diy solar

diy solar

SRNE Odd MPPT Charging Behavior - Pause Every 12 Minutes?

Mattb4 said:
From your symptoms it seems more like #1 since your icon for PV goes off (disconnect from array).
Click to expand...

I think that's a question you should ask whoever programmed this algorithm. The PV icon flashes during this procedure, this does not necessarily mean that the PV system is disconnected, it may mean that a scan is in progress. Only the person who programmed this algorithm knows that.

If the PV system was completely disconnected the PV power would drop to 0W for a moment which is not the case. The PV power goes from 202W to 161W before rising to 217W. For information, the measurement of the current A given by the EAsun for the PV input is false, in reality it is always a little more so at the lowest it can be 175W in reality.

The battery power graph coming from the BMS also shows that the charge decreases but does not stop. So I think it is indeed a full sweep without disconnection as mentioned in point 2.

I won't be at home for the next few days but I could do an additional check on Wednesday with an ammeter clamp and the return of the sun :cool:

1676212051548.png
 
  • Like
Reactions: mib
Sleeper85 said:
I think that's a question you should ask whoever programmed this algorithm.

...
Click to expand...
LOL. Programmers and why they choose to do something a certain way is beyond me to fathom. Brilliant folks by and large but their thought process is not similar to mine.
 
What's interesting is that @Aerial Gopher who had a SRNE 3Kw 24v system, 120 volt out model. ( HF2430U60-100 ? ) could see this scan every 12 minutes in Solar Assistant. Was the PV icon flashing at this time ?

Edit:

It appears to be the model HF2430U60-100 he purchased in May 2022 see post below.
It would be interesting to know what firmware his AIO had but we probably won't know because he sent it back to buy a Growatt.

diysolarforum.com

Wiring SRNE inverter/charger to a distribution panel

Hello, I have a SRNE HF2340U60-100 all in one inverter. It is going in an off-grid cabin, with a 120V generator wired to the AC input, 1400W solar array, and 120V AC output wired to a distribution panel. The problem I am having is that there is no ground terminal on the AC out, and I can't get...
diysolarforum.com diysolarforum.com

And by reading this post we can learn that MidNite sells the 24V-3000W and 48V-5000W 120V AC model for $785 / $1246 maybe with a magic firmware ? :rolleyes:
Reading the MidNite manual will convince you that it's indeed an SRNE.

Category Products - MidNite Solar's

MidNite Solar Category Products.
www.midnitesolar.com
 
Last edited:
  • Like
Reactions: mib
Trying to make special watch of my EAsun AIO it does have a MPPT cycle where it drops off loading (and the fans cycle down) for a bit and than starts back up max PV output. It does not cause the PV icon to flash off and it also seems to not drop wattage entirely. However if it is on a certain repeating time period I can not say. One issue is the battery charging program can also cause a change in PV demand as it polls the DC bus.

Without a graphing program I can not be definitive.
 
While it's demonstrating something strange, I think assessment of what's going on needs data collection at a much higher frequency than the 6 Hz Solar Assistant is capable of showing.
 
I know that @mib is tempted to connect his SRNE to Home Assistant, maybe we can have a nice graph soon?

The wifi datalogger will not allow this to be seen as it only takes data every 5 minutes.

If you are interested, I could make you a script in Python to analyze the data via modbus by connecting your PC to the USB port of the AIO?
 
  • Like
Reactions: mib
wattmatters said:
While it's demonstrating something strange, I think assessment of what's going on needs data collection at a much higher frequency than the 6 Hz Solar Assistant is capable of showing.
Click to expand...

Solar Assistant collects data every 2sec for the dashboard versus 10sec in the charts.

Via a Python script I could get even more info like every 0.5sec.
 
  • Like
Reactions: mib
Sleeper85 said:
I know that @mib is tempted to connect his SRNE to Home Assistant, maybe we can have a nice graph soon?

The wifi datalogger will not allow this to be seen as it only takes data every 5 minutes.

If you are interested, I could make you a script in Python to analyze the data via modbus by connecting your PC to the USB port of the AIO?
Click to expand...
:) Thanks for the offer but even though I am a Linux user of a couple of decades (computer user going back to DOS 5.0) now my abilities to understand how all these things operate is limited. I would say I am pretty much computer illiterate beyond simple follow step by step the recipe type of interactions.
 
wattmatters said:
at a much higher frequency than the 6 Hz Solar Assistant is capable of showing.
Click to expand...
Not sure what I was thinking, SA doesn't show at 6 Hz.
At best it's 0.5 Hz (display), while charting is at 0.1 Hz.

Sleeper85 said:
Via a Python script I could get even more info like every 0.5sec.
Click to expand...
That really only helps if the data stream being captured is updated with a high enough frequency to make that worth while.
 
  • Like
Reactions: mib
Sleeper85 said:
I know that @mib is tempted to connect his SRNE to Home Assistant, maybe we can have a nice graph soon?
Click to expand...
For now I'm just dreaming :)

maybe it can be seen with current clamps on the pv line?

I have two mppt srne ml2440 there is no such problem, for the reason that I am very satisfied with ml2440 I bought AIO Srne :)
 
Attached is the script, it works for me from Linux after installing the "minimalmodbus" lib.
Should also work on Windows. In any case, the COM port address (/dev/ttyUSB0) must be adapted to your environment.

I set the time to 1sec, this can be decreased but I think 1sec is good because it needs time for AIO to write to modbus registers.

You can see what's happening on screen and also in a CSV file (pv.csv in the same folder as the script) which you can open in Excel or OpenOffice to analyze the data.

CTRL+C to stop the script.


Installation of minimalmodbus on Debian/Ubuntu.
Code: 
sudo apt install python3-pip
pip3 install -U minimalmodbus

On Linux, connect your PC to the AIO via the USB port then give access rights to the COM port before running the script.
Code: 
sudo chmod 666 /dev/ttyUSB0

Run the script
Python: 
python3 pv.py

1676236696556.png

The information collected in the pv.csv file

1676236547547.png

And if you want to read other info, I also attach the SRNE modbus protocol with the list of registers.
 

Attachments

  • srne_pv_stats.zip
    774 bytes · Views: 5
  • SRNE hybrid solar inverter MODBUS protocol V1.7.pdf
    445.9 KB · Views: 10
Mattb4 said:
Trying to make special watch of my EAsun AIO it does have a MPPT cycle where it drops off loading (and the fans cycle down) for a bit and than starts back up max PV output. It does not cause the PV icon to flash off and it also seems to not drop wattage entirely. However if it is on a certain repeating time period I can not say.
Click to expand...

This is exactly what is happening, the reduction of the fans until the complete stop then a resumption of maximum power load.
Except that during this phase, my PV icon flashes => it's a firmware improvement :ROFLMAO:

I took a video for you => SRNE MPPT sweep video

Below are the graphs corresponding to the video.

The next time I measure the current on the PV input, if there is a total disconnection when the voltage rises near Voc then the current from the PV must be 0A and not 0.9A.

1676326708126.png

1676325056414.png

1676325076030.png

It's still a loss of production of 2 minutes, I find it ugly!

16763251207955


Data from the BMS :

Here too the BMS still registers a load of 346W at the lowest point. If there is a total disconnection, this charge may come from the capacitors ?
_
 
Last edited:
  • Like
Reactions: mib
Between the two sweeping phases A and C, we can see that the MPPT is able to automatically adapt the voltage by decreasing it to get more power. Why make such a brutal sweep every 30 minutes ?

10_PV_Volt - Scan every 30min.png

1676327992320.png

1676328052991.png
 
  • Like
Reactions: mib
Measurement with the current clamp. The current coming from the PV is at 5.55A before the "sweep" begins, it suddenly drops to 3.69A then gradually decreases to 1.34A (0.8A on the EAsun screen) while the voltage rises towards Voc then the current rises suddenly towards +/- 3.5A while the voltage falls then rises towards Vmp.

=> Video of the measurement with the current clamp
 
  • Like
Reactions: mib
The Victron MPPT SCC seem to operate the same way.

So, clearly on a beautiful cloudless sunny day these sweep do nothing.

On the other hand, during a day with a very change weather with sun and clouds we could gain in power.

But it's possible to make more efficient algorithm.

A parameter that might be useful would be to enter the max Vmp voltage of the string to prevent the MPPT from searching so high to find nothing.

Change MPPT Tracking Refresh Interval - VictronEnergy

community.victronenergy.com community.victronenergy.com
 
  • Like
Reactions: mib
During a mixed day with little sun, the power losses are less visible on the "PV Power" curve and I would even say that the sweep is beneficial if we look at what is happening at 13:54:10 when a sweep begins.

1676581118159.png

1676581061479.png

1676581176985.png
 
  • Like
Reactions: mib
Aerial Gopher said:
The array was 610 ( 2x310 ) and now is 1500 watt with more panels, it has been doing this before as well as now.
The max PV input is 1600 watt, and max Voc is 100vdc which the panels and set up is below both those limits.

So I have had a small array, and one that near maxes out the charge controller.

On a 170 watt load on the inverter, the total draw from batteries is between 80 - 90 watts more to run the inverter. Was not happy about that.
Im in talks with the seller on Aliexpress, and they claim all of this is normal, but we have all been down that road before.
Click to expand...
Could you help on how you wired your panels...a string of 2 panels in series and parallel?
 
Some people don't like seeing this MPP tracking every 30 minutes.

Below is what a Deye Hybrid does not do.

Same string, just a different inverter.

EAsun 27.05.2023 :

11_PV_MPPT_Power_Curve_27.05.2023.png

11_PV_MPPT_with_shadow_27.05.2023.png

Deye 06.09.2023 - Very nice curve but...

MPPT Power Curve.png

Production loss !

MPPT shadow Production loss.png

Deye test of 08.09.2023 :

MPPT shadow Power.png

MPPT shadow Voltage.png
 
Sleeper85 said:
Production loss !

MPPT shadow Production loss.png
Click to expand...

Each one of those chart grid squares is equivalent to ~ 42 Wh. I estimate approximately nine squares worth between your estimated pink missing production line and the actual yellow production line, which is 0.375 kWh.
 
mirekcafa said:
SRNE (SMXII 5,6KW) the problem is software V6.17. With software V6.35 the curve is already fine
Click to expand...

Interesting, so how does he react in the presence of shadow?

Actually it's not a problem, it's just that it searches for a better MPP every 30 minutes.

If with 6.35 he no longer does it then perhaps he is no longer looking for a better point or is he doing it differently?
 
You must log in or register to reply here.

Similar threads

A
SRNE Inverter Odd MPPT Charging Behavior - Pause Every 12 Minutes?
Replies
0
Views
1K
Aerial Gopher
A
Sleeper85
SRNE AIO - Winter production and MPPT behavior at low light
2
Replies
25
Views
2K
sethmad
S
I
Wind turbine 2000w charging 3 phase 15v on 4 x 12 v 105ah AGM battery's 48v, 4 x 375w solar panels MPPT 48v, Can I use a 24v inverter is it possible.
Replies
1
Views
96
sunshine_eggo
sunshine_eggo
B
Odd discharging, not sure if this is normal or not
Replies
10
Views
845
TomC4306
TomC4306
consolelog
Victron Orion Voltage Drop
Replies
22
Views
816
400bird
4

diy solar

diy solar
Back
Top