MFR0623 Axial resistors are only 1/4W rated

[johansen]

I have to say i successfully guessed the OP was farringdon from seeing nothing more than this on my phone :

"MFR0623 axial resistors are only 1/4W rated.
However, they are 6mm x 2.3mm in dimension,"

[tggzzz]

I have to say i successfully guessed the OP was farringdon from seeing nothing more than this on my phone :

"MFR0623 axial resistors are only 1/4W rated.
However, they are 6mm x 2.3mm in dimension,"

Accurate, and a good snark

[Xena E]

To Faringdon:
Snip the leads close to the body of the TH resistor and blob the SMD part in the gap, use HMP solder, and the jobs a goodun' lover boy.

☆☆☆☆☆☆☆☆☆☆☆☆☆☆☆

To the Forum:
The real issue here is that the manufacturers are going to rate their components according to any maxima that they are capable of, given the materials of construction: if nothing else it's good marketing.

The problem is that it doesn't help the designer who doesn't fully read the spec., sheets; it's a bit of a schoolboy error to assume for instance that power ratings in any particular resistors mean they are suitable for any duty provided that they meet V•I calculations.

In the conjunction that you go to the manufacturers data looking for suitability for a particular application, a temperature rise figure in free air is the nearest thing that would be of any use, and specifying maximum dissipation at an arbitrary ambient temperature may be a good quick look up, but wouldn't make good headline figures for them.

Floobydust's comments are valid, and I've sometimes shared that frustration with the 'banner ratings', though spec., sheets generally aren't BS however, the relevant information is usually down to the user to extrapolate, or they Gotcha! We've all seen burned up badly designed boards.

I will second the creation of spreadsheets for resistor ratings; something we also do for use in higher power scenarios, it's only a minor PITA.

OT:
The real problem that I have connected with this, particularly with SMD, is when the bean counters start questioning why, (for example), a design may have 10 parallel 10ohm 1W resistors when 1 single 1 ohm one will 'do'. I now refuse to repeat myself to them, but do have a previously enjoyed list of indexed design decisions on the system, and their reasons for existing. I just answer the official queries with bullet point references, (it saves typing, or even cutting and pasting as I do it from a crib sheet thats bluetacked behind my desk), also each time tell them its outside their remit to question engineering decisions: with little knowledge being a dangerous commodity.*

* anyone who has this particular time suck to deal with, can use this idea with free licence it let's the bean counters know that you consider their existence being as pointless to you as indicators on a BMW.

Regards
Xena.

[tggzzz]

...
The problem is that it doesn't help the designer who doesn't fully read the spec., sheets; it's a bit of a schoolboy error to assume for instance that power ratings in any particular resistors mean they are suitable for any duty provided that they meet V•I calculations.

You aren't finking of anyone here, are you? Shurely shome mishtake!

My pet amusement is those fleabay/aliexpress impossibly small switching PSUs claiming 6V-100V -> 5V 3A. The phrase "getting burned by" is literally relevant Good luck mounting them or attaching a heatsink!


EDIT: changed tag from [ i ] to [ img ], doh!

...
it let's the bean counters know that you consider their existence being as pointless to you as indicators on a BMW.

What is the sound of two birds being killed with one well-aimed stone

[ebastler]

The real issue here is that the manufacturers are going to rate their components according to any maxima that they are capable of, given the materials of construction: if nothing else it's good marketing.

The problem is that it doesn't help the designer who doesn't fully read the spec., sheets; it's a bit of a schoolboy error to assume for instance that power ratings in any particular resistors mean they are suitable for any duty provided that they meet V•I calculations.

In the conjunction that you go to the manufacturers data looking for suitability for a particular application, a temperature rise figure in free air is the nearest thing that would be of any use, and specifying maximum dissipation at an arbitrary ambient temperature may be a good quick look up, but wouldn't make good headline figures for them.

What kind of shadow-boxing discussion are we having here? The very first post by Faringdon did link to the relevant Vishay datasheet. And that datasheet does, of course, show hotspot temperature as a function of power dissipation, derating as a function of environmental temperature etc.. And it does point out the (rather high) maximum film temperature which they are basing their specs on, right on the first page in a prominent place.

The insinuations that (a) manufacturers are not telling the truth, and/or (b) engineers are too dumb to check the datasheets, are just not correct.

[Xena E]

The real issue here is that the manufacturers are going to rate their components according to any maxima that they are capable of, given the materials of construction: if nothing else it's good marketing.

The problem is that it doesn't help the designer who doesn't fully read the spec., sheets; it's a bit of a schoolboy error to assume for instance that power ratings in any particular resistors mean they are suitable for any duty provided that they meet V•I calculations.

In the conjunction that you go to the manufacturers data looking for suitability for a particular application, a temperature rise figure in free air is the nearest thing that would be of any use, and specifying maximum dissipation at an arbitrary ambient temperature may be a good quick look up, but wouldn't make good headline figures for them.

What kind of shadow-boxing discussion are we having here? The very first post by Faringdon did link to the relevant Vishay datasheet. And that datasheet does, of course, show hotspot temperature as a function of power dissipation, derating as a function of environmental temperature etc.. And it does point out the (rather high) maximum film temperature which they are basing their specs on, right on the first page in a prominent place.

The insinuations that (a) manufacturers are not telling the truth, and/or (b) engineers are too dumb to check the datasheets, are just not correct.

No shadow boxing...

No assertion that the manufacturers are pedalling bullshit... (reputable ones like Vishay anyhow).

Only labouring the point that designers need to carefully read the specs and translate them to their applications rather than go by the banner specification. It's a PITA but has to be done.

I deduce that someone is either:
A friend of Faringdon.

Or:

Drives a BMW?

[ebastler]

I deduce that someone is either:
A friend of Faringdon.
Or:
Drives a BMW?

Not really, and no.

But while Faringdon has often posted questions which were apparently off the cuff, without doing much research of his own before -- I wanted to do him justice in the present case, where he did link to a relevant datasheet right away.

And in a way, the subsequent discussion validates Faringdon's initial assumption: Besides real differences in resistor design (and specifically the acceptable operating temperature), there also are varying degrees of bullishness among manufacturers. Hence specs cannot alway be compared at face value, but only after looking into the spec details -- if the manufacturer provides them.

[Xena E]

I deduce that someone is either:
A friend of Faringdon.
Or:
Drives a BMW?

Not really, and no.

But while Faringdon has often posted questions which were apparently off the cuff, without doing much research of his own before -- I wanted to do him justice in the present case, where he did link to a relevant datasheet right away.

And in a way, the subsequent discussion validates Faringdon's initial assumption: Besides real differences in resistor design (and specifically the acceptable operating temperature), there also are varying degrees of bullishness among manufacturers. Hence specs cannot alway be compared at face value, but only after looking into the spec details -- if the manufacturer provides them.

Exactly my point concisely put, thank you.

[Xena E]

...
The problem is that it doesn't help the designer who doesn't fully read the spec., sheets; it's a bit of a schoolboy error to assume for instance that power ratings in any particular resistors mean they are suitable for any duty provided that they meet V•I calculations.

You aren't finking of anyone here, are you? Shurely shome mishtake!

My pet amusement is those fleabay/aliexpress impossibly small switching PSUs claiming 6V-100V -> 5V 3A. The phrase "getting burned by" is literally relevant Good luck mounting them or attaching a heatsink!

https://www.eevblog.com/forum/beginners/mfr0623-axial-resistors-are-only-14w-rated/?action=dlattach;attach=1920786

...
it let's the bean counters know that you consider their existence being as pointless to you as indicators on a BMW.

What is the sound of two birds being killed with one well-aimed stone

Your deduction is correct

Your comments on the ebay crappo is also amusing as well as accurate... I've had that shit held up as an example to me before, with comments such as if thay can do it why can't we?

We can do it! But do "we" want to ruin "our" reputation?

And to be clear I actually like proper BMW cars (RWD with at least a straight six), just not some of the cocks who drive them.

[tggzzz]

The insinuations that ... engineers are too dumb to check the datasheets, are just not correct.

The evidence is against you, in this thread at least!

[floobydust]

Trusting the resistor manufacturer's data I think is bad idea. Worse now with these super/ultra mini parts.

Did you see the picture Post #16 Yageo FMP100 "1W"? It's the size of typical 1/4W part lol.
This is what I mean for the race to the bottom- for size, cost requiring some deception because one manufacturer does it and others follow.
"Ultra miniature size" gives the usual derate curve above 70°C - yet Yageo gives NOTHING about temperature rise. Same for Yageo MFR Series.

Vishay SFR25 rated high 200K/W (in unknown test jig) so 80°C rise at rated 0.4W so what is the point of this part being good for 155°C if it's not going to run that hot? Why does it need the high temp materials then?

And how is the surface temperature test performed? Free air? On a PCB? Big copper traces/pads?
To what standard jig are they measuring this? IEC 60115-1 4.14 is not mentioned. It's important the test is realistic, practical as to how EE's will be mounting them, and that the temperature rise spec is real. Right now, EE's are clueless, naive, blindly trusting about the billboard power rating and it's just another marketing trap.
Vishay is good enough to provide temp rise numbers but we don't know the test conditions, if they are practical or magical values.

[ebastler]

Vishay SFR25 rated high 200K/W (in unknown test jig) so 80°C rise at rated 0.4W so what is the point of this part being good for 155°C if it's not going to run that hot? Why does it need the high temp materials then?

Are you now complaining that they are not running them hot enough?

Come on, you must have come across the difference between "absolute maximum ratings" and "recommended operating conditions" in other datasheets before. Would you prefer if they had specified it such that it would reach 155°C when run at the specified power at room temperature?

[floobydust]

Well, how hot is too hot - for a manufacturer to shrink a component's size and hide the fact it will burn your finger? 

Also pointing out the illogic in the datasheet's numbers or the part's claims = marketing bullshit.
Rated resistor power is P₇₀ or in a 70°C ambient, SFR25 is "0.4W". As a max. spec it appears to be achieved operating at 155°C, for the billboard power rating.
Temperature rise 200K/W says it's gonna rise 80°C above ambient... but derate the part above 70°C ambient. So the numbers appear off.
Effectively, this part is _W.
Can we figure this out, for the beginners and noobs to avoid the trap.

[ebastler]

Also pointing out the illogic in the datasheet's numbers or the part's claims = marketing bullshit.
Rated resistor power is P₇₀ or in a 70°C ambient, SFR25 is "0.4W". As a max. spec it appears to be achieved operating at 155°C, for the billboard power rating.
Temperature rise 200K/W says it's gonna rise 80°C above ambient... but derate the part above 70°C ambient. So the numbers appear off.
Effectively, this part is _W.
Can we figure this out, for the beginners and noobs to avoid the trap.

Now you got me confused. 70°C ambient plus 80°C above ambient (at 0.4W rated power) equals 150°C operating temperature. Is your concern that it is 150°C vs. 155°C maximum permissible film temperature?

[Siwastaja]

So when a manufacturer has a magical small size resistor with the same power rating - it's basically bullshit as you can't break the laws of thermodynamics.

What bullshit you are spewing.

Show me the law of thermodynamics that says a certain size of resistor can dissipate certain amount of power. Obviously it doesn't exist.

The following can be different:
* Maximum allowable temperature for internal materials. A simple example: polyethylene manages maybe 100degC without melting, but glass melts at 1700degC.
* Evenness of thermal dissipation. If the resistor hotspots somewhere, said maximum material temperature is locally exceeded even at lower average power density.
* Thermal conduction from the internal element to the surface
* Surface emissivity of the resistor

New resistors with increased power density are just better. And as a designer, your job is to make sure your surrounding parts can take the power, too. Nothing's changed; if you assumed with old resistors "oh, the resistor can take it, so my PCB can, too", you ended up with charred FR4. Many such examples from 1980's.

Trusting the resistor manufacturer's data I think is bad idea.

Trusting manufacturer data has served me pretty well for two decades, but before you can trust something, you have to understand what it means. It's a well known fact that using component datasheets requires some care and expertise, but that is not limited to resistors (or "new" resistors). For example, MOSFET datasheets are among the least understood IMHO.

But quite frankly, if 70+80=150 < 155 makes you confused, I don't know what to say. Maybe you should be doing something else? In my opinion, resistors front-page-rated to 70degC ambient is quite fair. Many MOSFET front page specs are at junction temperature forced to 25degC!

[Siwastaja]

My pet amusement is those fleabay/aliexpress impossibly small switching PSUs claiming 6V-100V -> 5V 3A. The phrase "getting burned by" is literally relevant Good luck mounting them or attaching a heatsink!

Yeah. And actually their power density (conversion power, not dissipated power) isn't impossibly high at all; if they were modern-day high-efficiency designs, something like 5V 3A would be easy-peasy in that form factor (I remember doing 5V 8A in half that size or so). But they are not modern; they aren't even synchronous (so diode loss at 5V is massive, and that small poor diode definitely can't handle 3A), so they simply dissipate too much for given output. Solution: do not heatsink, fix them to dissipate less!

The best kind of thermal design is minimizing dissipation to begin with, i.e. aiming for maximum possible efficiency, and because thermal design is difficult to some people (as evidenced in this thread) and heatsinks (plus mounting materials) expensive, total cost is usually lower even if you have to spend a bit more on silicon.

And this is what can be seen especially in power electronics designs (PSUs, amplifiers, motor drives etc.). 1980's stuff, linear or switched, had those large TO3 transistors, those "good old" toasty power resistor which toasted the FR4 or even melted the solder joints. Compared to that, even that Chinese module doesn't look too bad, and if you go from a $0.50 design to a $1 design, it can run pretty cool without heatsinking at all beyond the PCB itself. I'm pretty happy with the current times.

[floobydust]

So when a manufacturer has a magical small size resistor with the same power rating - it's basically bullshit as you can't break the laws of thermodynamics.

What bullshit you are spewing.

Show me the law of thermodynamics that says a certain size of resistor can dissipate certain amount of power. Obviously it doesn't exist.

The following can be different:
* Maximum allowable temperature for internal materials. A simple example: polyethylene manages maybe 100degC without melting, but glass melts at 1700degC.
* Evenness of thermal dissipation. If the resistor hotspots somewhere, said maximum material temperature is locally exceeded even at lower average power density.
* Thermal conduction from the internal element to the surface
* Surface emissivity of the resistor

New resistors with increased power density are just better. And as a designer, your job is to make sure your surrounding parts can take the power, too. Nothing's changed; if you assumed with old resistors "oh, the resistor can take it, so my PCB can, too", you ended up with charred FR4. Many such examples from 1980's.

Trusting the resistor manufacturer's data I think is bad idea.

Trusting manufacturer data has served me pretty well for two decades, but before you can trust something, you have to understand what it means. It's a well known fact that using component datasheets requires some care and expertise, but that is not limited to resistors (or "new" resistors). For example, MOSFET datasheets are among the least understood IMHO.

But quite frankly, if 70+80=150 < 155 makes you confused, I don't know what to say. Maybe you should be doing something else? In my opinion, resistors front-page-rated to 70degC ambient is quite fair. Many MOSFET front page specs are at junction temperature forced to 25degC!

It's not bullshit - surface area and heat radiation as well as convective losses are related. I'm not saying it's a limiting factor - just that a smaller part will run hotter.
It's not always OK, even if the resistor's materials are good for high temps that still ends up transferred to the PCB and nobody likes that to get discoloured or burnt. Do you know when standoffs are required? Were they part of the thermal resistance test? You don't know it all and acting like you do is annoying.

Insults and criticism aside, the datasheet thermal resistance number is from an unknown configuration.
Is that mounted to a PC board, open air, standoffs, forced air etc? No standard jig is mentioned. Maybe it goes back decades, who knows.
Because the SFR25H is the same size yet achieves a lower value, I suspect from better (copper) conduction heat losses to the jig, the resistor is mounted to "something". It also has a different coating which might radiate better.

Yet you get pissy about me not trusting that number. Who really blindly trusts any manufacturer nowadays? Many specs are stretched.
So I've asked Vishay for details, since people here don't know anything about it. IEC 60115 I don't have access to 4.14 if that is the test method.