A HID Xenon light on my bike

After I had finished my charger, the shunt resistor became very hot. The shunt resistor senses the curretn through the battery. The charger tries to keep the mean voltage over the shuntresistor at 160 mV. In my case I wanted to load with 3.2 A, so the resistor had to bee (0.160/3.2=) 0.05 Ohms. I expected that a resistor of 3.2 A x 0.160 V = 0.5 Watt would be sufficient. This appeared not to be true. The charger controls the current through the battery by chopping the sine current with thyristors. Through a trigger pulse on the gate of the thyristor a part of the rectified sine wave is used (see thick black line in figure 1). The gray line is the recified sinewave, the thick black line is the voltage over the shuntresistor.

Figuur 1

In the charger I build, the triggerpulse is at 8.83 msec after the zerocossing of the sine ("trigger time"). The mean voltage of the black line was indeed 0.160 Volt. The Root Mean Square (RMS) value then is 0.542 Volt. This high RMS value is caused by the high sharp shape of the black line. The dissipated power in the shunt resistor yields now to 0.542x0.542/0.05 = 5.9 Watt (Oeps!). Thats why I replaced the shuntresistor by a 30 cm , 2 mm diameter stainless steel wire of the same resistance at the original resistor (0.05 Ohms). The wire has a lot of convective cooling and does not get realy hot.

The ration between RMS voltage on the shuntresistor and the mean voltage depends on the time of triggering. The earlier the trigger is, the better the ratio RMS to mean voltage is (figure 2 and 3).

Figure 2

In my case the triggering occured after 8.8 msec. So the ration RMS to mean is about 3.3, so dissipation is about 3.3 x 3.3 = 10 times higher then I initially expected. I have a 25 V transformer causing the the bad RMS/Mean ratio and the the high power dissipation in the shunt resistor.