Thick film power resistors linearity & noise

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

n3mmr

Member
Joined
Jan 1, 2019
Messages
10
Are thick film high power noninductive resistors, such as produced for automotive use, linear enough for audio?
At least enough to allow using them as "external loads" with a QuantAsylum QA402 and QA451?
 
Can you find a data sheet for them?

As I recall from my last real job, we had to use non-inductive resistor loads for testing power amps on the bench and production line.

JR
 
“Linear enough”? I doubt linearity is an issue. It’s a resistor, after all. In the small-signal audio world at least, it shouldn’t be a concern.

Noise, however, is where you should be concerned. Thick film resistors are cited as having considerably higher current noise as compared to metal film (thin film) or wirewound in the attached analysis. This is the most thorough study of resistor noise that i’ve come across. While other factors may play more significantly in low noise audio design, I personally wouldn’t choose thick film when other types are just as common and cost-effective.
 

Attachments

  • Resistor Noise Tests.pdf
    3.2 MB · Views: 27
Last edited:
Sorry,my bad.

The QA451 is a load switching device used when measuring the output characteristics of power amplifiers.

So I'm talking about, say, an 8 ohm 500W resistor hooked up as a load to a high power amplifier during testing.

Even quite severe excess noise in high-power resistors with values typical for audio loads should not be a problem. Since I'm working at the output end, not the input, of the amplifier


Linearity could be a problem, if you're measuring IMd in the audio range at high currents.
 
I suspect another concern is to be truly resistive. Besides stability, an inductive or capacitive load could cause small frequency response errors that could show up in bench measurements.

JR
 
John Roberts,
Truly low stray inductance is one of the few strong points of thick film resistors. Ww resistors can be wound in an inductance minimising pattern, but thick film naturally has the inductance of a single straight wire of the length of the resistor package.
 
There are a couple of variables impacting resistor linearity: self-heating and voltage coefficient. They are two separate phenomenon. Both characteristics are generally substantially worse with thick film compared thin film resistors. In addition, thick film resistors tend to exhibit relatively high 1/f noise compared to thin film. For these reasons, thin film parts are almost always preferred where analog signal fidelity is a consideration. Places where thick film resistors tend to still get used are applications like digital termination resistors, pull-ups, very low cost products, etc.
 
John Roberts,
Truly low stray inductance is one of the few strong points of thick film resistors. Ww resistors can be wound in an inductance minimising pattern, but thick film naturally has the inductance of a single straight wire of the length of the resistor package.
I've built several power soaks and experimented with different styles.
The most satisfying used 51 off 820 ohm 50W aluminium shell resistors (for a 2500W 4-ohm load) mounted on both sides of a large (20-inch) aluminium plate. Due to their short length their inductance is minimized. Of course, the lead inductance is a factor. I used a Zobel circuit to more or less compensate it.
The big advantage is the low voltage coefficient, that is dominant with other technologies.
 
Last edited:
In general, low valued thick film resistors will be cleaner than higher valued ones. I'm using a Yageo 5025 case RC series thick film 100Ω part in a power supply location, but it does get signal current through it, so I gave it a bit of concern. The datasheet says that the excess noise is only 10dB for 100Ω and lower, whereas excess noise irises from 24dB to 46dB for values from 100Ω up to 1MΩ. This part was heavily filtered as well, so it seemed safe (and measures well enough). For thick film, excess noise and voltage coefficient are related, since they are both caused by the granular nature of the conducting material. Just like a carbon composition resistor, lower resistance values will have less noise and voltage coefficient.

So, if you can use a very low value, a thick film will probably be good enough.
 

Latest posts

Back
Top