Hi,
I have a 7S52P Powerbank (364 lithium cells 18650) and All the readings for each bank (7) goes into two ADS1115 and goes into Arduino Nano. All readings are "good" and tested/compared to multitester voltage reading for calibration.
BUT as the temperature of the day changes (calibration was at 9:00 in the night) somehow the readings change in a few millivolts also.
is this a know issued?
Thanks for the info!
So , with this, am I right to asume that calibration should it be at "working" temperature ? (during the day its get charged and every mV counts) and asume that during the night the calibration could go a little odd??
PD: I charge each bank of 52 x 18650 batterys till 4,1V and measure mV to calibrate all.
Not sure what your calibration is used for. I haven't worried about temperature effects (maybe I should). I have found the load changes and the variation in sunlight for charging to be bigger problems.
yes, I understand that...
but in this case for the load topic and a bigger battery reduces battery sag.
In the same way sunlight (and watts) to charge a bigger battery also reduces sag
so controlling the maximum voltage of the charging part (calibrated for the temperature of the day) would be enough in my case.
PS: as I read again your post I remember also that the size or diameter of the cable, and the amps it could freely transport, also affects the voltage and current flow.
Think we need to know a bit more if what you are doing to get able to comment sensibly
There are specific methods of charging Li cells and you should look that up
well after reading and nerding out I guess I found the problem in the readings
each bank goes into a Voltage Divider (two resistors) and ADS1115 does not shift with temperature, but RESISTORS DO!!!
So after this noob finding, I will have to change the carbon resistors into metal film resistors so minimize the shift in resistance and have a smaller gap of the readings shift
What are the values of those dividers.
Don't forget that an ADS1115 has a much lower input impedance than an Arduino A/D.
The PGA input has a typical impedance (to 0.7volt) of 6Megohm.
Leo..
510K + 100K (for 29,4V)
510K + 120K (for 25,2V)
510K + 150K (for 21,0V)
510K + 201K (150K+51k) (for 16,8V)
510K + 303K (270K+33K) (for 12,6V)
510K + 610K (510K+100K) (for 8.4V)
10K + 100K (for 4.2V)
by lower impedance you mean I should use less than 10K resistor instead of the bigger values I used?
The impedance and maybe other negative effects of the A/D become more apparent with higher resistor values. Up to your experiements how much that is and how much you can accept.
I see you calculated the dividers for ~4.8volt. Not what I would have done.
Calculating for ≤4.096volt, and using a PGA of 1 gives a higher resolution.
Leo..
now I have 50mV shift .. was expecting less than 10mV but maybe it can't be done.
as for resolution I guess I had enough to make it work
Would you advise to use 1/10 of the values in my voltage dividers to get better readings?
That would reduce the influence of the A/D, but add more drain to the battery.
Leo..
did the obvious test:
blow some air into the resistors and boy it shifts fast
thanks for the responses, will change the hole resistor array!
