Body
I am comparing the energy into the battery with the energy out. The battery is a 13 Ah. Energy out is done with a 33 ohm resistor for 24h. Assuming the battery is fixed at 12 V that makes
Energy dumped to the resistor when discharging:
E = Vbat^2 * h / R
In my case:
E = 12^2 * 24 / 33 ≈ 104 Wh
I have chosen Vbat = 12 V so that energy dumped is an underestimation since it started at 14 V.
Energy in is calculated using the average current into the battery:
I = C * f * (Vcap - Vbat)
Energy to the battery when charging is then:
E = C * f * (Vcap - Vbat) * Vbat * h
In my case if charging for 5 days:
E = 12E-6 * 50 * (110 - 15.4) * 15.4 * 5 * 24 ≈ 104 Wh
I have chosen Vbat = 15.4 V so that charging energy is an overestimation since it never reach that.
This gives COP ≈ 1
So if charging takes less than 5 days we can suspect OU at the battery-terminals.
I have tried all kinds of caps, but caps are not made the same, different kinds of caps with the same capacitance can give very different results. Some get hot others don't, some charge the battery really well on low voltage others don't even though the have the same capacitance. I tried a really high-capacitance cap at low voltage but the diodes fried!
A bigger gate-resistor requires a higher AC-voltage, otherwise it will not trigger, but once it triggers it just consumes energy so its god efficiency to have a high resistance. Note the possibility to switch the position between the SCR and the battery.
The current into the battery depends on capacitance*AC-voltage. But the energy in the cap depends on capacitance*AC-voltage^2 which makes me wonder whats going on. This together with the differences between cap-types makes it difficult to test. I need more caps. Motor-caps seems like the way to go.
/Hob