1
\$\begingroup\$

I build a circuit that charges a 400V 4.7µF capacitor to 307V. I have been discharging this capacitor by shorting the terminals using a large piece of metal. The capacitor makes a small spark and then my multimeter reads 0-5V from the capacitor.

I have heard in various places that this is bad for the capacitor because it causes a large current spike that can permanently reduce the capacitance and might blow up the capacitor. People also recommend using a resistor to discharge the capacitor.

I found a listing for a resistor on Amazon that has a resistance of 1K ohms and can dissipate 100W of heat.

I don't know if it is necessary to use a resistor like this or if a regular 1/4W resistor would suffice.

I calculated that at 1000ohms and 300V there should be a current of 0.3A and a wattage of 90W. This made me think that I need this big resistor.

\$\endgroup\$
  • \$\begingroup\$ Note that the rated power is in terms of continuous dissipation. You might find the following question and answers helpful: Resistor power rating for short pulses \$\endgroup\$ – Alfred Centauri Sep 4 '17 at 22:11
  • \$\begingroup\$ If you are patient, 1M works fine in 10 seconds to safe levels. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Sep 4 '17 at 22:30
  • \$\begingroup\$ Two 1/4W series parts will have a breakdown rating of 500V or more. So if you plan to do this often, 2 x 100k is a good choice. Cap current depends on rated V/ESR, MTBF depends how often you do this, material type and OEM quality \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Sep 4 '17 at 22:42
1
\$\begingroup\$

That resistor would be massive overkill. Yes, the instantaneous current and power dissipation is that high, but it only lasts for a very short time. After one R-C time constant (about 5 ms), the voltage and current have already decayed to 37% of their initial values, and the power is only 14% of its initial value.

The total energy stored in the capacitor is:

$$\frac{1}{2}C\cdot V^2 = 0.22 \text{ Joules}$$

Even if you discharged it once a second continuously, that would be less than 1/4 W average dissipation in the resistor.

However, you do need to pay attention to the voltage rating of the resistor as well as its power rating. Very small resistors cannot safely handle hundreds of volts.

\$\endgroup\$
1
\$\begingroup\$

You do need a power resistor, but anything above about a 1-2W rating would be sufficient for this application.

Look at the curve provided here: enter image description here

It's important that you use a resistor designed to support considerable power dissipation, and a wire wound is best. For example a wire wound resistor is ideal when compared to the smaller metal oxide resistors which may be laser trimmed and have a small dissipation area. Power resistors can usually absorb enormous short term power abuse, typically 100:1 or more of their average/continuous power rating.

You can get some understanding of the capability of power resistors by looking at datasheets such as this Vishay wirewound resistor. You'll notice here that the 'fusing limit' of the resistor is at up to 120 W power rating, and still takes many hundreds of milliseconds to fuse.

enter image description here

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.