What causes resistor distortion?

Jump to Latest
#44
Vovk Z said:
- But 10 VDC is the same as 10VAC, when we talk about heat.
Click to expand...

I guess you mean 10VAC RMS.
With AC, the resistor cycles through heating and cooling with the signal cycle. More so with lower frequencies.

Suppose with 10V DC the resistor heats up 10degrees. Now apply 10VAC. It will heat up on the rising wave (pos and neg), and cool of if the wave returns to zero. So it will not heat up 10degrees because before it gets there, the signal falls and the resistor cools a bit.

So with AC the temp rise will always be either equal to DC or lower, so DC is worst case.

Jan
 
#48
RF is much less demanding than audio for distortion. -50dB harmonics is very good at RF. At RF if a harmonic is a problem, you band-pass filter, problem solved.


There is a situation where small amounts of non-linearity do matter for RF, and thats in transmitting antennas and nearby metalwork, where non-linearity due to corrosion can lead to out-of-band emissions, or IM distortion that crosses from downlink to uplink frequencies, normally extremely well isolated by duplexers. Cellphone antennas are particularly sensitive due to being duplex (same antenna is transmitting and receiving at vastly different signal levels simultaneously).


Presumably solders for duplexers <-> antenna connection are where this is tested to the limit?
 
#51
I have tested different solder formulations. I used a PCB with 100 jumpers soldered in place for a total of 200 joints. Looking at the intermodulation distortion of around 90,000 & 93,000 hertz.

Conclusion, no significant distortion on any of the solders tested.
 
#53
Elvee said:
With AC, the temperature will be a waveform at twice the frequency, centered about the rms value, ie. DC of the same value.
It will thus oscillate about the DC value, sometimes higher sometimes lower
Click to expand...

The power will indeed be oscillating and sometimes be higher than the DC value. But the resistor has a thermal time constant. So the temperature will lag the power exactly the way the voltage on a capacitor lags the input current.

Unlike DC, at AC that cap voltage will not reach the final end value before the current reverses direction. Similarly, at AC the resistor will not reach the final end temperature before the power starts to drop again.

Jan
 
#55
jan.didden said:
I guess you mean 10VAC RMS.
With AC, the resistor cycles through heating and cooling with the signal
Click to expand...
- of cause, I'm talking about RMS value. Of cause, heat cycles.
There are at least two different frequency regions - low frequencies, under "termal resonance frequency", when temperature varies very much and rises max possible peak value, and frequencies several times higher that frequency, when there are only very small temperature variations. That " termal resonance frequency" usually is somewere at tenths-hundreds Hertz (for most usual parts. I may be wrong, and that frequency goes up to several kilohertz for small parts). Accurate termodynamics calculations isn't linear, so I don't want to talk about low-frequency AC. I'm talking about frequencies of 1000 Hz and higher - I believe there AC makes the same heat as DC - it's what RMS is.
 
Last edited:
#56
I think I would agree to that. My idea was that to determine the resistor variation resulting from a certain signal (power) level, you can use the DC-situation, and then the AC-situation will be either lower or the same.

The reason for my idea is that for instance for the Vishay Zfoil, only the DC PCR (Power Coefficient of Resistance) is available. And that I use as worst-case.

Jan
 
You must log in or register to reply here.
Top Bottom