Do computer PSU need a load?

[alm]

I think a temperature of 300 °C is dangerous. I think it's dangerous because you might be a dopey electronics tech like me and put the resistor dummy load on the end of the ATX power cable and lay it a bunch of papers or a wood desk.  When you start to see the paper smoke and the wood starts to smoke and burn, you'll understand why these power resistors can be unwittingly dangerous.  We're all use to using 1/4 watt resistors and they don't usually get hot plugged into a perfboard, but power resistors are a breed unto themselves.

So put them in a box with plenty of ventilation to avoid touching them. 1/4 W resistors can also get quite toasty near their rated power by the way, it's just that we tend to run them at << 1/4 W. Even though they may get 100 °C instead of 200 °C for a power resistor, 100 °C is still plenty to burn your skin.

I'm still waiting for an EE to calculate the thermal properties of a power resistor and show me how to calculate the heat generated of various wattage resistors run at the same current and voltage.  What specs do I need to use and what's the math?

As IanB I will assume you mean temperature and not heat, since heat equals power in this case.

The data sheet will sometimes contain a graph of temperature rise vs dissipated power. Add the temperature rise to the ambient temperature and you've got an estimate for the temperature it will reach. This is usually defined for the resistor sitting in free air, so it will be higher for an enclosed space. The maximum power is usually derated with ambient temperature, that 5 W resistor won't be able to handle 5 W at 150 °C ambient. If the data sheet does not contain a load vs temperature curve, check other data sheets of resistors with similar materials, construction and dimensions.

Another way would be to estimate the thermal resistance from the temperature derating curve. Temperature derating often just shows the maximum power that will keep the temperature below the max. operating temperature, so if a 20 W resistor can handle 100% of the power at 70 °C ambient, linearly decreasing to 0% at 270 °C, you might estimate the thermal resistance to be 200 K / 20 W = 10 K/W. This assumes that the thermal resistance is constant, which may not be accurate, so it may be too conservative (high).

[danz409]

my luck sucks... was starting day 2 of my power supply build. bought some new test leads that have a clip on attachement from radioshack.. (yea i know.. horrable place..)
but honestly the price wasn't too bad for once $15 AND they are fairly decent quality

however i went to go do some testing with my $4 Cen-Tech and apparently now there is a constant short between the leads somehow. i opened it up and i see nothing on the board that would cause the issue...  that's 2 of these chepos down and 1 left.. problem is... i spent the last 2 hours looking for it and i can't find it.

thankfully i got another coming in tomorrow with a heap of components from Jameco. which i have work so.
 yay for not soo productive weekend

[onewatt]

As IanB I will assume you mean temperature and not heat, since heat equals power in this case.

I remember from college physics, for example, that a swimming pool with 100 °C water temperature has a lot more heat than a cup of water at 100 °C  - even though both have the same temperature.

Thanks for explaining that Watts relates to heat.  I think I have a better understanding of what's going on.

[IanB]

I remember from college physics, for example, that a swimming pool with 100 °C water temperature has a lot more heat than a cup of water at 100 °C  - even though both have the same temperature.

Thanks for explaining that Watts relates to heat.  I think I have a better understanding of what's goin on.

There's a difference between stored heat and power. Power is the flow of energy, and heat flow is the flow of thermal energy. Thermodynamics says that all forms of energy are equivalent and energy is conserved when it changes from one form to another.

So in the case of a resistor, the electrical power consumed by the resistor (volts x amps) is equal to the thermal power (heat flow) being generated by the resistor. The heat being generated inside the resistor will cause it to heat up until the heat leaving the resistor balances the rate of heat generation. If you wrapped the resistor inside a thick layer of insulation so the heat could not escape, the resistor would get very hot indeed and might even melt.

[alm]

There's no difference between a heat sink and a power resistor as far as the thermal behavior is concerned. The only difference is that the thermal resistance of a power resistor is often not specified, you have to estimate it from other data like temperature rise or dimensions. Dave did some videos about thermal calculations for heat sinks, and I'm sure there are plenty of other resources.

[IanB]

The important thing to know with any design is that the "thermal resistance" between the device (heat sink or whatever) and the surroundings is not a fact. It is an estimate based on certain assumptions or presumptions about the environment and air flow around the device and how it is installed. For example there may be thermal conduction through the leads or soldered joints of the device and for SMT devices this could be critical to good performance.

Huge differences in heat transfer can exist when considering enclosed and open spaces, free convection or forced convection, nearby hot components, and so on. Just because the manufacturer quotes a number or provides graphs in a data sheet, it doesn't mean you can rely on those numbers without doing your own due diligence.

[alm]

Absolutely, but this due diligence likely consists of doing actual measurements, especially with the limited thermal specifications that you usually get with power resistors.

[vk6zgo]

This seems to be a practice confined to computer power supplies.

Most switchmode supplies used in other service don't care if they have a load or not!
Early Philips TV supplies needed a load,or they would go into the dreaded "hiccup" mode.
Philips,being a power unto themselves,never stopped gazing at their own navel long enough to realise nobody else had this problem,& Techs could reduce the load with abandon,in the search for a fault in other brands.

Computer PSU's may have more legitimate reasons for this practice--for instance,more critical voltage regulation,but they are definitely out of step with the rest of the Electronics industry.

VK6ZGO

[alm]

Most current Mean Well SMPSs with multiple outputs specify a minimum current for each output. Some switchers in Tek scopes also needed a minimum load to start reliably. I don't think it's limited to computer power supplies.

[amyk]

In many cases the minimum load is already provided by resistors internal to the PSU, next to the output.

(They're also known to cook the capacitors next to them.)

[danz409]

mine may need a load D:

http://youtu.be/yujYeMnXlNQ

[danz409]

i did some experiments WITH a load. and it still does the same thing... it may be some sort of safety feature