So, I found this … https://raw.githubusercontent.com/FurTrader/OverkillSolarBMS/master/Overkill_Solar_BMS_Instruction_Manual.pdf
Page 51, Appendix B: Calibration
Yes this is right, calibration just means making something read to a known trusted quantity.
In your case you want the BMS to read the same current reading as the BMV.
There would seem to be no reason to do a voltage calibration as those readings seem to marry up already.
Prior to doing this procedure:
I would do a Zero current calibration on the BMV using Victron connect app first.
After that I would use the BMV reading as the the reading to calibrate to.
I would also calibrate at as high current as would typically be your maximum charge/discharge.
Another trick is you could swap the + and - on the BMS so that you would be able to calibrate both discharge and charge from only a single primary current condition. Otherwise you could create genuine discharge and charge conditions. It’s up to you whichever is more convenient.
1 Like
Right, so with a car headlamp bulb, blew the previous version … don’t ask … connected to cells 3 and 4, at slow charge of 15amps, more and the cells jump out, less and the bulb takes too much, it seems to behave.
Green thin wire is there to turn the 2nd filament on.
Want to see if this works …
So far so good …
Will take it to 95% … or until I go to sleep, cause if the charge stops, the bulb must be removed.
Before I calibrate anything, I had a brain fart today.
Started here:
Already here:
Using “The Bulb v3” …
Discovered a feature see, where you can switch off, that it only balances on charge. So now it balances with no charge, or rather, maube a teeny weeny amount of amps.
Got to this point by telling the Venus the Min SOC is 100%.
That feature is nice to initially balance your battery. Charge the battery to where the first cell starts getting full, then stop the charge and switch on the option to always charge and leave for a day or 2. Charge again and repeat untill all the cells are nice and neat.
That takes longer than you “The Bulb” method, but the BMS does all the work for you and you don’t have to watch it all the time.
Now if that second BMS I ordered would just arrive I can do this on the cells sitting idle in the garage… 
1 Like
“The Bulb v3” works for the impatient one … 0.028v difference over the bank …
V4 … 2 x headlamp bulbs.
IF IF IF this works …
Who’s the boss? 
If I tried this with my old Daly BMS … DC Ripple after DC Ripple … aka DC Ripple’S.
Bringing them down in batches of 4 cells now … cell 3 and 4 my problem children.
Got a EV Peak C1 50w 6a DC Balance charger for Lipo/Liion/Life/Nicd/Nimh to sort them … and the other small cells I have.
Right, done for now … 0.017v difference … will leave it overnight, maybe for 24h … will then see what happens when it goes back into production.
If the cells behave … then this is next … the calibration:
Seconds later …
Sometimes +2 amps …
And If that works, the calibration between BMS and BMV … I will be The Master … again … of “WTF Not To Do”!
1 Like
Right, so with “The Bulb v3”, and the EV Peak set to 3.6v max and 6amps, I’m pulling those two cells up, one bit at a time.
Drain to like 20% SOC, do some charging there, without panels assisting. Normally evenings.
80% SOC, do some charging there, like the pic, solar assisting.
100% SOC, “The Bulb v3” comes out for the other cells, and some more charging by the EV Peak for cells 3 and 4.
Yes, it takes time, and one cannot leave it all on and walk away, but man(!) if this works … I’ll smile.
Sorting cells out on a live system, with no uncoupling then taking days/weeks … naaa, THAT is just looking for trouble … if you go by the moniker TTT. 
Once I’ve got this sorted, I have lots of holders for AAA and AA and 9v batteries, to then charge them.
That is AFTER I sorted them 2nd hand 2nd life bulging cell. It has “compressed” nicely over the months … will give some TLC there like “plug and forget” TLC, as the EV has a few options to preprogram all the above settings.
The 2nd hand 2nd Life cells, have 14 of them, 8 is perfect, sorting the last 6 now … using the 12v MP 500va and “The Bulb v3” x 2.
Once the system has run for a week or so, charging to 100%, discharging to 20%, each day, and the cells stay in line, as the other 8 did, I’ll smile broadly.
And If I can “fix” the bulging ones … last … that would be kinda grand methinks.
Must say, I’m enjoying this way too much. 
A 12mV delta is something I would have ignored…
EDIT: Wrong bank …
That 12mV is because of the EV charging cell 3 with like ±4 amps, cell 4 now “topped” up at 80% SOC.
The rest of the cells I sorted Saturday, because as cell 3 and 4 jumped, the SOC got confused and the charging algorithm was altered, so they never got properly charged.
Bank used to be on average 11mV out over 16 cells.
That 4 cell 12v bank … nice … 0.004 - 0.005v out now, after 14 issues yesterday.
At 94%, it is 0.006 - 0.008v difference.
The 16 cell bank … have not been here for a while … YES!
Next is the daily charge/discharge cycles, that they STAY in line. Normally I need 2-4 cycles with TLC during the times, for both the 16 and 4 cell banks, before all is “settled”, as cell 3 and 4 dropped faster than the rest, and shot up faster.
Then I can focus on the calibration of the BMV and BMS, once the cells behave in unison again:
Update: After many MANY hours, I think I have a “formula” to handle an unbalanced bank better, and reducing the risk of a cell or more “shooting out” when over 90%.
First a titbit or two to know about:
If you set a Victron inverters Grid SetPoint to Zero, it will drain the batteries every so slowly.
If you set the Grid SetPoint high like on a 500va, it will charge the batteries at about 78w.
What I deduced is that the best “protection” lies with @Louisvdw driver.
When the batts hit 91% SOC, the charge current is halved.
When the batteries hit 95%, charging is stopped … keeping in mind the two titbits above.
The above is aimed at giving the BMS more time to balanced at the top.
And also, when I know the house will idle when we are away for a weekend, I can drop the charge SOC to stop at like 85% … as a further precaution.
Ps. I would not be surprised if the brand names do similar, as it helps a damn lot keeping cells in line, by using the equipment vs “The Bulb”.
Yes. If there is an imbalance (one cell shoots out over 3.55V while others are still below 3.4V), then you need to put the BMS in the best spot to do its job.
With a passive balancer, this means a small amount of current must continue to flow, otherwise it has nothing to work with. A passive balancer works by allowing that small current to flow through the lower cells, while bypassing the higher cell.
You want to hold the voltage at a point wher the highest cell is somewhere between 3.55V and 3.75V (no higher), while allowing just enough current to pass to pull the other ones up.
Ideally you want that amount of current to be equal to the maximum the BMS can bypass, which is usually in the order of a few tens of milliamps. Since the Multi doesn’t allow you to do adjust it that accurately (anything below 2A is basically a guess and a prayer), you want to spend as much time as possible to get from 90% to 100%.
Basically, a bit like a good lover. There, I’ve put that image into your collective heads now.
Now I can never look at my Multiplus the same again. 
Expect an answer that will put politicians to shame…
Solar working at max array potential, pushing like +3.6kw to recharge main system and 12v bank, with both banks at 96% SOC, @Louisvdw BMS/Venus driver handling it:
12v 300ah bank sitting solidly at 0.004 - 0.008v Delta over 4 parallel/2 series x 150ah cells.
48v 100ah bank, 100ah and 150ah cells mixed (… they said …), solidly at a delta of between 0.006-0.008v.
… they said … Lifepo4 can mix and match, so I wanted to see how it “feels”. 
But bugger, you better make sure the cells are balanced.
Grid Setpoint at Zero.
# Change depending on the SOC values
if self.soc > 95:
self.control_allow_charge = False
else:
self.control_allow_charge = True
# Change depending on the SOC values
if 95 < self.soc <= 100:
self.control_charge_current = 0
elif 91 < self.soc <= 95:
self.control_charge_current = self.max_battery_current/2
else:
self.control_charge_current = self.max_battery_current