Heatsinking through hole resistors?

[kalel]

With SMD resistors, I assume this is different, but how would you heatsink through hole resistors?

I'm not sure how commonly this is done, and also don't have a specific project in mind, but it might make sense in some projects. If it doesn't make any sense, that would be good to know too.

I got the question from thinking about those battery tester units supplied with resistors for load. But those are different shape and material (ceramic) than e.g. metal film transistors and I'm curious about those too. Surely they don't need to be heatsinked if used properly, but I'm not sure if that might help to keep things slightly more stable. Of course, a simple fan might be helpful as well.

[tautech]

I've seen a few solutions, mostly in old gear where power consumption was not such an issue as it is today.

Resistors with stout enough leads are stood off the PCB with no other support provided.
Stacking ceramic beads on the leads to provide a rigid standoff.
Metallic brackets, mostly used with 5W+ ceramic resistors to stand them in on end.
Use a larger W ratings than necessary.

[T3sl4co1l]

Only metal case resistors need heatsinking.  Hmm, there's also SOT-227 resistors available, and such.  Anyway, anything that's not meant to be bolted, bonded or pressed to a heatsink, mmmkay?

As mentioned above, the usual approach is to insulate, so the PCB doesn't get stupidly hot.  The resistor is made of materials that can operate over 200C, while the PCB turns brown above 150C (and solder melts above 200C).

Providing a large copper dissipation area, for the leads, will help reduce hot spots on the PCB, and is often a necessary step to use high power SMT resistors.

Tim

[SeanB]

Regular metal film and carbon film resistors are meant to operate at a relatively high surface temperature, so the best way to run them at full rating is to mount them at least the body diameter off the board on the leads. There were lead formers that did this with a single tool, but a good compromise is some small needle nose pliers to put a kink in the leads at the right point so they are spaced there for soldering. As Tim said, a large copper pour on the solder leads will keep those cool, and help prevent solder joint failure, but also have a large solder mask relief as well, or leave the whole pad free and solder fill the pad entirely.

Power resistors ( typically a white ceramic boat with the element inside and filled with a cement, or a green cement coat on the resistor wound on a ceramic or glass fibre former with pressed on end caps) are best mounted vertically, with use of a special tinned steel mounting that provides one lead connection, along with holding the resistor from moving with vibration, that spaces it typically 1cm from the board. Otherwise mount them 3cm from the board, using either small ceramic beads or a ceramic or metal tube, though there are solder in rivets that combine the tube with support as well. Power resistors typically are rated for full dissipation at a surface temperature of something between 175C to 200C, and really should be derated to reduce the high surface temperature so you do not scorch the board, or run them mounted to tag strips so that the heat is off the board, and use wires to connect them.

Worst thing you can do is have power resistors mounted right next to electrolytic capacitors, where you cook the capacitors to failure in short order. Even worse id to have them touching, or use silastic or glue to hold them to the board or other components, which will cook the board brown and make tracks peel off when you try to rework there

[TimFox]

Heathkit used to specify mounting power resistors (typically rectangular "cement" cases) by cutting a bit of cardboard from the shipping box, spacing the resistor from the board during soldering, and then removing the cardboard.

[cvanc]

Heathkit used to specify mounting power resistors (typically rectangular "cement" cases) by cutting a bit of cardboard from the shipping box, spacing the resistor from the board during soldering, and then removing the cardboard.

You know what this means?  It means you're as old as I am because I remember this too   

[T3sl4co1l]

Also if you're feeling fancy, remember you can still buy those ceramic spacers to slide over the leads, e.g. http://www.essentracomponents.com.sg/images-line/pdfs/CER-X.pdf

Gives more support, and sets the component height on assembly without needing kinked leads.  (Which you can also get, if you find the right part numbers -- most manufacturers provide lead form options, but distributors mostly stock the straight axial format, so if you want just a few, instead of a box of reels, you have to snoop around carefully for the right part numbers.)

Tim

[0xPIT]

You typically don't need to heatsink resistors, if they operate within their design specs.

A common way to manage heat at resistors is to let the resistor stand off the PCB as long as possible.
Sometimes ceramic standoffs are used.

The wires act as a heatsink, this usually seems to be sufficient, at least to prevent self-desoldering.

[alm]

The wires do not so much act as a heat sink, but rather as a thermal resistance to drop the temperature at the solder joint so the PCB substrate does not get overheated. I doubt they contribute much to the resistor case-ambient thermal resistance. If anything, shorter leads would allow more heat transfer through the PCB, decreasing the case-ambient resistance. Obviously this would heat up the PCB, which is usually the limiting factor. The resistor may easily handle 200°C continuously, unlike most PCBs. By putting thermal resistance (thin wires) between the resistor and the PCB, you protect the most sensitive part (the PCB).

[retrolefty]

You typically don't need to heatsink resistors, if they operate within their design specs.

 This is the 'official' policy one should take.

However I know that many a thrifty Ham Radio types have frequently taken an inexpensive route in the construction of 'full-gallon' (1K watts) RF dummy loads out of a bunch of 2-5 watt that is series/parallel wired for 50 ohm load and hanging the resistor array into a gallon paint can and immersing them into transformer oil == 1,000 watt 50 ohm resistor. Works fine for the HF bands but not very good at UHF and above.



 

[SeanB]

Oil cooled can dummy loads are also typically only used for a very short time, so the oil and the resistors do not get close to the boiling point of the oil. This allows a short high power input to be handled without making a lot of smoke then fire, as the thermal capacity of the oil absorbs the power. They typically have a duty cycle of under 3 minutes per hour, as the oil takes a long time to cool.

[colorado.rob]

https://www.vishay.com/resistors-fixed/heatsink/

[Beamin]

I have wondered this myself. I would take a little roll of copper wrap it around it and leave a tab hanging off one end. Fill the middle with heat sink goop. Simple and effective. Solder one side of the copper to one of the leads.

[stj]

these days you can get metal-film resistors in a TO220 or TO218 package!

[Beamin]

these days you can get metal-film resistors in a TO220 or TO218 package!

That would be a strange product to tear down. All transistors by the look of it. I might order some of those I have a thing about keeping resistors cool.

[Simon]

erm, resistors are in themselves rated to a certain power, if you need more power you specify a resistor for that power, once you get into serious power (over 5W) you have to heat-sink the resistor anyway and it will be designed to accept a heat-sink. You can get resistors rated to hundreds of watts that can't do that power unless put on a heat-sink.

[Cliff Matthews]

I thought these 225-watt ceramic resistors didn't need a heat sink.. oh well, MTDOC has 20 of them now 
https://www.eevblog.com/forum/buysellwanted/fs-ohmite-new-old-stock-225w-2ohm-resistors/msg1082793/#msg1082793

[CatalinaWOW]

You can cheat a little bit by making little copper or aluminum tabs a folding them over the leads of the resistor and crimping them in place (or solder if copper is used).  I have seen a similar tab shaped around the body of a resistor to buy a little more power dissipation capacity.

This might be a chosen approach for a hobby project where you were unwilling to wait for an appropriate part to be delivered.  Although I have actually seen it in commercial products, the only way I would personally approve of this would be if the resistor were rated for the dissipated power and the heat sink was intended to lower operating temperature for improved reliability.  It is hard to imagine where this would be more cost effective or space effective than just specifying a higher power rated resistor.

[thermistor-guy]

I've seen a few solutions, mostly in old gear where power consumption was not such an issue as it is today.

Resistors with stout enough leads are stood off the PCB with no other support provided.
Stacking ceramic beads on the leads to provide a rigid standoff...

Old (failed) steam irons are a good source of ceramic spacers. Steam irons also have useful items like terminal blocks and high-temperature wiring.

Next time your steam iron fails, scavenge it for parts.

[Simon]

I thought these 225-watt ceramic resistors didn't need a heat sink.. oh well, MTDOC has 20 of them now 
https://www.eevblog.com/forum/buysellwanted/fs-ohmite-new-old-stock-225w-2ohm-resistors/msg1082793/#msg1082793

There will be some I grant you that do high power all by themselves, but I bet they are bigger than their little aluminium counterparts that are deceptively small because they are meant to be put on a heatsink or can only do less power.

[T3sl4co1l]

Yes, ceramic resistors are usually somewhat larger than aluminum body resistors.

Only somewhat so, because they are able to operate at MUCH higher temperatures.  Air is a proportionally more effective heatsink (power per volume) at high temperatures.  In fact even more than proportional, because of natural convection.

Aluminum body resistors tend to be slightly cheaper, or at least competitive.  But one must keep in mind it is a false economy: the heatsinking is NOT OPTIONAL.  You must include the cost of extra heatsinking volume, machining steps, hardware and assembly.  They're a big waste, overall!

Tim

[tszaboo]

For those 5-10W cemented wirewound resistors, you can cool them with just some airflow.

There are chassis mount resistors, these go up for some 50-100W

This, you drill holes to the heatsink, and bolt it on it.

And then there are the big beasts, which go up to 1KW

Again, no heatsinking.

There are TO220 and other transistor case style resistors. IMHO they are not very good price/performance wise. For a single one, you can  just use a dozen cemented resistors. They are fine, if you are using them as a precision current sense.

[Simon]

The aluminium ones are probably used to get heat out of the case by screwing them to the inside of the case wall so that heat can go through the wall and out possibly with an additional heat-sink, a ceramic resistor would heat up the inside of the enclosure unless it is forced air cooled or the resistor is put outside.

[Beamin]

Heres a variable high wattage resistor with built in heat sink. Caution it gets really hot.

[kalel]

Heres a variable high wattage resistor with built in heat sink. Caution it gets really hot.

Needs good ventilation.