Fusible resistors-When do they fuse?

[djsb]

I am replacing 1 off 68 Ohm 250mW fusible resistor with a 22 Ohm and a 47 Ohm resistor in series (each rated at 1/3 watt) in the power supply of an FM tuner. It is in series with a 79M12A -12v (0.5 amp) regulator and has probably blown due to a mains surge.

https://www.vishay.com/docs/28737/nfr25.pdf

I can't make sense of the FUSING CHARACTERISTICS in the data sheet above. I will do some actual tests on them but I'm wondering if they will be OK. I've attached the schematic to show where the fusible resistor is (R288 lower right of schematic next to C220). I understand that there will be different voltage drops across each resistor and that each will therefore dissipate power differently as well. Just wondering which one would blow first and at what current? Hope someone has experience with these? Thanks.

[Miyuki]

They will melt somewhere in that striped area
So time variation is huge

They are meant more as last resort to prevent big bang when all other goes wrong

When you need defined blow characteristics you need a fuse

[tszaboo]

In my experience: never. I was testing some 2 or 3W rated though hole fusible resistors (name brand), as I was curious myself. It didn't have this nice graph. So I was running it on rated power -> fine. Overloaded it 200% -> fine. Overloaded it 10x the rated power-> The resistor started glowing yellow, and it was burning the desk. There was a black charred area on my workbench from that point, reminding me not to trust these components, ever.

[Miyuki]

In my experience: never. I was testing some 2 or 3W rated though hole fusible resistors (name brand), as I was curious myself. It didn't have this nice graph. So I was running it on rated power -> fine. Overloaded it 200% -> fine. Overloaded it 10x the rated power-> The resistor started glowing yellow, and it was burning the desk. There was a black charred area on my workbench from that point, reminding me not to trust these components, ever.

That is right according to the 3W Vishay datasheet  https://www.vishay.com/docs/28909/acxxcs.pdf

It won't fuse at 30W, it needs at least 50W to fuse

[bdunham7]

I am replacing 1 off 68 Ohm 250mW fusible resistor with a 22 Ohm and a 47 Ohm resistor in series (each rated at 1/3 watt) in the power supply of an FM tuner. It is in series with a 7912 -12v (1 amp) regulator and has probably blown due to a mains surge.

https://www.vishay.com/docs/28737/nfr25.pdf

I can't make sense of the FUSING CHARACTERISTICS in the data sheet above. I will do some actual tests on them but I'm wondering if they will be OK. I've attached the schematic to show where the fusible resistor is (R288 lower right of schematic next to C220). I understand that there will be different voltage drops across each resistor and that each will therefore dissipate power differently as well. Just wondering which one would blow first and at what current? Hope someone has experience with these? Thanks.

You find those things all over in audio stuff especially from the 70s and 80s.  The fusing characteristics are vague and typically have a long 'tail', so that they will eventually start to fail even if operated below their maximum ratings.  They are almost always used in low-power areas such as secondary power supplies where their characteristics--and often resistance--just aren't critical.  It is common to replace them with slightly more robust versions and in many cases, slightly lower resistances. 

To answer you question, in your case the 47R resistor would be dissipating over twice the power, so it ought to be the first to melt at high power.  However, that is so variable in real life that I wouldn't count on it.  Outside (below) the datasheet cutoff of 3W, their lifetime is pretty unpredictable, at least as far as I know.  You may be able to just use the 47R resistor alone and everything wil be fine. If you have a model and schematic, I can look at it and give you an opinion.  Is this resistor before or after the 7912?  And is the 7912 not heat-sinked?

[tszaboo]

In my experience: never. I was testing some 2 or 3W rated though hole fusible resistors (name brand), as I was curious myself. It didn't have this nice graph. So I was running it on rated power -> fine. Overloaded it 200% -> fine. Overloaded it 10x the rated power-> The resistor started glowing yellow, and it was burning the desk. There was a black charred area on my workbench from that point, reminding me not to trust these components, ever.

That is right according to the 3W Vishay datasheet  https://www.vishay.com/docs/28909/acxxcs.pdf

It won't fuse at 30W, it needs at least 50W to fuse

As I said, it wasnt Vishay, and it didn't have the curve. And I dont see the point. When the resistor acts as a yellow lightbulb, and it can easily set something else on fire, I dont see the reason/benefit to use it as a protective component.

[bdunham7]

As I said, it wasnt Vishay, and it didn't have the curve. And I dont see the point. When the resistor acts as a yellow lightbulb, and it can easily set something else on fire, I dont see the reason/benefit to use it as a protective component.

Flameproof and fusible resistors are two different things, although some resistors are both.  And neither is a fuse.  I've replaced quite a few and I've yet to see anything burned up from a hot fusible resistor that was simply overloaded.  Was your example a very low resistance?

[tszaboo]

As I said, it wasnt Vishay, and it didn't have the curve. And I dont see the point. When the resistor acts as a yellow lightbulb, and it can easily set something else on fire, I dont see the reason/benefit to use it as a protective component.

Flameproof and fusible resistors are two different things, although some resistors are both.  And neither is a fuse.  I've replaced quite a few and I've yet to see anything burned up from a hot fusible resistor that was simply overloaded.  Was your example a very low resistance?

10 Ohm, as far as I remember. This was a few years ago.

[djsb]

I am replacing 1 off 68 Ohm 250mW fusible resistor with a 22 Ohm and a 47 Ohm resistor in series (each rated at 1/3 watt) in the power supply of an FM tuner. It is in series with a 7912 -12v (1 amp) regulator and has probably blown due to a mains surge.

https://www.vishay.com/docs/28737/nfr25.pdf

I can't make sense of the FUSING CHARACTERISTICS in the data sheet above. I will do some actual tests on them but I'm wondering if they will be OK. I've attached the schematic to show where the fusible resistor is (R288 lower right of schematic next to C220). I understand that there will be different voltage drops across each resistor and that each will therefore dissipate power differently as well. Just wondering which one would blow first and at what current? Hope someone has experience with these? Thanks.

Is this resistor before or after the 7912?  And is the 7912 not heat-sinked?

It's before the input (although it's a -12v regulator so the current flows in the opposite direction). The -12v regulator is NOT heatsinked. The +12v regulator is (and it is getting nice and hot).

[mag_therm]

It would be useful if the data sheets published surface temperature vs time at various currents,
in with the data curves they already show.

[bdunham7]

It's before the input (although it's a -12v regulator so the current flows in the opposite direction). The -12v regulator is NOT heatsinked. The +12v regulator is (and it is getting nice and hot).

Looking at the schematic this appears to be simply a dropping resistor to minimize the dissipation of the 7912.  It probably has 3-5 volts dropped across it in normal operation and the circuit is probably 40 to 60 mA.  Fusibility is just traditional in these secondary supplies and since the whole thing is protected by 400mA fuses, this resistor is pretty clearly not critical. What you have represents a nuisance failure, not a power surge or anything like that.  If I were fixing it I might just use the 47R fusible that you have and add a very small heat sink to the 7912, or just use a 68R non-fusible 1/4W metal film resistor, 1/2W if you have one and can fit it in neatly.

[trobbins]

The failed part could have been a nuisance failure, but have you confirmed the 12V neg regulator is ok, and that the load current on it is "nominal" at 12V - ie. have you confirmed there is no other failed circuitry?

The original part would have been smaller and cheaper than any fuse at that time, but the designers likely determined that the upstream capability, even with 400-500mA fuse would not have avoided the R288 from destructing itself for some common faults.  Although the part is termed 'fusible', its feature is that it won't flame under any overload - whether it open's or not is a secondary issue.

If the original part is 0.25W and 68 ohm, then that indicates nominal feed current is noticeably below 60mA (ie. 4V drop).  The part would only reach fusible conditions at likely 10x the 60mA capability (ie. 0.6A), and that level may not take out the upstream fuse for at least an hour, depending on characteristic and how high the over-current ratio is, and whether the fault was downstream of the regulator and the regulator was current limiting or going in to thermal shutdown.

Perhaps check the nominal overvoltage available to the regulator, as you may be able to install a 91 or 100 ohm nfr25 part, and so operate it at near the original conditions as far as rated resistor wattage is concerned.  Also note the height above the pcb that the original part was installed.

[T3sl4co1l]

10 Ohm, as far as I remember. This was a few years ago.

Such types, in such values, are commonly seen in mains inrush protection for light* loads, lamps for example.

*Pun intended.

There, the overload is many times higher -- as stated above, and therein lies your explanation.

Personally, I've had a fusible resistor blow under a failed (shorted) vacuum tube.  It promptly blew out a small spark and became open circuit.  Nice.  Again, high voltage, in this case with capacitors handy to deliver the fault current.

If you need fusing at low voltages, or more generally, low peak powers, the answer is simple -- don't use a fusible resistor, use a more accurate type.

Tim

[Alti]

I am replacing 1 off 68 Ohm 250mW fusible resistor with a 22 Ohm and a 47 Ohm resistor in series (each rated at 1/3 watt) in the power supply of an FM tuner. It is in series with a 7912 -12v (1 amp) regulator and has probably blown due to a mains surge.

1A * 1A * 68 = 68W

[bdunham7]

1A * 1A * 68 = 68W

That's just the max current rating of the regulator (with a heat sink), but in this circuit the current is much lower.

[TheMG]

Fusible and flameproof resistors look alike. They can be difficult if not impossible to tell apart visually.

Most resistors of this type you are likely to encounter are simply flameproof resistors, not the fusible type.

The flameproof characteristic means when overloaded the resistor will not emit flame. It can still get very very hot.

As for the fusible type, I've mostly seen these used in small SMPS, where they perform a dual function of limiting inrush current and also acting like a fuse in case of a gross overload (primary switching transistor shorting). In that case, the exact value at which the resistor fuses open is not particularly important. Mostly a cost saving thing instead of using a separate fuse.

[Siwastaja]

The flameproof characteristic means when overloaded the resistor will not emit flame. It can still get very very hot.

... and set something else on fire!

As for the fusible type, I've mostly seen these used in small SMPS, where they perform a dual function of limiting inrush current and also acting like a fuse in case of a gross overload (primary switching transistor shorting). In that case, the exact value at which the resistor fuses open is not particularly important. Mostly a cost saving thing instead of using a separate fuse.

Yes. And as always, cost-saving on these bulk SMPS is based on not skimping on design at all, they are sold in millions so they have ample resources to pour say $10000 worth of time to save just $0.10 on the thing! Hobbyists and small-scale production professionals beware, we usually don't have such resources to prove a marginal cost-optimized design is actually safe, so just use an actual fuse whenever one is needed (and fuses are great!)

[djsb]

Thanks for all your replies. I've corrected my original post as I stated that the regulator was a 1 amp part. It is in fact a 0.5 amp regulator. I'm currently waiting for the fusible resistors to arrive. Meanwhile, I've checked for shorts across the supply pins of all the IC's and found non so far. I'll post more later.

[Alti]

I've corrected my original post as I stated that the regulator was a 1 amp part. It is in fact a 0.5 amp regulator.

Ok, so what does that change in the context of blown fused resistor? This is still 17W from a 0.25W resistor.
For a 0.25W you need 0.68 ohm resistor (680 miliohm) to survive the load.

[Siwastaja]

A 0.5A rated regulator doesn't necessarily mean it has a 0.5A load. Maybe the said negative rail uses very little current - at least normally. The reason for failure could be anything.

[TheMG]

The flameproof characteristic means when overloaded the resistor will not emit flame. It can still get very very hot.

... and set something else on fire!

Which is one of the reasons the "FR" in commonly used FR4 circuit boards means "Flame Retardant". Most other components are also made of flame retardant materials.

It is the product designer's responsibility to ensure that components that could become very hot either during normal operation or in a fault condition not be located in proximity to flammable materials, such as a plastic housing made of non-FR material.

As for the non-flameproof resistors, I've seen it happen where a 2W carbon-composition resistor overheated and caught on fire, and with the circuit board mounted vertically this caused several other carbon composition resistors to combust. The fire was contained since the circuit board was inside an aluminum case (for RF shielding and heatsinking purposes), but the board was pretty much ruined beyond any hope of repair. Had flameproof resistors been used, at worst the PCB would have been heat damaged in the area around the resistor but this would have been repairable.

Also picture a scenario where the above mentioned board would have been mounted in a device with a slotted metal case for ventilation, and say someone had left something combustible (paper for instance) on top of the case. The flames from the resistors burning would have extended out through the ventilation slots and the rest is history...

[djsb]

The tuner has been fixed now. I replaced the 5v and -12v regulators and kept the +12v as it was unharmed. I replaced the 68 ohm fusible resistor (0.25w) with a 47 ohm (0.5w) one. I did some tests on the 22 ohm and 47 ohm resistors using my power supply. The 22 ohm one fused at 0.6 amps. The 47 ohm resistor (in series with the 22 ohm one) fused at approx 0.37 Amps (the 22 ohm survived but got a bit charred). The 47 Ohm 0.3 W fused at approx 0.4 Amps and the 47 Ohm 0.5W one fused at approx 0.47 Amps. As the 47 Ohm 0.5W is closest to the regulators maximum current I fitted this one as a replacement. I measured the voltage across the 47 ohm replacement as the tuner is operating correctly (listening to Sounds of the seventies on BBC radio 2 as I write this) and measured 0.93 Volts (0.020 Amps current). Thanks everyone.