about AAA or LR03

for obvious reasons i didn't want to run the tests myself so i relied on THIS - likely the only thing the Google can currently allocate - suitable for this purpose

a bit processing ::



and we have some fuzzy insights based on my own LR03 stats. - based on 60 AAA-s @ various discharge using ε(r) relation ::

Load (Ω)	dt (s)	Eε(U) (V)	Er (Ω)	I (mA)	ΔC (mA·h)
3.9	23.4	1.49(1.11)	1.31	285	735
5.1	32.5	1.47(1.14)	1.44	224	876
20	132	1.32(1.14)	3.22	56.9	1058
24	194	1.37(1.25)	2.34	52.2	1025
75	619	1.29(1.23)	4.07	16.3	1180

note: the prefix E is used for Expected value in probability theory - such as EX = Σ(x_i·p_i) = [here assuming all p_i = 1/n] = 1/n·Σ(x_i) , /!\ /!\ /!\

Load (Ω)	Fresh			Exhausted
	ε(U) (V)	r (Ω)	I (mA)	ε(U) (V)	r (Ω)	I (mA)
3.9	1.69*(1.49)	0.533	381	1.38(.902)	2.07	231
5.1	1.67*(1.5)	0.573	294	1.34(.907)	2.46	178
20	1.57(1.5)	0.88	75.1	1.17(.901)	6.09	45
24	1.62(1.57)	0.71	65.5	1.21(.998)	5.07	41.6
75	1.6(1.57)	0.771	21	1.05(.902)	12.7	12

reprocessed web LR03 stats. ::

* --

/!\

The source site refers their device falling into initial EMF range of 1.58...1.65V - since the corresponding values for 3.9 and 5.1 Ω tests exceed the range it must be assumed that the test resistance used was of the inappropriate power range e.g. it's resistance significantly increased during the beginning of the test due thermal effects . Since there would be two unknowns to find and match to process time at all times i won't provide the appropriate corrections here. I suggest the graphs being more adequate at the end of tests. update : as i always post-error-check my ramblings (((a programmers special disease :: the prog-s're huge - the 'puter responce is always at some extent unpredictable - so it makes more sense to write your code - then multiple pass dis-errorize it - you can't avoid unknown error conditions in advance - even if there's no programmatic (code-flow) errors - you should still verify it does what it suppose to e.c e.c))) - so the update - it seems the QC passed a quite common 1 "a sign error" - it appares the higher the TEST resistance the higher the EMF(electro-motoric force for the battery (we need it actually to be lower here)) - /!\ there is actually such intermediate term phenomena as the more current is drawn from battery the better it performs (the higher the EMF temporarily climbs) - which might explain the higer startup EMF for lower resistance TESTs - i must make tests of my own to verify such -- for NOW what is known is that if we /!\ /!\ ALTER the "battery formula" ε = 1.5371 r-0.1493 >> ε = 1.25 r-0.5 then it'll show better correlation for that very dependency throughout the different TESTs . . . /!\ yes ? - should it ...

/!\

NB! -- this is just apx. guide line -- there is no any real instance of LR03 presented on those graphs
PS! -- notice that the unit for the internal resistance here is not the conventional and is varying by graph.