SoC accuracy and consequences?

45 views
Skip to first unread message

Will OBrien

unread,
May 4, 2021, 8:04:15 PMMay 4
to electrodacus
I'm suspicious of the shunt calibration on my setup.
My main shunt is a 75mv 400A, and my PV shunt is a 75mv 250A.

I have entered the following for each:
.075/250 = .0003Ohm OR .3 mOhm
And
.075/400 =  .0001875 OR .1875 mOhm 

That said, I'm not seeing data quite align with what's measured by my Victron BMV712.
It's close, but I'm seeing about 1A difference. This could be down to the difference in time to report.

This got me thinking a bit, Dacian, are the significant consequences to having SoC on the SBMS not be quite accurate? It's most likely the cheap shunts that I have more than anything else. 

Solar is a bit to jagged right now for good comparisons, so I'll try to check it one day when I have a constant current source for comparison.

Will OBrien

unread,
May 4, 2021, 8:09:49 PMMay 4
to electrodacus
Of course right after I said that, the clouds cleared, now I'm only seeing about .3A difference in readings which is pretty decent.
Still it got me curious.

Dacian Todea

unread,
May 5, 2021, 12:16:23 PM (13 days ago) May 5
to electrodacus
Will,

It depends on the accuracy of your current shunts. Some lower cost current shunts may have tolerances in resistance of up to around 5%  and so if you use calculate resistance values you are trusting the manufacturers spec. Even if the current shunts are the cheaper less accurate ones (as they are not properly factory trimmed)  you can adjust the resistance value live to make the current match a known accurate measurement device that way basically calibrating to correct resistance value.
For example if you disconnect all your loads the PV shunt and battery shunt should show the same value within around 50mA as there is a SBMS self consumption that you need to consider.
Reply all
Reply to author
Forward
0 new messages