Resistor wattage differences

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
G

Guest

Guest
I'm building the SSL and Mouser has a few resistors backordered, so I called DigiKey, who said they DO have the right values, but not at the right watts.

From what I've read, the watts measure the maximum wattage it can handle, so if I need a 27K metal film resistor, 1/2 watt, 1%, would it be ok if I got one at 1 watt? I think that would make the tolerance 5%, so that might not be ok.

Another one I need from them is a carbon film resistor, which already has a 5% tolerance. So I would assume a higher wattage would be ok since the tolerance is already that big.

Am I on the right track here at least? :?:
 
Higher wattage R's with everything else the same are almost always fine. Of course look at the package size to make sure they will fit.

The one difference which is usually of no consequence is that the higher wattage R's may have a higher self-capacitance. I can't think of an audio app where this would be significant except for the high-meg R's in condensor mics.

Whether a 5% tolerance part will do where a 1% is specified is another whole subject. But if the circuit specified a 27k, which is traditionally a 5% or worse tolerance value, the 5 percenter is likely fine. Need to see the circuit to know for sure.
 
In other words-Tolerence doesn't refer to the wattage. It refers to the accuracy of the resistance .

Joel
 
> the watts measure the maximum wattage it can handle, so if I need a 27K metal film resistor, 1/2 watt, 1%, would it be ok if I got one at 1 watt? I think that would make the tolerance 5%, so that might not be ok. Another one I need from them is a carbon film resistor, which already has a 5% tolerance. So I would assume a higher wattage would be ok since the tolerance is already that big.

Tolerance and Watts are VERY different things.

Yes, there are trends. Precision resistors are usually "small-medium" power. High-power resistors are usually needed in places where no great precision is required. However you can find, or order, high power resistors of very good precision. It used to be that you could not get high-precision in very-very small Watts: if a precision resistor runs hot, it isn't so precision; anyway americans can't do tiny and exact at the same time.

Generally, the watts shown on the plan is a minimum. A higher-Watt resistor will work, within reason (a 100 Watt resistor in a wide bandwidth amp may add too much stray parasitics). Actually, most designers don't bother to calculate the watts of every resistor. And if you are working with 30V (or +/-15V) power, then a 27K resistor probably can't be dissipating any more than 30V*30V/27K= 0.030 Watts. A 1/2W, 1/4W, 1/8W, or even a 1/16th watt resistor would be plenty. But it is insanely tedious to calculate all the watts, and buy different parts: most designers pick the cheapest size and use it everywhere they don't need something bigger.

When I was a boy, 1/2W was cheapest, 1W and 2W cost more, and 1/4W didn't exist. I see now that 1/4W is often cheaper. If I specified a 1/16W resistor, because it only needs 1/16 Watts, I probably could not find one and might have to pay a lot of money.

Tolerance is often over-rated. In most places in audio, 10% tolerance will work great. 5% is now the low-price standard, and you sure can do good things with just 1/4W 5% resistors. There are a few places where a pair of resistors can both be off by 5%, but should not be off more than 1% from each other. In production you specify tight tolerance so any two resistors will work. In DIY it is often expedient to get a strip of 5% resistors and measure-out matched pairs.

Metal film and carbon film.... I can't believe it will make a BIG difference either way. Metal is good in many ways, but very expensive. And some very fine gear was made with carbon dust (composition), which is much less perfect than modern carbon-film. There is apparently enough "looseness" in the old carbon-composition that it is prized by certain guitar-amp builders, it gives a special tone. But modern dirt-cheap carbon-film is really better than any non-exotic resistors available 30 years ago.

Metal film will hold together at higher temperatures than carbon-film, which may be why you are finding it in the larger wattages.

I think an SSL could be built with nothing but 1/4W carbon-film resistors, which are cheap even in 2% tolerance.
 
i was drifting through some audiophile design sites , and noticed a

few statements about using larger watts "was better sounding"

like 1 to 2 watt resistors instead of 1/4 or 1/2.

so reading above , this is ok to do .

not saying id pay 2000$ for a powercord to my amp or anything.

heh.


?
 
...larger watts rsistors as "better sounding" generally is typical audiophoolery nonsense IMHO, though it may have been true in one specific situation with specific brand resistors...

Carefull generalisation is not what the hifi-world is good at..

Jakob E.
 
Agree with Jakob for the most part. Whenever I hear people raving about the huge differences of what are nearly perfect parts I get skeptical.

What could make a larger resistor sound different than a small? Besides the self-capacitance being larger, mentioned earlier as a tiny effect except for super-high values, there are three other effects: voltage coefficient, excess noise, and self-heating shifts.

Voltage coefficient means that the resistance value is a function of the voltage across. It's typically tiny for any other than very high values, but when significant is an odd-order distortion mechanism for a.c. signals.

Excess noise, also sometimes confusingly called current noise, has typically a low-frequency "1/f" spectrum with a level dependent on the current through the resistor. Wirewounds are generally the lowest excess noise parts, with the bulk metal foils being close. When you see people specifying "low noise" resistors they are talking about ones selected for low excess noise---of course all resistors have a well-defined thermal noise spectrum based on the resistance value and temperature.

Self-heating shifts occur when the resistor has a temperature coefficient and the signal changes the power dissipation. When the thermal time constants are getting anywhere near the audio freq range you get distortion at low frequencies. Audio Precision has had to deal with this as a measureable effect in their analyzers when they migrated to SMD at the same time trying to push the residuals down below the few ppm level. Even with low tempco R's the dissipation swing and small thermal tau produced effects that spoiled performance.

Again, these are generally not large effects!
 
One further comment: If you are using wirewound for high power be wary of the self-inductance. There are ways to wind them for low inductance but the typical ones are just a bunch of wire wound on a form, that is, wound as an ceramic-core or sand-core solenoid.

I had a Phi*ips integrated amp oscillating in test the other day with sand-filled 4 ohm load resistors. At 7MHz, where the amp channels were screaming, the Q of the inductance of the resistors was surprisingly high, quite a bit higher than the intended speaker loads.
 
Larger wattages will help keep noise down in a tube circuit that uses carbon for plate resistors (old Fenders, etc). Changing the 1/2 watt 100K's to 1 watt 100 k's is a standard mod for all old Fender amps. It keeps the resistors from developing that crackling noise.
 
gyraf said:
...larger watts rsistors as "better sounding" generally is typical audiophoolery nonsense IMHO, though it may have been true in one specific situation with specific brand resistors...

Carefull generalisation is not what the hifi-world is good at..

Jakob E.

Ok, but if i made a mistake by populating the full audio path with 1w resistors, should i buy a new turret and start again? They are 1w aller Bradley cc resistors for la2as and pultecs...
 
JohnRoberts said:
Using higher wattage resistors should not be a problem as long as their specifications meet all requirements.

JR

Thanks you, so i let them like this, we Will see.
 

Latest posts

Back
Top