Post by thread:[EE] Overdriving power resistors

I don't know the answer to your question, but I note that with the latest extruder head even the Makerbot people have abandoned using resistors as heaters. they've gone to the little canister elements that everyone else is using. Somewhere before I blew up my computer I had a link to a great website that showed pictures of the resistor cores migrating out of the aluminum heat sinks and destroying themselves. I personally had problems because every time I fired the thing up the solder would soften up and the resistors would disconnect themselves. You're probably not running them that hot. but generally I would (and did) use nichrome wire for heating up large flat things, even though it's a royal pain to wire. Thanks, DougM On Sat, Oct 1, 2011 at 6:44 PM, PICdude <spam_OUTpicdude3TakeThisOuTnarwani.org> wrote: {Quote hidden}> I have a few chassis-mount resistors mounted to an aluminum block > specifically for the purpose of heating the aluminum block (3D printer > extruder head). Can't I run much more power (than rated) through the > resistors, as long as I don't exceed the max operating temp? > > Specifically, the resistors are 5-ohm 10W, and if I power it from 12V > (lead-acid battery) I get 2.4A per resistor, and 28.8W. I would PWM > it to reduce the average power, but with feedback. So it would run at > 28.8W until it comes up to temp (I need <150 deg-C), then reduce > average power. The resistors are rated at 250 deg-C max, so I'd be > well within that. > > Cheers, > -Neil. > > >

On Saturday, October 01, 2011 6:44 PM, "PICdude" wrote: > I have a few chassis-mount resistors mounted to an aluminum block > specifically for the purpose of heating the aluminum block (3D printer > extruder head). Can't I run much more power (than rated) through the > resistors, as long as I don't exceed the max operating temp? > > Specifically, the resistors are 5-ohm 10W, and if I power it from 12V > (lead-acid battery) I get 2.4A per resistor, and 28.8W. I would PWM > it to reduce the average power, but with feedback. So it would run at > 28.8W until it comes up to temp (I need <150 deg-C), then reduce > average power. The resistors are rated at 250 deg-C max, so I'd be > well within that. Hi Neil, I have some experience using cheap ceramic resistors as heaters in a commercial product. To sum up my experience, you will experience a higher-than-desirable failure rate if you significantly overdrive them, even if the temperature is low, the duration is short, and feedback limits the temperature. I was using 5 watt resistors. 6.4 watts under worst-case seemed to be the sweet spot for me. More than that and I could expect some failures. So 28.8 watts from a 10 watt resistor is about 15 watts more than I would allow in a volume product. You could probably hand-select resistors or use "qualilty" brands and do OK. But don't expect cheap ones to all survive. Cheerful regards, Bob -- http://www.fastmail.fm - Access all of your messages and folders wherever you are

part 1 3022 bytes content-type:text/plain; charset="iso-8859-1" (decoded quoted-printable) Background: I set my high-school engineering mentees off to design an extruder head, by researching the DIY (makerbot, reprap, etc) units out there that work, and the resulting design looks like this ... http://wiki.makerbot.com/plastruder-mk5 . The nichrome-wire wrapped was discussed, but finding high-temp cement was difficult, so they went with the resistor block type. Attaching a pic of ours, which I hope will make it through to the list. Adding kapton tape helped a lot (not shown in that pic). My engineering mentees are off for a bit, so I'm on it for now. I've seen those resistor-type units work well, so I should be able to tweak it into submission. Ours melts the ABS plastic, though just not enough to flow freely. Currently, the resistors are two-in-series, paralleled with the other two in series, so 7.2W of heat is generated by each resistor (each resistor is rated at 10W. They're get no where near hot enough to melt solder currently. IIRC ABS melts in the low 200's (deg-F) range. Looking for tweaks to get some more oomph out of it, but starting with the electrical things, before going to mechanical things (reducing thermal mass, etc). Cheers, -Neil. Quoting doug metzler <.....doug.metzlerKILLspam@spam@gmail.com>: {Quote hidden}> I don't know the answer to your question, but I note that with the > latest extruder head even the Makerbot people have abandoned using > resistors as heaters. they've gone to the little canister elements > that everyone else is using. > > Somewhere before I blew up my computer I had a link to a great website > that showed pictures of the resistor cores migrating out of the > aluminum heat sinks and destroying themselves. > > I personally had problems because every time I fired the thing up the > solder would soften up and the resistors would disconnect themselves. > You're probably not running them that hot. > > but generally I would (and did) use nichrome wire for heating up large > flat things, even though it's a royal pain to wire. > > Thanks, > > DougM > > On Sat, Oct 1, 2011 at 6:44 PM, PICdude <picdude3KILLspamnarwani.org> wrote: >> I have a few chassis-mount resistors mounted to an aluminum block >> specifically for the purpose of heating the aluminum block (3D printer >> extruder head). �Can't I run much more power (than rated) through the >> resistors, as long as I don't exceed the max operating temp? >> >> Specifically, the resistors are 5-ohm 10W, and if I power it from 12V >> (lead-acid battery) I get 2.4A per resistor, and 28.8W. �I would PWM >> it to reduce the average power, but with feedback. �So it would run at >> 28.8W until it comes up to temp (I need <150 deg-C), then reduce >> average power. �The resistors are rated at 250 deg-C max, so I'd be >> well within that. >> >> Cheers, >> -Neil. >> >> >> --

How do these things fail? This is not a volume or commercial product, but a one-off 3D printer head. Resistors are these ones from Yageo... http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=810F5R0E-ND So I have some options now... (a) Use an 18V source (perhaps add a 6V battery in series with the 12V battery, so each resistor would generate 10.8W of heat), (b) Build a PWM switcher now and put all 4 resistors in parallel, then adjust for 10W to each resistor, (c) Swap the resistors to 15-ohm units (same physical size and 10W), which in parallel will be 9.6W per resistor, (d) Run 28.8W through each resistor and see if I can get it to work for some time before it fails. (a) is complicated cause I'll have to buy a 6V battery to experiment. (b) will take time to get components for a PWM switcher, but probably makes most sense. (c) is pointless, cause I might as well try (d) first and if they fail I'd then do (c). Think I'm leaning to see if I can dig up a power mosfet to build a PWM-switcher temporarily using a function generator. Cheers, -Neil. Quoting Bob Blick <.....bobblickKILLspam.....ftml.net>: {Quote hidden}> Hi Neil, > > I have some experience using cheap ceramic resistors as heaters in a > commercial product. To sum up my experience, you will experience a > higher-than-desirable failure rate if you significantly overdrive them, > even if the temperature is low, the duration is short, and feedback > limits the temperature. I was using 5 watt resistors. 6.4 watts under > worst-case seemed to be the sweet spot for me. More than that and I > could expect some failures. > > So 28.8 watts from a 10 watt resistor is about 15 watts more than I > would allow in a volume product. > > You could probably hand-select resistors or use "qualilty" brands and do > OK. But don't expect cheap ones to all survive. > > Cheerful regards, > > Bob > > -- > http://www.fastmail.fm - Access all of your messages and folders > wherever you are > >

> but generally I would (and did) use nichrome wire for heating up large > flat things, even though it's a royal pain to wire. I'll second the use of Nichrome for high power levels. As I noted yesterday, I had occasion to dissipate up to 500 Watts (not for long) or 100 - 200 Watts for longish periods, and suitably thick Nichrome is very abuse resistant. It is run at red hot temperatures by design in eg toaster and radiant heater elements and seems to take such treatment well. I ran air cooled coils at below visbly hot (just you-really-don't-want-to-touch-this hot) and reliability was never a problem. As Doug says - connection is "annoying". There are a number of webpages that provide soldering advice. I used mechanical clamps (connector block principle but heavier duty screws. ) Russell McMaho

Hi Neil, They fail open. The voids in the molding process and looseness at the weld point between nichrome and lead-wire are the fail points. Remember I am using ceramic-sand-and-concrete resistors. But in my past I have seen resistors like you are using fail in similar ways. If this is not a commercial product, I'd say "just go for it". If you get a dud once in a while, you'll fix it. My application is on a diesel truck with serious vibration and operation down to -40 degrees, so the thermal shock is considerable. Cheerful regards, Bob On Sunday, October 02, 2011 6:30 AM, "PICdude" wrote: {Quote hidden}> How do these things fail? This is not a volume or commercial product, > but a one-off 3D printer head. Resistors are these ones from Yageo... > search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=810F5R0E-ND > > So I have some options now... > (a) Use an 18V source (perhaps add a 6V battery in series with the 12V > battery, so each resistor would generate 10.8W of heat), > (b) Build a PWM switcher now and put all 4 resistors in parallel, then > adjust for 10W to each resistor, > (c) Swap the resistors to 15-ohm units (same physical size and 10W), > which in parallel will be 9.6W per resistor, > (d) Run 28.8W through each resistor and see if I can get it to work > for some time before it fails. > > (a) is complicated cause I'll have to buy a 6V battery to experiment. > (b) will take time to get components for a PWM switcher, but probably > makes most sense. (c) is pointless, cause I might as well try (d) > first and if they fail I'd then do (c). > > Think I'm leaning to see if I can dig up a power mosfet to build a > PWM-switcher temporarily using a function generator. > > Cheers, > -Neil. > > > > Quoting Bob Blick <EraseMEbobblickspam_OUTTakeThisOuTftml.net>: > > Hi Neil, > > > > I have some experience using cheap ceramic resistors as heaters in a > > commercial product. To sum up my experience, you will experience a > > higher-than-desirable failure rate if you significantly overdrive them, > > even if the temperature is low, the duration is short, and feedback > > limits the temperature. I was using 5 watt resistors. 6.4 watts under > > worst-case seemed to be the sweet spot for me. More than that and I > > could expect some failures. > > > > So 28.8 watts from a 10 watt resistor is about 15 watts more than I > > would allow in a volume product. > > > > You could probably hand-select resistors or use "qualilty" brands and do > > OK. But don't expect cheap ones to all survive. > > > > Cheerful regards, > > > > Bob > > > > -- > > http://www.fastmail.fm - Access all of your messages and folders > > wherever you are > > > > --

On 3 October 2011 02:30, PICdude <picdude3spam_OUTnarwani.org> wrote: > How do these things fail? This is not a volume or commercial product, I used commercial white ceramic 5W and 10W resistors to "power" a Fluidyne Stirling Engine (small and for fun). These were run at about rate wattage with the aim of being "as hot as possible". From memory these were placed in contact with a metal tube, possibly layered over with high temperature silicone rubber and wrapped in fibreglass insulation (intended for in-wall insulation). These lasted "for a while". Given that the engine only got occasional use I'd suspect they lasted 10's of hours of run time. This was OK for the application. Temperature unknown. Failure O/C. If you used Nichrome and brought the ends out to a cooler location you'd not have the problems that Bob mentions. Russell

part 1 1572 bytes content-type:text/plain; charset="iso-8859-1" (decoded quoted-printable) We've done similar things with 20W and 100W resistors. The most consistent failure was the soldered joint from resistor tabs to wires. At any elevated temperature above 80degC we noticed that the joints, over time, would fail and become open cct. This was cured by either bolting or crimping wires to the resistor tabs. We've run 20W resistors (ceramic casing) up to 100degC for up to 10+ hrs before element failure. 100W up to 160degC for close to 40+ hrs before element failure (ceramic/aluminium casing). Nino. On 3/10/2011 2:14 PM, RussellMc wrote: {Quote hidden}> On 3 October 2011 02:30, PICdude<@spam@picdude3KILLspamnarwani.org> wrote: >> How do these things fail? This is not a volume or commercial product, > I used commercial white ceramic 5W and 10W resistors to "power" a > Fluidyne Stirling Engine (small and for fun). These were run at about > rate wattage with the aim of being "as hot as possible". From memory > these were placed in contact with a metal tube, possibly layered over > with high temperature silicone rubber and wrapped in fibreglass > insulation (intended for in-wall insulation). These lasted "for a > while". Given that the engine only got occasional use I'd suspect they > lasted 10's of hours of run time. This was OK for the application. > Temperature unknown. Failure O/C. > > If you used Nichrome and brought the ends out to a cooler location > you'd not have the problems that Bob mentions. > > > Russell > part 2 823 bytes content-type:text/x-vcard; name="Nino_Benci.vcf" (decoded base64) begin:vcard fn:Antonio L. Benci n:Benci;Antonio L. org:Monash University;School of Physics adr:;;BLD27 / RM118-119;Monash University;VIC;3800;Australia email;internet:KILLspamnino.benciKILLspammonash.edu title:Professional Officer / E&IS Manager / OHS Representative tel;work:+613 9905 3649 tel;fax:+613 9905 3637 note;quoted-printable:Albert Einstein=0D=0A= =E2=80=9CIf a cluttered desk is a sign of a cluttered mind, then what are= we to think of an empty desk?=E2=80=9D=0D=0A= =0D=0A= Rick Cook=0D=0A= "Programming today is a race between software engineers striving to build= bigger and better idiot-proof programs, and the universe trying to buildb= igger and better idiots. So far the universe is winning" url:http://www.physics.monash.edu.au/ version:2.1 end:vcard part 3 181 bytes content-type:text/plain; name="ATT00001.txt" (decoded base64) -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at mailman.mit.edu/mailman/listinfo/piclist

Here's what's confusing me... I tend to think of the wattage rating as the ability (or rate) of the device to dissipate heat (given it's package size, surface area, materials, etc), which is why I imagine higher wattage being no problem as long as the heat is being sucked away. For your situation and anyone else who has had power resistors fail, were they within specified max operating temp (but higher than rated wattage)? Or did the temps go over the specified max operating temp? Cheers, -Neil. Quoting Bob Blick <RemoveMEbobblickTakeThisOuTftml.net>: {Quote hidden}> Hi Neil, > > They fail open. The voids in the molding process and looseness at the > weld point between nichrome and lead-wire are the fail points. Remember > I am using ceramic-sand-and-concrete resistors. But in my past I have > seen resistors like you are using fail in similar ways. > > If this is not a commercial product, I'd say "just go for it". If you > get a dud once in a while, you'll fix it. My application is on a diesel > truck with serious vibration and operation down to -40 degrees, so the > thermal shock is considerable. > > Cheerful regards, > > Bob > > > On Sunday, October 02, 2011 6:30 AM, "PICdude" wrote: >> How do these things fail? This is not a volume or commercial product, >> but a one-off 3D printer head. Resistors are these ones from Yageo... >> search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=810F5R0E-ND >> >> So I have some options now... >> (a) Use an 18V source (perhaps add a 6V battery in series with the 12V >> battery, so each resistor would generate 10.8W of heat), >> (b) Build a PWM switcher now and put all 4 resistors in parallel, then >> adjust for 10W to each resistor, >> (c) Swap the resistors to 15-ohm units (same physical size and 10W), >> which in parallel will be 9.6W per resistor, >> (d) Run 28.8W through each resistor and see if I can get it to work >> for some time before it fails. >> >> (a) is complicated cause I'll have to buy a 6V battery to experiment. >> (b) will take time to get components for a PWM switcher, but probably >> makes most sense. (c) is pointless, cause I might as well try (d) >> first and if they fail I'd then do (c). >> >> Think I'm leaning to see if I can dig up a power mosfet to build a >> PWM-switcher temporarily using a function generator. >> >> Cheers, >> -Neil. >> >> >> >> Quoting Bob Blick <spamBeGonebobblickspamBeGoneftml.net>: >> > Hi Neil, >> > >> > I have some experience using cheap ceramic resistors as heaters in a >> > commercial product. To sum up my experience, you will experience a >> > higher-than-desirable failure rate if you significantly overdrive them, >> > even if the temperature is low, the duration is short, and feedback >> > limits the temperature. I was using 5 watt resistors. 6.4 watts under >> > worst-case seemed to be the sweet spot for me. More than that and I >> > could expect some failures. >> > >> > So 28.8 watts from a 10 watt resistor is about 15 watts more than I >> > would allow in a volume product. >> > >> > You could probably hand-select resistors or use "qualilty" brands and do >> > OK. But don't expect cheap ones to all survive. >> > >> > Cheerful regards, >> > >> > Bob >> > >> > -- >> > http://www.fastmail.fm - Access all of your messages and folders >> > wherever you are >> > >> > --

PICdude wrote: > Here's what's confusing me... I tend to think of the wattage rating as the > ability (or rate) of the device to dissipate heat (given it's package size, > surface area, materials, etc), which is why I imagine higher wattage being > no problem as long as the heat is being sucked away. Yes, but it isn't as simple as you think. The manufacturer comes up with the overall steady-state power rating with full knowledge of the internal structure and materials in the resistor. They know that if the external surface is at the specified temperature, the internal parts won't be stressed to failure at the specified power level. When you overdrive the resistor, the internal temperature distribution changes, and some spots may become hotter much faster than the external surface does. You really need to consider all of the paths and materials that lie between the spot where the heat is produced and the spot where it is dissipated to "ambient". It's all about thermal resistances -- and, in dynamic situations, the specific heat of materials (thermal capacitances, if you will) comes into play as well. -- Dave Twee

On Mon, 03 Oct 2011 06:17:40 -0700, you wrote: >Here's what's confusing me... I tend to think of the wattage rating as >the ability (or rate) of the device to dissipate heat (given it's >package size, surface area, materials, etc), which is why I imagine >higher wattage being no problem as long as the heat is being sucked >away. True as long as the thermal resistance from the element to the outside world is not high enough to be an issue. As mentioned elsewhere, welds may be a weak point. If you want to look at in detail, you could measure the element temperature via its change in resistance. I've seen a grossly overloaded metal-clad aluminium resistor violently fire one of its ends off, leaving a taser-like trail of (live) wire behind it...!

Interesting data. Do you know what the max specified operating temp was on these? And I assume the power to these resistors was higher than rated right? This answers a lot. Cheers, -Neil Quoting Antonio Benci <RemoveMEnino.benciTakeThisOuTmonash.edu>: {Quote hidden}> We've done similar things with 20W and 100W resistors. > > The most consistent failure was the soldered joint from resistor > tabs to wires. At any elevated temperature above 80degC we noticed > that the joints, over time, would fail and become open cct. This was > cured by either bolting or crimping wires to the resistor tabs. > > We've run 20W resistors (ceramic casing) up to 100degC for up to 10+ > hrs before element failure. 100W up to 160degC for close to 40+ hrs > before element failure (ceramic/aluminium casing). > > Nino. > > On 3/10/2011 2:14 PM, RussellMc wrote: >> On 3 October 2011 02:30, PICdude<picdude3EraseME.....narwani.org> wrote: >>> How do these things fail? This is not a volume or commercial product, >> I used commercial white ceramic 5W and 10W resistors to "power" a >> Fluidyne Stirling Engine (small and for fun). These were run at about >> rate wattage with the aim of being "as hot as possible". From memory >> these were placed in contact with a metal tube, possibly layered over >> with high temperature silicone rubber and wrapped in fibreglass >> insulation (intended for in-wall insulation). These lasted "for a >> while". Given that the engine only got occasional use I'd suspect they >> lasted 10's of hours of run time. This was OK for the application. >> Temperature unknown. Failure O/C. >> >> If you used Nichrome and brought the ends out to a cooler location >> you'd not have the problems that Bob mentions. >> >> >> Russell >> >

On Monday, October 03, 2011 6:17 AM, "PICdude" wrote: > Here's what's confusing me... I tend to think of the wattage rating as > the ability (or rate) of the device to dissipate heat (given it's > package size, surface area, materials, etc), which is why I imagine > higher wattage being no problem as long as the heat is being sucked > away. For your situation and anyone else who has had power resistors > fail, were they within specified max operating temp (but higher than > rated wattage)? Or did the temps go over the specified max operating > temp? They were part of a de-icer, clamped and glued to a metal plate, and shut off when they reached a little over room temperature. So they never got very hot(externally). Bob -- http://www.fastmail.fm - One of many happy users: http://www.fastmail.fm/docs/quotes.html

Not shown in that pic is a bunch of kapton tape wrapper around the resistors and aluminum block. It helped a lot, but it needs just a bit more. If I can find fire-cement or similar in a small quantity I would try that too, or just build a nichrome-wire head. But that cement has been the reason for the resistor-heater design. Think it's time I hit the 3D-printer forums. Cheers, -Neil. Quoting EraseMEalan.b.pearcestfc.ac.uk: {Quote hidden}> I have had failures on that style of resistor where the magic smoke > came out. I think there was an impulse current not too much higher > than the current required for maximum dissipation. > > But looking at the photo of your head, I suspect you could get it a > lot hotter by simply putting some Rockwool or similar insulation > around it. You are going to be losing most of your therms to the > atmosphere, and some simple, but high temperature insulation will > make a world of difference, especially as you say the temperature is > 'almost there'. Personally I would try this before attempting to get > more heat out of the resistors. > > -- > Scanned by iCritical. > >

> Not shown in that pic is a bunch of kapton tape wrapper around the > resistors and aluminum block. It helped a lot, but it needs just a > bit more. If I can find fire-cement or similar in a small quantity I > would try that too, or just build a nichrome-wire head. But that > cement has been the reason for the resistor-heater design. Think it's > time I hit the 3D-printer forums. I doubt that just going to the white cement ones will change much for you, AFAICT they are the same in the channel in the cement as fed into the inside of the aluminium block and cemented in using similar cement. The Al block will have a lower thermal resistance from element to outside surface. The problem you have is that most of the surface of the resistor is not in contact with the part you are attempting to heat, so 75% (roughly) of the resistor surface area is wasting heat for you. It will take somewhat more than a few layers of kapton tape to sort it, although that may give enough thermal resistance to show what is really needed. -- Scanned by iCritical.

Exactly why I *was* okay with them blowing up, but originally asked to see if there was any way to reduce potential damage. Now though, I am concerned with the one report of a resistor violently blowing up. The automated workcell was built for multiple purposes and we thought we could make a 3D printer out of it too with the addition of an extruder head. This is just for the fun of it and I really don't need it for anything useful, sensitive to downtime = almost zero. Cheers, -Neil. Quoting David VanHorn <RemoveMEmicrobrixspam_OUTKILLspamgmail.com>: > Is it really worth all this sweat? How hard are they to replace? > How sensitive are you to failure? (cost and downtime) > > Try running a pair at the anticipated power level 24/7 and see how > long they last. >

> -----Original Message----- > From: EraseMEpiclist-bouncesspamspamBeGonemit.edu On Behalf Of PICdude > Sent: Sunday, October 02, 2011 9:31 AM > > How do these things fail? This is not a volume or commercial product, > but a one-off 3D printer head. Resistors are these ones from Yageo... > search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=8 > 10F5R0E-ND When thinking about using a power resistor as a heater don't forget the power derating for temperature. The datasheet linked from the DigiKey page has the derating specs a bit out of place (just above electrical instead of below power). The spec is linear 100% @ 25C to 0% @ 275C, so @ 150C the power handling is only 50% (5 Watts for the 10 Watt version). Paul Hutch {Quote hidden}> > So I have some options now... > (a) Use an 18V source (perhaps add a 6V battery in series with the 12V > battery, so each resistor would generate 10.8W of heat), > (b) Build a PWM switcher now and put all 4 resistors in parallel, then > adjust for 10W to each resistor, > (c) Swap the resistors to 15-ohm units (same physical size and 10W), > which in parallel will be 9.6W per resistor, > (d) Run 28.8W through each resistor and see if I can get it to work > for some time before it fails. > > (a) is complicated cause I'll have to buy a 6V battery to experiment. > (b) will take time to get components for a PWM switcher, but probably > makes most sense. (c) is pointless, cause I might as well try (d) > first and if they fail I'd then do (c). > > Think I'm leaning to see if I can dig up a power mosfet to build a > PWM-switcher temporarily using a function generator. > > Cheers, > -Neil.

Heat loss as pointed out earlier is worth thinking about. (1) The resistors are designed to cool themselves, so mill the "heatsink fins" off. (2) If the block could be milled for a slot you could get more contact between the resistor and the block. (3) It almost looks like you could bolt another two resistors on; even if smaller, since every bit helps. The other thought I had that wasn't covered already is that if you have a cheap plentiful supply of resistors, you could subject them to large overdrives and observe their failure modes. (May be some good YouTube material here :) Then do successively lower power tests and watch how the destruction time increases. Maybe you can then predict how long they would last at your desired power level

I *think* that such power derating curves are usually intended for a certain set of conditions - normally the resistor in its typical mounting, surrounded by open air, so that the temperature mentioned in the derating curve is the ambient air temperature, not the resistor surface temperature. Otherwise, that would imply that the resistor needs active cooling in order to use it at its normal rated power. Some components, like MOSFETs, are typically rated this way (i.e., they give you some ridiculous max power dissipation for a *case* temp of 25C, which would only be achievable if you had it attached to a cold plate with liquid nitrogen running through it). However, I'm under the impression that resistors are not rated that way, especially those which are not designed to be heat-sunk. That is to say, if you really are sure that the ambient air will never be more than 25C, then a 10W resistor *can* actually dissipate 10W and live for its full normal rated lifetime. Someone please correct me if I'm wrong! Sean On Mon, Oct 3, 2011 at 11:14 AM, Paul Hutchinson <spamBeGonepaullhutchinsonSTOPspamEraseMEyahoo.com> wrote: > When thinking about using a power resistor as a heater don't forget the > power derating for temperature. The datasheet linked from the DigiKey page > has the derating specs a bit out of place (just above electrical instead of > below power). The spec is linear 100% @ 25C to 0% @ 275C, so @ 150C the > power handling is only 50% (5 Watts for the 10 Watt version). > > Paul Hutch

Yes you are correct and I should have written "ambient temperature" instead of just "temperature" in my first sentence. What I was responding to was the "How do these things fail?" opening that was quoted in my post. I was trying to point to a possible explanation for why the resistor would fail even when the nominal power rating was never exceeded. Let me, hopefully, clarify my point. When you use a resistor as a heating element you will either be unintentionally or, in the case of an enclosed space heater, intentionally increasing the ambient temperature. An increase in ambient temperature causes a decrease in power handling capacity for most if not all resistors. I've seen a number of cases where this effect is why a power resistor used as a heater causes the resistor to fail when used near its nominal power rating. The particular 10 Watt resistor that was pointed out has a derating factor of 0.4% per degree rise in ambient temperature. So that resistor would only have a 9.5 Watt capability with a mere 12.5 degree increase above the rated ambient temperature. If you can ensure that the air entering this resistors heat sink never exceeds the 25 degree ambient specification then it should not fail with 10 Watts applied. Looking at what I just wrote, I'm thinking I may not have clarified the point well. I suck at writing quickly and I'm not going to spend the time it takes for me to write well. Paul Hutch > {Original Message removed}

> Looking at what I just wrote, I'm thinking I may not have clarified the > point well. I suck at writing quickly and I'm not going to spend the time it > takes for me to write well. Paul - came across perfectly clearly. You said: Resistors are rated at full power only at say 25C and as ambient temperature rises their power handling capability is derated. So, if a resistor is used to add energy to air but the incoming air that it heats is always at 25C then it should operate safely at full power. But, if a resistor is surrounded by the air that it heats then it's ambient temperature will rise as the air heats, and the manufacturer's derating formulae or tables must be consulted. Russell