I long felt it was easily possible to have a microcontroller notify you at the end of a desulfation cycle and be able to predict how long it would take.
While testing a new pulser prototype, I discovered that desulfation occurs at a random rate. This sort of knocks the wind out of being able to predict when the desulfation cycle will end.
Two batteries were selected for testing. Battery A had a substantially
higher standing voltage in the 11 volt region. Battery B had an initial standing voltage of only 10.47v.
While desulfating a pair of batteries in a controlled environment, one thing became very clear. The rate of desulfation
occurs at a random rate.
I was hoping the rate would be linear. If so, a battery finishing time might be predicted based on early voltage change data.
Battery A:
13.45v 1:30pm
13.88v 2:30pm +0.43v
14.06v 3:30pm +0.18v
Battery B:
12.84v 1:00am Start
12.86v 4:00am +3 hours from start
12.88v 5:00am +4 hours from start
I had earlier thought that battery A might have some irregular increases possibly due to some interference from gas bubbles clinging to the plates or electrolyte remixing resulting in SG changes due to stirring. This seems to be eliminated in B battery results in the pre-gaseous state.
Another odd phenomenon was observed. Batteries that are severely sulfated will result in a fast rising voltage that eventually lowers and bottoms out before seeing a steady voltage climb.
Battery B:
10.47v Standing
13.35v 1:00pm Start
14.56v 2:00pm +1 hour
12.60v 2:00am +13 hours from start
Conclusions:
Desulfation cycle times cannot be predicted from early or late stage voltage changes
Time to end of cycle cannot be predicted
