# 100W resistors. Are they really 100W? [Solved]

#### ballsystemlord

Joined Nov 19, 2018
102
Hello,

(When talking about power, I will be referring to continuous load. Pulsed is another discussion, and isn't relevant to my problem.)

When I learned electronics it was that an XXX watt resistor could be loaded up to XXX watts. So, when I purchased some aluminum encased 100W resistors, I thought I could load them at 100W.
Today I placed two of them in parallel under a 200W load (100W per resistor -- 10v @ 10a). I was trying to measure the temperature another part of the circuit and so I had a laser thermometer in hand. I smelt a funny scent, so I checked their temperature. They heated up, and not just by a little. They hit 274C! Now each of them had a dedicated 12v 120mm fan pointing at them (rated at ~2300rpm, 44CFM). The room temperature couldn't have been higher than 30C (the thermostat on the AC said 27.8C). Obviously, I disconnected the power when I saw how hot they were.

Here's the resistor type I have: https://www.aliexpress.com/item/1005003040560587.html

Are 100W resistors really 100W?
Is there some kind of derating curve that my book on electronics didn't go into?
How hot should these aluminum encased resistors get?

Thanks!

#### ElectricSpidey

Joined Dec 2, 2017
2,181
The type of resistor you are referring to normally get mounted onto a heatsink.

#### Tonyr1084

Joined Sep 24, 2015
7,199
I placed two of them in parallel under a 200W load (100W per resistor -- 10v @ 10a).
So assuming you pushed 100 watts through those 100 watt resistors, at 10 volts that would be 10 amps. You said you put two of them in parallel. Well, at 10 volts (divided by) 10 amps is going to be 1.0Ω. Think carefully about what happens to resistance when you put two resistors in parallel. IF you have two resistors of 1.0Ω each, the total resistance would be 0.5Ω. So 10V divided by 0.5Ω equals 20 amps. Something is going to get hot at that kind of amperage. Each resistor should handle half the load, assuming they're the same ohmage. The resistors you linked to are 10Ω, not 1.0Ω. Your numbers suggest the resistors are 1.0Ω.

At this point I don't know for sure what you were smelling. Could have been wiring getting hot, could have been just the resistors alone. But like ElectricSpidey said, they're supposed to be mounted to some sort of heat sink. If you didn't then they had to dissipate all that energy by themselves. A heat sink of sufficient size would certainly have made things a whole lot cooler.

Are your resistors 1.0Ω? Or are they 10Ω like the link you posted said?

#### dl324

Joined Mar 30, 2015
15,446
Is there some kind of derating curve that my book on electronics didn't go into?
Power dissipation rating is always derated for temperature. Conservative designers won't use any component at 100% of it's rated value.

This is for Ohmite high power resistors:

#### ballsystemlord

Joined Nov 19, 2018
102
So assuming you pushed 100 watts through those 100 watt resistors, at 10 volts that would be 10 amps. You said you put two of them in parallel. Well, at 10 volts (divided by) 10 amps is going to be 1.0Ω.
<snip>

Are your resistors 1.0Ω? Or are they 10Ω like the link you posted said?
Sorry to confuse you. They are both 1.0Ω. They each get 10v @ 10a for 200W total, 100w per resistor.
The link I posted sells a selection of values. 1R is among them. I just double checked.

Last edited:

#### ballsystemlord

Joined Nov 19, 2018
102
The type of resistor you are referring to normally get mounted onto a heatsink.
I've never seen a resistor heatsink before. Could you point me to an example? Or maybe there's an official spec?
Thanks!

#### nsaspook

Joined Aug 27, 2009
10,679

#### MrChips

Joined Oct 2, 2009
27,679
100W is 100W. The electrical power is being converted to thermal energy regardless of the power rating of the resistor. 100W has to go somewhere otherwise the resistor will get hot!

I bet that you can't hold on to a 25W soldering iron with bare hands.

#### crutschow

Joined Mar 14, 2008
31,116
It should have been somewhat apparent that a device that small couldn't dissipate 100W to air without getting very, very hot.
If you look at the specs for that type of resistor, the 100W rating is when it is mounted on a heat-sink that is maintained at a 25°C temperature.
If the heat sink goes above that temperature (which it will do unless it is water cooled), the power must be derated linearly to its maximum operating temperature.
The heatsink can be any type, such as the finned aluminum ones used for power transistors.
It will need to be pretty good size to dissipate near 100W, or else have a fan blowing on it.
Those with a built in fan designed to cool high power PC CPUs may work.

Below is the specification for a similar resistor.
Note that the 100W unit is rated for 30W with no heatsink (the surface temperature will be about 200°C at 30W).

The data sheet is here.

So the lesson is that you don't use an electronic device, even a power resistor, without reading the data sheet.

Last edited:

#### MrSalts

Joined Apr 2, 2020
2,611
Sometimes, it is just easier to use a 12V, 100W incandescent projector bulb, 55W automotive headlight, or similar. Note, they get hot, too. If you're pumping 100W into something, and it is not getting hot, you're doing something wrong.

I was working with a short-throw projector in a cabinet for a lighting test. Someone decided to center the vent hole on the cabinet door for aesthetics even though the vent hole was designed to perfectly align with the projector's fans. They burned two \$150 bulbs in about 5 minutes before they wondered what was wrong.

Lesson: Even if something is designed to get hot, it is still possible to get too hot. Like a 100W resistor without a heat sink.

#### dl324

Joined Mar 30, 2015
15,446
Also note that not all resistors are rated for operation at 200C.

This graph is easier to read:

Full power rating at room temperature with proper heatsinking.

For example, MP820 is rated for 20W maximum, but at 25C ambient with no heatsink, it can only handle 2.25W. The '*' denotes the note with the derating curve.