resistors with long legs

[AndrewT]

My statement is NOT an opinion.
It is not my opinion.
It is a comment stating a FACT.
"The conduction to the PCB MAY be better for cooling than air circulation alone."

.........................
Generally what your saying is simply wrong....................

My statement is not an opinion. It is a fact.
It is not wrong.
Read what it says and then explain what is wrong about it.

 

[Fast Eddie D]

sometimes big, high temp rated power resistors (multi watt metal oxide or wirewound) that get really hot are purposely mounted with long leads to avoid scorching the pcb

but for smaller R the heat conduction through short copper leads to the pcb trace can be part of the power rating

I've been working on some older vintage equipment. I have seen lower power resistors (quarter or eighth watt) mounted a half inch or an inch from the board, with insulating sleeves on the leads. I assumed that this was for power dissipation, because I can think of no other reason to do so (and one bad reason as mentioned - parasitic pickup). I figured that the manufacturer saved a few pennies by not upgrading to a half watt or one watt resistor.

Slightly off topic, heating of a resistor in a precision application can have a significant effect on circuit operation. In the case of a feedback resistor, the larger resistor will dissipate more power and thus drift more. While this is not generally an issue, it can affect performance in parallel circuits and bridge amps where precision is much more critical. Conversely, if a resistor is in a current limiting application - like an emitter resistor - the drift can actually have a slight beneficial effect on the circuit.

When repairing equipment, it is advisable to keep the physical configuration as original as possible, unless you are intentionally altering a parameter or "upgrading" a component or configuration.

 

[Fast Eddie D]

My statement is not an opinion. It is a fact.
It is not wrong.
Read what it says and then explain what is wrong about it.

Andrew, the leads of a component and how they are configured physically on the board are part of the thermal gradient profile. In other words, the leads conduct a significant proportion of the heat away from the device. (This is especially crucial with op amps.) For smaller components like small resistors and op amps, this represents a significant and important proportion of that profile. For larger resistors, it is less significant and the thermal gradient can be improved by mounting them farther from the board where more air can circulate around them.

 

[wg_ski]

High(ish) power resistors mounted up a bit from the board do not actually run cooler. Not any hotter, either. But it prevents the PCB from getting cooked - case temps can easily get to 100C or more, even at half rated dissipation. When the board cooks it can turn to carbon, which is partially conducting. And the solder joints oxidize and go bad giving a further source of heat. Until something gives and you've got an amp to fix.

Not just resistors, but 1 and 2 watt zener diodes, too.

 

[audiostrat]

AndrewT, yes I used that word. But as I said again, I used it instead of saying "post". I am not a native speaker, so please bear with me. I meant post. My mistake if you understood something else! 🙂

Ending the off-topic, I think that wg_ski's post contains significant information. And as I asked earlier, will this placement also aid in not transferring heat from the resistors to other components?

Say you have the reservoir capacitor and right next to it an RC filter, with R being a high power resistor. Assume that this resistor operates at a certain wattage during operation, and gets hot by a certain amount. What is better for the reservoir, in terms of letting him run cooler? Positioning the resistor close to the pcb, so that the whole board gets hot and conducts heat to the cap, or allowing a bit of space to ensure only air heat transfer? Also bearing in mind that hot air goes upwards, so it is less likely to affect the cap?

Or it makes no significant difference? 🙂

 

[DF96]

There are no simple answers to these questions. It all depends on PCB design (maximum copper, minimum copper?), lead thickness, air flow etc. If you are worried about heating other components then spacing and low temperature of the hottest item are the key. A 2W resistor dissipating 1W will probably get hotter than a 5W resistor dissipating 1W; the heat energy is concentrated into a small volume.

If you really need to know the answer then a full multiphysics simulation is needed. This may be expensive!

 

[audiostrat]

I imagined that there can't be a simple question, just seeking for rules of thumb maybe.

I plan to place 5W resistors close to 105C capacitors, but these shall not exceed 0.5W continuous dissipation. I left one of them running at 0.5W for a quarter, and I was able to touch it - notably hot, but I could keep my finger on it for ages. So they should be around 50C, and since ambient is around 21-22C, I think they should not get more than 30C above ambient. So I don't think they would actually get hotter than 70C for my needs.

So, it seems like I have no trouble. But hey, 5C can double the capacitor's life - so why not even slightly lower temperatures? I was always fond of practices that can give you a good result by means of just moving something or what. 🙂

 

[sreten]

My statement is not an opinion. It is a fact. It is not wrong.
Read what it says and then explain what is wrong about it.

Hi, It is fact but the wrong physics, rgds, sreten.

 

[AndrewT]

Sreten,
there is nothing wrong with the physics in my comment.
DF agrees. Many factors should be taken into account to model each situation.

I am NOT saying contact WILL be cooler, That would be a conclusion requiring evidence.
I am saying MAY and experiment or modeling would be required to show which in any particular situation would lead to cooler running.

This also applies to Audiostrat's question
"what effect on capacitor temperature does a nearby heat source have?" Thise needs to be modeled or measured.