4
\$\begingroup\$

I don't have an explosion chamber, so I cannot test critical values.

  1. If you explode resistors with too high current-voltage combination, does it damage your equipment, such as multimeter and DC-supplier, even in the explosion chamber?
  2. How do I notice when a critical point for resistors is achieved, not touching the hot component?
  3. I want to test a Zener diode for increasing voltage. I have a great amount of different diodes, but I have no idea which ones are Zener diodes. Which resistors and Zener diodes should I use and how do I find the right ones without exploding them?
\$\endgroup\$
14
\$\begingroup\$

In my experience, resistors and diodes burn, but they don't explode.

The only components that I've experienced exploding are tantalum capacitors when placed with the wrong polarity and transient voltage suppressors ("tranzorbs") when exposed to ~2x their rated voltage.

To determine which resistors or diodes to use, you can check their power rating. For example a typical through-hole resistor is rated for 0.25 W. Depending on how you're using the resistor, there are three equations that are useful to calculate the power dissipation:

  1. Watts = volts * amps
  2. Watts = volts * volts / ohms
  3. Watts = amps * amps * ohms

For example, suppose you have a 10k resistor that you're using as a pull-up resistor for a pin on a 5 V microprocessor. The maximum power drawn by the chip through the resistor (by equation #2) will be 5 * 5 / 10000 = 0.0025 W, or 2.5 mW. That's fine even the tiniest surface mount resistors.

\$\endgroup\$
  • 1
    \$\begingroup\$ The power dissipation is 25 W. If that's below the manufacturer's rating, everything is OK. If you don't know the manufacturer's rating, and you can't estimate it by the package dimensions, you should monitor the current and voltage for a while to see if the resistance is staying constant-- if it changes, either it will start to burn, or your device will stop working. If the resistance is constant, the resistor is doing its job. \$\endgroup\$ – pingswept Jun 5 '10 at 18:56
  • 4
    \$\begingroup\$ Electrolytic capacitors explode, too. I have a small scar to prove it. :) \$\endgroup\$ – endolith Jun 5 '10 at 22:21
  • 6
    \$\begingroup\$ When I mentioned estimating by package dimensions, I meant that even if a resistor comes from an unknown source, its ability to dissipate heat is going to be roughly proportional to surface area, so you can make a good guess. For example, if it's a small cylinder, 2 mm diameter and 5 mm long, it can probably handle around 0.25 W. If it's the size of a hot dog, it's up in the 100 W area. \$\endgroup\$ – pingswept Jun 5 '10 at 22:35
  • 1
    \$\begingroup\$ I've had a TDA2030 explode (explode seems a bit generous for a little snapping sound) and threw shrapnel at my shirt. I've also seen some modems with split chips after a thunderstorm. I presume other IC's with hard shells can do the same if overloaded. \$\endgroup\$ – XTL Jul 16 '10 at 15:49
  • 1
    \$\begingroup\$ I've had an NPN explode (sending shrapnel flying - not good!!), as well as a small IC. Oh, and the occasional electrolytic. \$\endgroup\$ – Thomas O Oct 1 '10 at 22:27
5
\$\begingroup\$

Manufacturers typically give a maximum voltage for a particular resistor range, such as 200V for a 250 mW resistor. That needs to be taken into account, as well as the power.

\$\endgroup\$
4
\$\begingroup\$

The maximum voltage has more to do with the package dimensions than blow-up issues - if the maximum voltage across the device is exceeded it may arc across the terminals effectively bypassing the device.

Resistors can fail both open and short depending on the nature of the stress exposed to it.

A typical multimeter in voltage mode is tens of megohms. Blowing a resistor while measuring its voltage will likely only damage the probes. If you're measuring current, it will depend on the meter. Is it fused? If so, the fuse should protect the meter.

The "critical point" for a resistor is solely based on its temperature. If it gets too hot, the internal structure breaks down and the part dies.

If you have some sort of current-limited DC source, set it to 5mA or so and use it directly to test for the zener effect in your unlabelled diodes. Otherwise, use the highest resistor you can find to keep the current to a minimum (recognize that the zener voltage varies with load current) and test away, reducing gradually once you get an idea of the approximate voltage. Keep the total power in the zener below a few hundred milliwatts for safety's sake (if you don't know what power they can handle).

\$\endgroup\$
  • 1
    \$\begingroup\$ I think that using "the highest resistor you can find" won't work in some cases, particularly when that would be a 10MOhm resistor. I'd suggest using 20K ones, as that's only 1mA of current (and 20mW of power dissipated) even at 20V, which is probably the maximum voltage you would like to have zener diodes for. \$\endgroup\$ – whitequark Oct 2 '10 at 16:24
  • \$\begingroup\$ Point taken. The highest resistor you can find "within reason" would be a better answer. \$\endgroup\$ – Adam Lawrence Oct 3 '10 at 18:48
2
\$\begingroup\$

I've seen a 1/4-watt 1K resistor explode when one end got connected inadvertently to line voltage while the other end was connected to a ground-referenced circuit. Interestingly, all the current that went through the resistor also went through an ADC input on a Microchip PIC (16C73 if I recall) but the processor chip suffered no apparent damage.

\$\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.