Hi,
As above I see that resistors are specd in PPM
I.e.
5ppm
10ppm
100ppm
Which should I buy, for on a valve amp project.
Thanks
As above I see that resistors are specd in PPM
I.e.
5ppm
10ppm
100ppm
Which should I buy, for on a valve amp project.
Thanks
That's probably the temperature coefficient in ppm/K (or ppm/degree C, which is the same). The lower the better for a discrete multibit DAC or for the gain setting resistors of an extremely-low-distortion amplifier, but utterly unimportant for a typical valve amplifier.
PPM stands for "parts per million". SO if your resistor changes in value 10 parts per million and it is a 220k resistor, consider what an incredibly tiny effect we are going to have. One ten millionth of 220,000 ohms. Per degree of course.
I can't say which should you buy but I can say that 100ppm-300ppm is a general grade TCR. And 5-10 ppm (<=50ppm) - means a precision type resistor.
I may wrong but in a good old lamp days, nobody used 5-10 ppm resistors.
If you don't use <1% resistors, it really doesn't matter what the tempco is, it's within the tolerance of the part.
FWIW in the 50's and 60's 20% tolerance was good enough, and I am talking *factory* suggestions.
let´s do a little Math:
* suppose your resistor lives inside a tube amp chassis, and internal temperature rises 50C above ambient.
* suppose you have a perfect resistor, measured value is exactly as printed.
* you want to know what PPM is so bad, that it will throw this otherwise perfect resistor *outside* of 20% tolerance?
Total PPM=20% (delta or variation)/50C=0.2/50=200000 (millionths)/(50 * 1000000)=200000 PPM (total variation)
To get variation per deg C: 200000/50=4000 PPM (per degree C)
So for a resistor to be considered unusable in a Tube amp, it should have worse than 4000 PPM spec.
Worst among suggested is mere 100 PPM? ....it is INCREDIBLY GOOD!!!!!!!
Old time Amp designers would have MURDERED to get such good parts.
let´s do a little Math:
* suppose your resistor lives inside a tube amp chassis, and internal temperature rises 50C above ambient.
* suppose you have a perfect resistor, measured value is exactly as printed.
* you want to know what PPM is so bad, that it will throw this otherwise perfect resistor *outside* of 20% tolerance?
Total PPM=20% (delta or variation)/50C=0.2/50=200000 (millionths)/(50 * 1000000)=200000 PPM (total variation)
To get variation per deg C: 200000/50=4000 PPM (per degree C)
So for a resistor to be considered unusable in a Tube amp, it should have worse than 4000 PPM spec.
Worst among suggested is mere 100 PPM? ....it is INCREDIBLY GOOD!!!!!!!
Old time Amp designers would have MURDERED to get such good parts.
5ppm means that the resistor will change its resistance 5 parts per million for every degree C of change in its temperature. So if you have a 1 meg resistor, its R will change by 5 ohms for every degree of temp change. 5ppm is very expensive, with 100ppm, what most metal films do, is common. 100ppm is way better than you need for any tube amp. Use 5ppm, such as Vishay's bulk metal models, only for very accurate instrumentation. It won't hurt at all to use those in a tube amp, but you'll never benefit from them in that application, and they will deplete your wallet in no time.
Certainly guitar amp designers wouldn't have lifted a drink to get super-PPM resistors; they knew it did not matter.
Stereo Hi-Fi may justify <200ppm on specific gain-setting resistors to get <1dB error side to side.
Even in gain-set resistors or those in RIAA phono eq. circuits, you'd be hard pressed to hear the difference. The temperature will typically not change enough to cause the change to be heard. If the resistor is a power resistor getting hot, that may be different, but that is not a gain setting sort of application.