@gabriel the switching times between solar and grid is largely dictated by your battery capacity and of course the weather. I have 450AH capacity and switch to solar at 1:30 and to grid at 17:00 at present. My batteries were at 98.3% at 17:00 yesterday so some grid power was used to bring them to 100% during the evening. This happens because I have parameter 32 set to 90 minutes to force grid charging to make sure the batteries are at 100% before supplying power at 1:30 every day. This is in addition to a BMV700 controlling my RPi ICC.

This morning at about 8:00 I switched to grid as the batteries were at 80.4% with 100% cloud cover and even a few drops of rain as I am sure you have too. This happens on cloudy days and I will check later during the day to see how solar generation is doing and if sufficient and depending on battery charge state, I may switch back to solar which ICC with my present settings will switch to grid at 17:00. It is not exact science but my objective is to use as little as possible grid power for battery charging. Given that the weather is unpredictable, some human intervention is required. So I guess a fully automatic system is not possible yet unless ICC becomes substantially more complex. However it is not a hardship to check on ICC at times during the day and intervene if necessary. I hope this helps but feel free to ask if I can help further.

3 minutes ago, ebrsa said:

to solar at 1:30

i.e. 13:30 i presume, and if so why not earlier, say 10:00?

where do i do the setting to start getting my batteries to 100% soc, as i have also changed over to grid via icc because of the clouds and yet my batteries are still losing soc/discharging, see images

thanks!

ps we had more than 4mm rain since 19:00 yesterday, how are things there?

 

2 hours ago, gabriel said:

what i cant figure is why my battery soc slopes downward over the last 24h

... because you're drawing around 1 amp from them?

It's pretty common, I easily lose about 1% SoC overnight because the Multi needs around 10W and the CCGX and MPPTs have a combined draw of around 6W or so.

According to your Battery settings your system should have switched to grid at 5pm but your SOC graph doesn't show that happening - so it either didn't change to grid, or it did change and something made it change back.

If you look at battery volts on the middle graph, it appears that it did change for a short time around 5pm, and then changed back (if you switch off the pink amps graph then it will be easier to see). I think something caused your battery voltage to spike up to the Axpert back to battery voltage, overriding the ICC time control. But I dont use Time Control at all, so not very familiar with it.

 

Oops, I answered your 1st post and didnt see additional info/posts until after answering.

4 hours ago, gabriel said:

i.e. 13:30 i presume, and if so why not earlier, say 10:00?

Actually I switch to solar/batteries at 1:30 or one and a half hours after midnight. The system then runs on solar/batteries until 17:00 on that day. In the morning, on sunny days when the panels produce enough output to power the house base load of around 800W, the batteries are near 80% SOC and as more current is generated, the batteries are recharged. I prefer not to discharge below 80% SOC, except when we have a grid blackout, to try and get maximum life out of the batteries.

I have flashed both my Axperts (Proline branded) with the Coulomb/Weber firmware ver. 73.00 beta firmware and have been using all their previous versions over time. As my batteries are Trojan T105, that requires a bulk charge voltage of 59.3V and the Axpert maximum charge voltage is 58.4V, I have set parameter 32 to 90 minutes. My batteries are usually at 100% SOC when switching to grid at 17:00. However because of the parameter 32 setting they get a boost from the grid as soon as the solar panels stop producing power. The charge current falls very quickly to about 4A and decreases marginally for the full 90 minutes with the batteries at 100% SOC. That is less than 0.3 KWh per day so probably worth it. Parameter 16 is set to CSO or solar before grid. I am only doing this because the Trojan factory in the USA advised that I may charge at a voltage lower than the recommended bulk charge voltage of 59.3 but then for a longer time. Also if the panels did not fully charge the batteries, the shortfall is made up with grid charging.  So when switching to solar/battery at 1:30 at night, the batteries are still close to 100%SOC.

Today is of course a train smash with the cloud cover, with no rain for some hours and now at 14:30 my batteries are only at 88.3% SOC. Also I have been running off grid since 8:00 so the total output of the panels is used to charge the batteries. However the parameter 32 setting will ensure 100% SOC before switching to solar again at 1:30.

I have no idea how much rain we had last night as my weather station packed up a while ago and I have not replaced it yet. Still trying to find something that does not break the bank. A Davis will nice but is just too expensive for my usage. Perhaps something repairable based on Raspberry Pi, but so far I have not found anything that looks promising.

43 minutes ago, ebrsa said:

I have flashed both my Axperts

i will have to try my hand on this as my axpert only has 31 parameters - and some cant be set.

i attach a pic of the relevant axpert settings, are both of your boxes ticked as well? i initially thought its an either or situation until i re-read the icc manual.

i agree with you on treating the batteries with kid's gloves as they are pretty pricey, in fact the other day i had to replace one of mine!

 

51 minutes ago, ebrsa said:

A Davis will nice but is just too expensive

well, as with many a thing i had to google that, and yes, the price is ok if that is your profession; but as for me an ordinary agri rainmeter is just fine.

thanks for sharing your info!

@gabriel I only have Use time for control ticked as from past expereince it seems that one must use either one or the other but not both options. Another scenario that I have considered is to use SOC for control with To Solar set at 100% and To Grid at 80%. Then also set the Axpert parameter 16 to OSO, solar charging only, and set the BMV700 to charge for 90 minutes at bulk charge voltage. That would ensure that grid is not used for battery charging and batteries won't be used before they have been fully charged. The only downside is that my wife uses appliances with high demand for the evening main meal cooking. Drawing such large currents from the batteries only reduces their AH capacity. By using the time based switching, the house is already on grid when she starts cooking and we also use the dishwasher when on grid. The washing machine uses little power as little heating is needed. I should add that hot water is heated with a 12 evacuated tube solar heater controlled by a Geyserwise. My Owl energy meter on the geyser shows that so far about 0.5 KWh was used for water heating this year. Last month my total grid account was for 261 KWh so I don't mind switching the Axpert manually between grid and solar as required on cloudy days.

Flashing the Axperts with the Coulomb/Weber firmware has been quite easy and I have also done it twice for a client of @Mike in the village. The instructions are included in the 73.00 firmware download. You will require a USB to Serial converter unless the computer you use has an RS232 serial port. I use a laptop with the battery fully charged as a power failure during flashing will leave you with a dead Axpert. However there is always risk that things may go wrong although I have thankfully not had any disasters so far.

2 hours ago, ebrsa said:

I only have Use time for control ticked as from past expereince it seems that one must use either one or the other but not both options

@ebrsa now that makes sense to me! in that regard the icc manual seems ambiguous, @Manie i'd like your opinion here please.

questions:

the rs232 port on the axpert has to be used for the firmware update?

can i use a 5e network cable as my pc is 15m removed?

must this network cable be converted to usb for the pc or can i use my pc's lan port?

and for now last question:

how can i force either grid or pv to get my batteries up to 100% soc as they are at 77% and falling?

pardon any 'stupid' questions, but it seems to me i should have stayed in the 'solar for beginners' section...

 

 

@gabriel except for the experts in this business, I guess we all live and learn. 

The RS232 port on the Axpert looks like a RJ45 LAN cable port and the unit comes with a serial cable with an RJ5 plug and a 9-pin RS232 female plug on the other side. If you do not have a laptop that has an RS232 plug built in, which was the case on older laptops, you will need an RS232 to USB converter. That must be plugged into the Axpert serial cable and gives you a USB plug to plug into the laptop. The other side ie the RJ45 plug of the Axpert cable plugs into the serial port of the Axpert which is actually an RJ45 LAN plug. As long as you use the Axpert cable, I cannot see how you can go wrong. Study the documentation carefully and follow it to the letter. Also it takes about 8 to 10 minutes to flash the Axpert inverter firmware. If you also have to flash the MPPT solar controller, make sure you use COM1 on the laptop. It is all well described in the documentation. Also check the website on which @Coulomb has posted the firmware, the link is http://forums.aeva.asn.au/viewtopic.php?p=66665#p66665 for lead acid and all other batteries except LFP batteries. I have been using the Version 73.00 firmware for almost 3 weeks and it works flawlessly.

For flashing the inverter firmware, I always turn all power connections to the Axpert off, that is grid, batteries and solar panels. When flashing the MPPT firmware, the batteries and solar panels need to be connected. For flashing the inverter firmware only the batteries should be connected. However study all the documentation carefully.

To get your batteries to 100% SOC you could set parameter 16 on the Axpert to SNU as that will allow solar and grid to charge the batteries. Also set parameter 13 to the fully charged voltage of your batteries. It seems to me from the screen captures you posted above that you are using ICC for Windows. I on the other hand use a Raspberry Pi with a Victron BMV700 connected to control my 2 Axperts connected in parallel and monitor the Axpert with the ICC MQTT app on a Windows7 computer in my study.

Just ask if something is unclear.

27 minutes ago, ebrsa said:

I guess we all live and learn

yes, and even the experts lurk here

thanks for all the info @ebrsa, i will definitely be the wiser for it!

God bless on this Good Friday!

g

11 hours ago, gabriel said:

 

the rs232 port on the axpert has to be used for the firmware update?

Yes. You can't use the USB port, for reasons that make no sense to me.

11 hours ago, gabriel said:

can i use a 5e network cable as my pc is 15m removed?

Hmmm. I guess you could use a CAT-5(e) extension cable (male to female) if it's a pain to move your PC, and you don't have a laptop handy.

11 hours ago, gabriel said:

must this network cable be converted to usb for the pc or can i use my pc's lan port?

It's not a LAN connection, it's RS232 on the wrong connector. I wish they wouldn't break standards like that. So yes, you need the USB to serial adapter, and you can't use the LAN port on your PC.

11 hours ago, gabriel said:

how can i force either grid or pv to get my batteries up to 100% soc as they are at 77% and falling?

If you install the patched firmware, that should happen automatically. In the meantime, the easiest way is probably to force a grid charge by changing setting 01 (output source priority) to Uti and setting 16 (charge source priority) to Cut. Probably have ICC disconnected when doing this, as it may override those settings. When done, put the settings back to what they were, or just reconnect ICC and it will eventually set them back again.

thanks for the info @Coulomb, i'll use my wife's laptop to patch the firmware.

my axpert's charge light indicates my batteries are fully charged, yet my soc on the icc indicates 77% and declining - is there something i miss here?

@gabriel I will answer your questions on this thread including the ones in your personal message so that others can comment if they don't agree or have  better solutions.

Below is a screenshot of a 100AH Sonic battery that I found with Google.

Your Sonics have an capacity rating of 234AH if discharged over 10 hours as is reflected on the data sheet. That is more than double that of the illustration above and would imply that a maximum charge current of 50 Amps would be acceptable for you. However gel batteries should not be charged at too high a rate as it will cause gassing with premature failure resulting. Personally I would limit the charge current to 40 Amps just to be safe. To control that you should set parameter 2, maximum solar and grid charge current on the Axpert, to 40A. You should also set parameter 11, maximum grid charge current to either 40A if you want the grid to top up batteries in case you panels did not fully charge them or 2A if you just want the grid to trickle charge..

Since I don't know if you have a Victron BMV700 or 702 connected to your computer on which ICC Windows is running, it is a bit tricky to suggest voltages on the Axpert which is notoriously useless in doing a good job with voltage related matters. I would recommend that you invest in a BMV702 and the special Victron USB cable for it. Someone on the forum sold them in the past for a good price. If your computer is 15 meters from the Axpert, you are going to need a USB extender kit togeter with a CAT5 LAN cable of suitable length.

Here are the best suggestions that I can come up with but others may not agree. Make sure parameter 5 is set to USE  as this will allow you to set parameters 26,27 and 29. Parameter 12, voltage back to grid, should be set to a low value or it will interfere with ICC, perhaps 44 or 45V if you have a BMV, otherwise perhaps 48V. Set parameter 13 to FUL, 16 to what suits you. Mine is set to CSO so that solar will charge as a first priority and grid will charge only if no solar is available. Set 26 to 58.4V, the highest available but perfect for your Sonics based on the bulk charge range of 14.4V - 14.9V as it equates to 14.6 x 4 = 58.4. Parameter 27 you could set to 54.8V which again is in the middle of the range on the picture above or 13.7 x 4 = 54.8V. Parameter 29, the low battery cut off voltage you could set to 44V or 46V.

I hope this helps but I have found that time based control of ICC, with some manual intervention, works well for me. My objective is to use as little grid as possible and also make sure that my batteries are at 100% SOC when ICC switches to solar/batteries at night. As I explained before, I also try to avoid drawing very high currents like more than 40A from the batteries. Since I have 2 strings of 225AH each in parallel that is only 20A per string.

Regarding your question why your batteries SOC stays at around 77%, it is a failing of the Axpert. Only by installing a BMV700 or 702 will you know the true state of charge and ICC will have better control over the Axpert. They are a bit expensive but worth it compared to replacing a set of batteries prematurely.

hi @ebrsa, 1st off, thanks for a thorough reply, must have taken you some time, i appreciate it!

my batteries are as in your image, just the 250ah [nps 250-12]  model.

at present i have max charge current on 40amp as per suggestion of @plonkster, which is 16% of my bank ah rating, so parameter 2 is sorted out, it also matches your suggestion. parameter 11 is also set at 40amp as that should not exceed the no 2 setting according to @Don's manual on the axpert. whilst mentioning his name, does someone have a copy of his 'solar terminology' document?

let me mention at this point that i have a june 2015 axpert 5kv which still has the old firmware - i'm waiting for the usb to serial converter to update it. i do also have the bmv702 , icc and the ah02 battery balancer, so in that department things look not too bad.

back to your suggestions. i will set par 5 to use, it is set for agm at the moment which fixes bulk charge to 56.4v and float to 54v.

due to the fact that i have the bmv702 i will set par 12 to 45v, it was on 48v.

par 13  gives me no option of FUL and i will leave it on 55v - maybe this is something which the 73.9b firmware addresses.

par 16 is set to cso, same as yours due to mutual objective.

the rest of the settings in your 3rd last paragraph will be set now.

as soon as i have updated the firmware as mentioned above i will let you know if things run more smoothly.

things which bug me at this stage includethe fact that although the axpert indicates a fully charged battery on the led display, icc [bmv702 value] reports about 70.3% and declining soc.

God bless

g

 

 

 

 

 

On 3/30/2018 at 6:24 PM, ebrsa said:

To get your batteries to 100% SOC you could set parameter 16 on the Axpert to SNU as that will allow solar and grid to charge the batteries. Also set parameter 13 to the fully charged voltage of your batteries.

that did not work, the soc still dropped.

i also changed the other settings [program 5 to use etc] as you mentioned which also saw no improvement in the soc.

soc is now at 69.3 and falling although the batteries just draw 1amp, see pic. the battery voltage now is 54.93 and the back to grid parameter is set to 48v, so i presume charging will only commence then - or should i rather click the icc "use soc for control" with a 80 and 100 setting and see what happens? maybe the axpert will only start reacting correct with the new firmware?

ok so i clicked the icc soc control as just mentioned and set the axpert as before, see pic - lets have a look if i can get my soc up again

Hi @gabriel,

There seems to be a problem with one of your reporting devices, either Axpert or BMV is lying and I am beginning to suspect the BMV (because we all know that in an Axpert/Victron environment the failure is highly unlikely to be on the Axpert side).

Looking at your Dashboard output above something doesn't quite gel - A lead acid battery at 69.5% SOC and that is busy discharging is never going to be at 54.94V, the numbers don't tie up.

Please send a pic of your [inverter/bmv values] page so that we can compare the outputs from the 2 devices, and also measure the bank voltage with a multi-meter at the same time and charge/discharge state as when the screen grab is taken, the multi-meter will be used to catch the liar. 

The BMV is an incredibly sensitive measurement device. Consider for a second that that is a 500 ampere 50mV shunt, in other words at 500A it only has a 50mV drop, so at 1 ampere there is a 100nV drop that has to be measured. It's so tiny most oscilloscopes can't even measure it. There have been cases where BMVs required recalibration, and if you check the manual it explains how to do it. It usually ships calibrated from the factory and you should not have to do anything... but if you have to, the screw is there if you need to turn it :-)

2 minutes ago, plonkster said:

The BMV is an incredibly sensitive measurement device.

I have found the BMV to be indispensable to the correct operation of my Axpert installation - but something is providing incorrect data in @gabriel installation and checking the data output against a manual multimeter reading should expose the culprit.

13 minutes ago, plonkster said:

but if you have to, the screw is there if you need to turn it :-)

If the voltage is correctly displayed then the SOC is way off, if the SOC is correct then the voltage is way off.

My guess is that amps in/out are no longer being correctly reported so the voltage shown is correct and the Axpert is reporting fully charged, but amps measured is sitting at -1 all the time so SOC as per BMV is declining (actual SOC is much higher).

But it is all still a guess until the requested info is provided.

ok folks, i appreciate you looking at this,

batteries are 13.4v 13.4v 13.5v 13.5v

axpert shows fully charged

images attached include present axpert settings

 

27 minutes ago, pilotfish said:

I have found the BMV to be indispensable

Oh agreed, I'm just saying... it's a sensitive thing, and sensitive things may need calibration. I looked up the specs after typing that post, that thing can measure down to 10mA, we're talking microvolts.

First step would be to check if the ICC reported number and the number on the BMV's LCD correlates. Just to rule that out as well.

7 minutes ago, plonkster said:

First step would be to check if the ICC reported number and the number on the BMV's LCD correlates

they do

The measured voltages agree with the voltage output of both Axpert and BMV, now do me a favour and put the Axpert into Battery mode so that we know there is a current being drawn, and lets see what is reported by BMV

I think that the BMV is not measuring current at all and defaulting to -1 so SOC slowly drops off.