|
|
re: battery curves
2 may 2000
dave woolcock wrote:
>>...but the point was to determine
>>the state of the battery, ie to measure its approximate series
>>and parallel resistance, not to predict the state of its charge.
>er no. my original query was to try and determine the state of charge
>of the battery
my point was how to automatically find and bypass a bad (open or leaky)
battery in a series/parallel network.
the single-slope e-meter algorithm seems to answer your query, although
trojan specifies two discharge slopes for t-105s: i = 4436.6t^-0.8596
for a discharge current i > 85 a and a discharge time t in minutes, and
i = 1168.2t^-0.5726 for i <= 85a. the second slope doesn't match their
measured low-current data well... i = 4782.9t^-0.8759 fits much better.
nick
|
|
