HI @Coulomb

 

I see the LC and LF firmware.

 

Question - what would the implications be if the "wrong" version is loaded?

 

Operationally, what is the difference in the firmware?

 

Thanks,

 

Gert

12 hours ago, Gert-Claassen said:

Question - what would the implications be if the "wrong" version is loaded?

If you loaded LC and needed LF or vice versa, it would still work, it would just be a nuisance.

12 hours ago, Gert-Claassen said:

Operationally, what is the difference in the firmware?

All it does is changes some thresholds and constants.

For example, it's easier to get back to bulk charging with an LFP battery, because you don't have to wait for the battery voltage to drop by 4.0 V, only 1.2 V. 4.0 V below the float voltage setting in an LFP battery would make it very close to dead flat.

The main change is the threshold for exiting absorb stage to float stage. For lead acid, this is C/5 amps, but to fully charge an LFP battery, C/12 is more appropriate, so that's what we use.

This old post has a summary of most of the changes. See the AEVA topic index for links to the documentation for the patches, e.g. the AussieView™ and KettleKomp™ manual.

Edit: re the worst case if you load the wrong firmware version (patched or factory), the worst thing is you have to get a new control board. These are "only" R1000 or so, but it's a royal pain getting the exact right one, sending photos of the sticker and endless emails and delays.

Edited August 7, 20214 yr by Coulomb

Hi @Coulomb - Thanks for your answer and sharing the article - great insights of what is changed.....

 

OK, so I guess I need to add this portion to my question:

 

I used to use 4 x 12V Lead Acid batteries. But have now replaced them with a lithium 15S bank, with a Daly BMS on it.

 

So it works great... - Do I need to change my inverter firmware to accommodate the Lithium pack now, or just leave it as is?

 

Thanks,

 

Gert

On 2021/08/16 at 11:06 PM, Gert-Claassen said:

So it works great... - Do I need to change my inverter firmware to accommodate the Lithium pack now, or just leave it as is?

It would benefit from the LFP flavour of the patched firmware. The factory firmware doesn't make any concessions for the different voltage behaviour of an LFP battery. That's if patched firmware is available for your model.

Hi @Coulomb

 

Currently running "LC1 74 40E" (as per LCD display).

 

What would you recommend?

 

Gert

18 hours ago, Gert-Claassen said:

Currently running "LC1 74 40E" (as per LCD display).

Good man!

18 hours ago, Gert-Claassen said:

What would you recommend?

The LFP version of that firmware, i.e. LF1-74.40e. It will be the other downloadable file at the end of that same post you would have gotten LC1-74.40e. The sibling of the one you have, if you like.

Hi,

 

OK, I have loaded the LF1 version - works good, but either I am not looking at the correct things - but I am not really noting any difference?

 

Gert

@Coulomb I have been running LC73.00e since it became available on my 2 x Axpert MKS PF 0.8 cluster with Trojan T105 batteries. The Axperts are controlled by ICC-Pi on a RPi with BMV700 connected. Voltage switching levels are set to minimums on the Axperts so as not to interfere with ICC. As voltages are set on ICC, I guess there is not any benefit in reflashing the inverters with LF73.00e if I replace the Trojans with banks of 16 series connected LiFePO4 cells. I ask to make doubly sure that the LiFePO4 cells are not damaged and would appreciate your advice. 

Are you using any controller other than the Axperts on your own system, which as I recall from past discussions on the forum, Is the same as mine except for ICC-Pi and batteries. 

20 hours ago, Gert-Claassen said:

works good, but either I am not looking at the correct things - but I am not really noting any difference?

The differences are a bit subtle, such as the ease with which you can switch from float back to bulk stages, interference between setting 29 and setting 12, how long the battery stays in absorb stage (how low the current falls before switching to float); those sorts of things.

17 hours ago, ebrsa said:

Voltage switching levels are set to minimums on the Axperts so as not to interfere with ICC. As voltages are set on ICC, I guess there is not any benefit in reflashing the inverters with LF73.00e if I replace the Trojans with banks of 16 series connected LiFePO4 cells.

No, as per the above, there are other subtle changes that make it worth getting the correct flavour of patched firmware (LF1 versus LC1).

17 hours ago, ebrsa said:

Are you using any controller other than the Axperts on your own system

Yes, I have a BeagleBone Black (similar to a Raspberry Pi, but not identical) with custom Node Red monitoring software written by a colleague.

So I get graphs like the below (sorry, not that exciting since it's night time here):

 

That's only half the graphs, and there are another 4 or so pages of dials, settings, a debug screen, a 5 dial overview, etc. All data is logged, so I can go back to data from any earlier day as a .csv file, though it's a slight pain to dig it out.

The custom monitoring software talks to the custom BMS that we designed as well. We had grand ideas of selling it all, until we realised the hassle of supporting it all. All except the Node red stuff is open source and open hardware, for the extremely brave.

@Coulomb Thank you for your advice and interesting description and graphs of your control system. I am, at my advanced age, not brave enough to tackle a DIY controller so will continue using the RPi and ICC-Pi control software that has served me well for many years. Clearly I will have to bite the bullet and re-flash my inverters with your and @weber's LF1-73.00e firmware.

I have made a note of your voltage settings, which your have advised on this forum some time ago, in my Aussieview manual. How have they worked for you with your LFP battery bank. My own DIY battery of 16 serial connected parallel pairs of  120AH cells with one Smart Ant BMS will hopefully be up and running in about 10 days. So any modifications to your original voltages will be much appreciated.

1 hour ago, ebrsa said:

I have made a note of your voltage settings, which your have advised on this forum some time ago, in my Aussieview manual. How have they worked for you with your LFP battery bank. My own DIY battery of 16 serial connected parallel pairs of  120AH cells with one Smart Ant BMS will hopefully be up and running in about 10 days. So any modifications to your original voltages will be much appreciated.

Weber and I are still using 55.2 V and 53.7 V. My cells were made in 2009, so now 12 years old, and I find that occasionally I have to change these to about 54.8 and 54.7 for a day or so, to allow them to balance better. If I don't do this I get "the clunks", where the inverter overshoots the voltage target, and by the time that the current is reduced enough, the stress level of the cells has reached 12 and the charging sources (AC-in and PV-in) are disconnected. It stays that way until the stress reaches 7 (neutral), then they turn back on (one by one so that the power supply isn't overwhelmed by the surge current of several contactors coming on at once). I could probably modify the Node Red software to do this automatically, but it doesn't happen often enough to make it worth the trouble.

As far as those voltages are concerned (55.2 and 53.7 for 16S LFP), I'm still 100% happy with them.

Thanks very much for your advice @Coulomb. I will try your voltages and see how it pans out with the Ant BMS. Just one question, does your controller communicate with your inverters and does the controller rather than the inverter determine when switching from grid to solar or vice versa occurs.

@Gert-Claassen I hope that my posts on your thread will assist in optimising you own similar solar system, which is probably a major objective of all of us. 

2 hours ago, ebrsa said:

does your controller communicate with your inverters

Yes. It reads values like PV power, etc.

Quote

and does the controller rather than the inverter determine when switching from grid to solar or vice versa occurs.

No, we don't attempt that. I occasionally have to tweak the settings on rainy days. All settings can be changed from a Node Red page.

We do disable utility charging during the day, however. This is mainly for a colleague with a time-of-use tariff, and I may go to time-of-use myself one day.

Edited August 22, 20214 yr by Coulomb

@Coulomb thank you for the information about your controller. My ICC-Pi does control the inverters' switching based on state of charge as per the BMV700 as well as time isettings. But it would seem to me that flashing the inverters with the LF1-73.00e firmware, as you recommended, would add voltage to the control. From everything that I have read about LFP batteries, that seems to be the crucial factor. It will be interesting to see how I manage to optimise the switching as soon as the LFP batteries are connected.