.  

 

 

On 2023/01/13 at 10:49 PM, I84RiS said:

@I84RiS: How many panels do you have and what are there ratings?  How are they wired up, all on one MPPT or split across both?

I have 6 x 545W JA solar. 3 per string, which means they deliver about 130-132V per string once the sun gets going. SunSynk min string voltage is 125V. 

 

On 2023/01/14 at 10:17 AM, jumper said:

@jumper: If the battery is full and there is no load the PV will idle as there is nowhere for the power to go.

I'll include one here.  Anyway, I was aware of what you're saying, and was only looking at scenarios where SoC was not at 100% and load was higher than PV, hence draining the battery while the sun was shining...

This was another sunny day. In this example you can see it started well, got the battery charged by 9:00am. but from 10:30 onward load was consistently higher than PV, draining the battery with load less than 1KW most of the time, and PV staying below 500W, where it had reached over 1.5KW in the morning. It's like PV says "I'm only doing this once a day, then I'm chilling...".

On another note: I met another installer over the weekend and bounced this off him. He was thinking the BMS might be sending "mixed signals". Even though it reports SoC below 100%, it might still be throttling back on the input current. This sounds perverse to me - does anyone else have any thoughts on this?

Anyway, his two suggestions were:

1. Try reboot everything and see if that helps.
2. Tell the Inverter the battery is Pb (make sure all voltages are correctly configured) and let the system run like that for a while. If it then charges properly, this may indicate it is a BMS problem.

I'm not entirely comfortable with (2).  Even if the Inverter treats the battery as a PB and manages it by voltage, surely the BMS will still attempt to "protect" the battery?

Thoughts?

44 minutes ago, Patrick OReilly said:

545W JA solar

at 13Imp these aren't exactly ideal, also if they are all pointing in the same directiong, make them one string on one MPPT, since I reckon your Voltage may be skirting the low end here and be part of the issue...

Does your inverter have a fan, or only a big heatsink?

17 minutes ago, Kalahari Meerkat said:

at 13Imp these aren't exactly ideal, also if they are all pointing in the same directiong, make them one string on one MPPT, since I reckon your Voltage may be skirting the low end here and be part of the issue...

Does your inverter have a fan, or only a big heatsink?

I also zoomed in on panels that are very close to operating level. 

I was going to ask if the 132V is assumed voltage of 3 series panels or actually measured during the day. 

I for one would not run the inverter on 3 series per MPPT. 6 in series would be my call too if in the same direction or 4 panels if in different directions. 

Edited January 18, 20233 yr by Scorp007

I think the general consensus would be that you need more than 3 panels in series per string.

Looking at it from the point of view of the panels specs, the JAM72S30 545W panel at NOCT (representing real-world high temperature of 45degC) produces maximum power at Vmp = 39.20V, while Voc = 46.55V.

If you are running 3 panels, your Vmp is below the startup voltage of the inverter. With Voc being enough to start up the panels, the problem is that while you are operating at 130V-132V, or 43.3V-44V per panel, you are approaching the open-circuit voltage, at which the panel current is dropping current flow, and hence overall power.

My gut feeling is that if you add in enough panels in series that the Voltage is well above the minimum, the inverter will have the freedom to set the operating voltage at a point where the voltage across each panel is closer to Vmp. Pure speculation, but I'd propose putting all 6 panels in series on one MPPT, facing East. Or else one additional panel on each side.

Thanks @GreenFields, @Kalahari Meerkat and @Scorp007.

I have had someone advise me to try improving the strings before, even going 6x1 string rather than the 3x2 that I have. All 6 panels are on one roof, at the same orientation, which is NNE. I'm space constrained there, so more panels are not an option right now.

I did ask my installer about this, and he was quite adamant that this is not really an issue on a grid-tied system where the "start-up" voltage does not really apply (according to him, because the inverter is running already from grid, so does not need the panels/MPPT to "start up").  I, naturally, don't really know if what he's saying is correct or not. But as I now have 4 different people who have advised the same it's sounding like this is a must-try.

FYI, below is a graph of the voltages recorded today. I note that the voltage remains much the same in the morning (when the power generation is good) and in the afternoon (when it tanks...).

@Kalahari Meerkat: The Sunsynk 5KW inverter I have has a large heatsink behind it. To aid cooling I've positioned a fan on the floor blowing at it till I determine whether heat is the root cause or not. So far this has cooled the inverter, but has had no impact on the problem.

 

 

Edited January 18, 20233 yr by Patrick OReilly

I should note one other thing: For a long time (2 months) the battery dip switches were not correctly set, which means the Inverter and BMS were not communicating correctly. After correcting the dip switches (I forget the date, but it was early in this month) I've seen battery temps in the stats for the first time, and have observed how the BMS manages the rate of charge (slows it down) as the battery approaches full charge. I have noted this evening that the problematic behavior of this topic seems to have started roughly about the time that we corrected the dip switches, so that the Dyness BMS now knows it is talking to a Sunsynk (Deye). I'm not sure if this is at all related, but I thought I should mention it for full context.

Hi Patrick

Just out of interest: What is the battery temperature you are seeing ?

2 minutes ago, sjp said:

What is the battery temperature you are seeing ?

It ranged from 27 to 33 deg C today...

 

Edited January 18, 20233 yr by Patrick OReilly

6 hours ago, Patrick OReilly said:

So far this has cooled the inverter, but has had no impact on the problem.

Ok, my Sunsynk has a fan and runs a bit cooler than yours, pre your adding the external fan, but either way, since this seems to not have had any effect on your end, keep up the cooling, but rewire your six panels into one serial string on one of the two MPPTs and I suspect this should make things work more correctly. I think the higher temperatures at the moment (on the panels), together with the marginal Voltage on the 3S strings is causing things to not quite work as wanted.

@Patrick OReilly

It does seem as if your installer refuses to accept what the real problem is. Not all of us are easy to admit a problem. A grid tied does have the grid only to know at what levels for Volts and frequency to synch. In order to produce it does need the voltage as per the spec. If this voltage also drops below the minimum it will stop generating. In my case I can see it clearly and my grid tied produces a few watt up to 19h45 in the evening. At about the same time I can hear my Axpert hunting by seeing PV and then nothing. Then enough to start and then not enough. Clicking relays every couple of seconds. I then switch off my PV to stop this. 

Thanks for the voltage graph. We hope it is accurate. Confirm the voltage with a multimeter anytime during the evening. 

Edited January 19, 20233 yr by Scorp007

17 hours ago, Patrick OReilly said:

I have had someone advise me to try improving the strings before, even going 6x1 string rather than the 3x2 that I have. All 6 panels are on one roof, at the same orientation, which is NNE. I'm space constrained there, so more panels are not an option right now.

I did ask my installer about this, and he was quite adamant that this is not really an issue on a grid-tied system where the "start-up" voltage does not really apply (according to him, because the inverter is running already from grid, so does not need the panels/MPPT to "start up").  I, naturally, don't really know if what he's saying is correct or not. But as I now have 4 different people who have advised the same it's sounding like this is a must-try.

FYI, below is a graph of the voltages recorded today. I note that the voltage remains much the same in the morning (when the power generation is good) and in the afternoon (when it tanks...).

Definitely put all your 6 panels into a single string and onto 1 MPPT. That will beef up your string voltage - which I suspect will solve your problems. My guess is that your current string voltage is so low that the inverter is unable to ramp up the MPPT when your load increases during the afternoon.

Get someone to re-wire your panels and see if your system then runs as expected.

20 hours ago, Patrick OReilly said:

On another note: I met another installer over the weekend and bounced this off him. He was thinking the BMS might be sending "mixed signals". Even though it reports SoC below 100%, it might still be throttling back on the input current. This sounds perverse to me - does anyone else have any thoughts on this?

I have the same behavior on my Sunsynk with Pylons. Once the Pylons get around the 'famous' 89% SOC, the BMS throttles the charge current. I'll let the experts here explain the details, but it is normal behavior.

But this is not the reason for what you are experiencing.

20 hours ago, Patrick OReilly said:

Anyway, his two suggestions were:

1. Try reboot everything and see if that helps.
2. Tell the Inverter the battery is Pb (make sure all voltages are correctly configured) and let the system run like that for a while. If it then charges properly, this may indicate it is a BMS problem.

I'm not entirely comfortable with (2).  Even if the Inverter treats the battery as a PB and manages it by voltage, surely the BMS will still attempt to "protect" the battery?

I would not do that either. The Sunsynk can directly communicate with your Dyness BMS, so why would you take away that ability?

2 hours ago, wolfandy said:

Definitely put all your 6 panels into a single string and onto 1 MPPT. That will beef up your string voltage - which I suspect will solve your problems. My guess is that your current string voltage is so low that the inverter is unable to ramp up the MPPT when your load increases during the afternoon.

Get someone to re-wire your panels and see if your system then runs as expected.

 

Actually on a Sunsynk with the 2x3S string each going to a MPPT on can just do a series connection at the inverter and connect the 2 remaining wires into MPPT1. Negative from MPPT1 connected to positive of MPPT2. Negative now on MPPT2 goes to negative of MPPT1. Positive on MPPT1 stays put. Easy 🤔🤔🤔

Edited January 19, 20233 yr by Scorp007

17 hours ago, Scorp007 said:

Actually on a Sunsynk with the 2x3S string each going to a MPPT on can just do a series connection at the inverter and connect the 2 remaining wires into MPPT1. Negative from MPPT1 connected to positive of MPPT2. Negative now on MPPT2 goes to negative of MPPT1. Positive on MPPT1 stays put. Easy 🤔🤔🤔

@Scorp007: Good idea for an easy test. If it works well then it is worth the effort of shortening the wires by linking the strings on the roof instead.

I lost track in your words 🙂, but this is what I understand:

• MPPT 1 Negative connected to String 1 Negative.
• String 1 Positive series connected to String 2 Negative.
• String 2 Positive connected to MPPT 1 Positive.
• MPPT2 unused.

Thanks to everyone for your input.  I'll keep you posted on results...

4 hours ago, Patrick OReilly said:

@Scorp007: Good idea for an easy test. If it works well then it is worth the effort of shortening the wires by linking the strings on the roof instead.

I lost track in your words 🙂, but this is what I understand:

• MPPT 1 Negative connected to String 1 Negative.
• String 1 Positive series connected to String 2 Negative.
• String 2 Positive connected to MPPT 1 Positive.
• MPPT2 unused.

Thanks to everyone for your input.  I'll keep you posted on results...

Disconnect wires 1,2 and 3.

Connect 1 to 2.

Connect 3 as per arrow to MPPT1 negative. 

Dotted lines means disconnect. 

 

23 hours ago, Scorp007 said:

Disconnect wires 1,2 and 3.

Connect 1 to 2.

Connect 3 as per arrow to MPPT1 negative. 

Dotted lines means disconnect. 

 

A picture says a thousand words!

Thanks @Scorp007 - that is very clear. My installer is suddenly on board with this idea (not sure what changed since I last mentioned it to him), but we'll do this change tomorrow. I'll definitely let you all know what transpires.

37 minutes ago, Patrick OReilly said:

My installer is suddenly on board with this idea (not sure what changed since I last mentioned it to him), but we'll do this change tomorrow. I'll definitely let you all know what transpires.

I'd do this after sunset, if I were you, and see tomorrow morning how things look... either way good luck and lets hope it mostly sorts you out...

On 2023/01/18 at 9:39 PM, Patrick OReilly said:

I did ask my installer about this, and he was quite adamant that this is not really an issue on a grid-tied system where the "start-up" voltage does not really apply (according to him, because the inverter is running already from grid, so does not need the panels/MPPT to "start up").

He is talking rubbish. The MPPT's startup voltage is 125V, after this it will generate a trickle, and it needs a few more volts than that (about 25 more in fact) to produce its maximum. You haven't provided your panels' specifics (just saying 545W isn't good enough, JA Solar makes hundreds of 545W panels :-)) , but as long as combined VoC does not exceed 450V, you should consider putting all 6 on one series string, and on one MPPT. The advantages are that it will start producing earlier in the day and end later, and overall efficiency (higher voltage means less voltage drop).

On 2023/01/18 at 6:09 PM, Patrick OReilly said:

Sorry guys - I was traveling for a few days...

 

We have run for several days with the fan blowing. Temps are consistently 15-20 deg C lower (good!), but I still have the same behavior.  

 

 

I have 6 x 545W JA solar. 3 per string, which means they deliver about 130-132V per string once the sun gets going. SunSynk min string voltage is 125V. 

 

I'll include one here.  Anyway, I was aware of what you're saying, and was only looking at scenarios where SoC was not at 100% and load was higher than PV, hence draining the battery while the sun was shining...

This was another sunny day. In this example you can see it started well, got the battery charged by 9:00am. but from 10:30 onward load was consistently higher than PV, draining the battery with load less than 1KW most of the time, and PV staying below 500W, where it had reached over 1.5KW in the morning. It's like PV says "I'm only doing this once a day, then I'm chilling...".

I do not see too much of a problem with the above graph, the batteries are charging, and once they around 80-90% the BMS will throttle the inverter back in any case. Your loads are generally very low: 300-500W, just the 3 demands at around 12H00, but they were not sustained. The MPPT controller is fairly slow to respond to sudden demand, so the 3 spiky bits around 12H00 will initially be taken from the grid (or another source) in anycase, and the PV will slowly be combed in, so that is something to take into consideration.

I agree that it could in fact be that the PV string's voltage is borderline. As load increases, it also has the effect of pushing the voltage lower, and this, coupled with panel heating later in the day, could be pushing it lower still, causing the MPPT to become reluctant to produce. The only way to check is to either provide a continuous, consistent load, or to export to grid (when no load shedding), that way the PV can produce to its max possible and the system can be troubleshooted. At the risk of stating the obvious, it's also a good idea to keep an eye on cloud movements outside too😄