Posted by: @robl

↑

@ggw I had a lot of trouble with the sunsynk originally(ok, I still do…).  New LiFePo4 battery, with the Seplos bms that talked with it, and yet I found the 0-100% doc just wasn’t right - it would charge to 100% and charge for loads more still!  Then run to 0%, and still have more to give:  I ended up running the system on AGM voltage mode, 55.2V top and 42V bottom across the 16 cells.  After running like this for a few months the sunsynk bms re-evaluated the cell Ah, and now it works well in Li mode.

I recommend running in AGM voltage mode for a while and letting it find the battery size.  Ours evaluated the batt as 425Ah - it was sold as 304Ah.  I assume the sunsynk has an inaccurate current shunt on the battery current measurement, hence the difficulty with this.

Good luck.

Ok, I have ths mode turned on. How did you know what voltages to set on the inverter?

is the issue being caused by the battery not being fully balanced and its taking time to identify the paramaters and its the battery/bms as opposed to the inverter?

i top balanced all my cells before connecting by battery (albeit i had a different inverter at the time) but when i changed to a Sunsynk, i didnt get any of the issues you seem to be having?

Hey 

I ran active balancer for the first 3 cycles. Voltage on the screen looks in good shape per cell.

Figuring out the Sunsynk however...

Try this agmV mode and see how I get on 

@ziaulh it would seem that's it's going be now against code to have battery in the loft.

SITREP #2 

Progress, but not logical progress on the Sunsynk.

So I changed the Sunsynk over to AGM-V

Good - I was able to charge to 100% and discharge to 1%. Good job that way.

However, the control of this setup is sub-optimal. The SOC that the inverter thinks it is at is wildly off. Eg. Showing 83% on sunsynk when the battery is 100%.

My timer is setup with voltage rather than SOC. It's a rough get around. And the same situation is still occuring that the Inverter voltage reading is 0.2-0.3V higher than the battery OS. I choose to believe Fogstar BMS it is more accurate.

Changed it back to Lithium protocol this morning and it was trash readings. Inverter doing funky stuff. After 30mins I gave up and turned on AGM-V again. The battery capacity never changed, different to what @robl stated on his. 

At a bit of a loss now in how to proceed.

@ggw

I left ours in AGM volts mode for maybe 2 months, the adaption was automatic but slow.  I did try and hurry it along by manually tweaking the Ah value it used a few times too!  These are the values for the batt I used when in AGM V mode (16S 304Ah lifepo4)

Agree that the SOC values given by sunsynk are terrible when using AGM v mode; it only gets full and empty correct when used like this.

Posted by: @robl

↑

Agree that the SOC values given by sunsynk are terrible when using AGM v mode; it only gets full and empty correct when used like this.

Indeed, for LiFePO4 batteries voltage is a poor proxy for SoC except at 0% (2.5V per cell) and 100% (3.65V per cell). This is why, after briefly thinking about the use of AGM mode (or was it the coublomb counting version?), I went back to BMS communication.

Posted by: @ggw

↑

Changed it back to Lithium protocol this morning and it was trash readings. Inverter doing funky stuff.

It seems as if communication between the Fogstar battery and inverter is not happening properly. Have you tried selecting the BMS_Err_Stop option and seeing whether there are intermittent errors? I can't see from the Fogstar pages whether there is a specific BMS protocol that should be used for Sunsynk/Deye, although I would hope the default Pylontech setting should just work. Would it be worth double-checking the values the BMS is sending using the PBMS tool on  https://www.fogstar.co.uk/pages/specification-sheets/ ? Finally I guess a call to Fogstar would be worthwhile, in case they have received feedback from other users!

@misterb @robl 

SITREP #4?

Blimey, what haven't I tried

Voltage mode, wasn't working very well. It's a bad SOC replacement. I tried every single setting change on BMS/Lithium mode - I couldn't make the battery charge to 100%.

So I gave up, contacted Sunsynk. They factory reset my unit. I quickly installed the basic setting in grid, battery & user timer.

That night - it charged too 100%! Magical. Brilliant.

5 days later, it's creeping down. Only got to 95% last night.

Same behaviour. Have BMS_Err_Stop running, no errors.

The Sunsynk is learning to/capping the charge, why?

@ggw 

Have you contacted Fogstar?  That would be what I would be doing - it seems to be battery related not Sunsynk as we have 2 different Sunsynk inverters with Seplos boxes and don't experience that behaviour.  If Ben @ Fogstar has a Sunsynk too then that'll narrow down where the problem is.

i think Andy at off grid garage has had issues with the Seplos, perhaps they are similar issues - might be worth watching some of his you tube videos (if you havent already done so) 

i think his solution was related to setting the cells to 'over charge' so it reset the BMS, but dont take my word for it, check it yourself !!!!!

@misterb @vaugi 

OK, I'll contacted Fogstar

Were I fall down that it is the battery - is that when the Sunsynk is factory reset - it charged the battery to 100% (twice)

Does this not mean it can charge it to 100% but after a week+ - it is deciding not too? 93% last night.

Im not factory resetting the battery at any stage

So much for fit & forget...

Hi there to all those on this forum thread. I'm currently thinking about getting a Fogstar 15.5kWh battery solution and doing a DIY install.  I've already got a GivEnergy 8.2 Battery with an AC connected AC3 GivEnergy inverter which works fine for my needs for the summer months when my ASHP is not needed for heating.  But now my fixed tariff from EDF has finished (10 hrs a day and all weekend cheap rate) I've moved over to Octopus Intelligent Go for my EV charging (via Zappi 2). So now I only have 6 hrs overnight cheap rate to run my house and my calcs show I will need to either double my current GivEnergy system capacity to 2x 8.2kWh batteries or go for another separate battery storage system.  An additional GivEnergy battery installation (add on) will be in the region of £3500, but I'm very interested in what it would take to go the DIY route (I'm experienced with building LiPo packs etc) and go for something like the Fogstar unit with a Victron 8000 inverter. This would be a grid tied AC connection like my current system and mostly just charging from grid overnight on cheap rate with no export back to grid planned.  

Is this a viable route? is the Victron inverter solution recommended?