Charge Conservation Anomaly (7th Mar '08)

my first tests were set up to switch charge from an input capacitor to an output capacitor

the input voltage supply was a capacitor formed from 3 x 1F caps in series (= 0.3333F);  the output load was a single 1F capacitor

the switching was achieved using transistors connected as Darlington pairs driven by an oscillator constructed using CMOS 4093 logic (Quad Schmitt NAND gate); the switching circuitry was powered from the input capacitor - no external supplies or Sig Gens were used

i charged the input capacitor stack to 6V and operated the switching circuit until the voltage on the output stack  increased from 0 to  2V

the 'charge sheet' was as follows:
input stack started at 6V  (2 Coulombs)
& finished at 2.7V  (0.9 C)

output cap started at 0V (0 C)
& finished at 2V (2 C)

Total starting charge = 2 C
Total end charge = 2.9 C

charge would be expected to be conserved in such operations - however, the results show that although energy was expended, net charge increased

see the Wikipedia Electric Charge entry -  http://en.wikipedia.org/wiki/Electric_charge  - for a commentary on Conservation of Charge

"This law is inherent to all processes known to physics" - apparently not !!

i extended my circuit & included the primary of a transformer in series with the switching capacitor and connected the rectified output of the secondary to a capacitor load;

circuit extended, adding inductance &  'feed-forward' transformer action

i repeated the charge experiment, but this time the input discharge was arbitrarily limited to a 1.5V drop from 8V
(the 2nd switch was disconnected so that the switching cap just accumulated charge and didn't transfer its charge to the load)

the input cap, switching cap and load cap were each formed with  4 x 1F in series

'scope trace of capacitor voltages 

the trace above shows the voltage on the input capacitor stack discharging and then i sampled the final voltage values on the switched capacitor stack & the additional load capacitor stack fed from the transformer secondary output

the input was discharged from 8V to 6.5V; the switched cap charged from 0 to 2.5V; the additional load cap charged from 0 to 0.6V

at start:
8V  2Coulombs (input cap)

at end:
6.5V 1.625C (input cap)
2.5V 0.625C (switched cap)
0.6V 0.15C (load cap)

Total starting charge: 2C
Total end charge 2.4C

(i repeated the test with the capacitor stacks swapped  - the results showed the same effect)

this anomaly was determined not to go away

so - if charge really is misbehaving what effect will that have on energy?   let's move on to the energy experiment...