cheap pots over wire-wound for uC controls?

Status
Not open for further replies.

oddson

Well-known member
I bought an expensive wire-wound potentiometer (10KOhm) to see how much BETTER performance it would bring to a voltage-divider-as-controller.

I thought there may be little or no difference but instead the expensive pot is much worse in terms of induced noise on a stationary control in the middle of its range.

The 10mm PC-board-mounting cheap controls I normally use typically vary in the LSB only (10-bit readings) with occasional glitches to 2 bits but the wire wound has many 4 bit changes while stationary.

I'm wondering if the wire coil is actually picking up noise itself (despite what looks like good shielding).

Anyone else observe this?
 
The case always remains so in all models. Whatever the potentiometer it is and whoever the manufacturer is 95% of the times it is problematic and will continue to do so. For your specific circuit try different models somehow you'll get the perfect one for you. I do use potentiometers as last option (Since in my industry I get a lot of noise using them). Instead I prefer to use a simple resisitve based voltage divider, after doing calculations from some calc like this, I make my circuit. But in my case I always need constant voltage for example lets say I will be going with 5.6 V whole week. In case you opting for variable one you should use a pot and test a few models to find the best one. However for microcontrollers I'll recommend the voltage divider, also you can find some ready made voltage dividers providing a clean look to circuit as well as rid'ng you to get involved in design.
 
Do you have a 10n capacitor in there somewhere?

Not yet...

I was just an experiment... but if the 50-cent pot doesn't need one but the $6 one does...

Unless it's performance is actually better except for high frequency noise but it didn't appear that way from sampling at various rates.

I was just curious about whether wire-wound pot induced noise is a known, common thing.
 
Last edited:
Does not really matter whether its a pot, a wire, or a trace on a PCB when measuring voltages you need some capacitance or your going to be subject to any RF noise in the area.
 
It was the magnitude of the difference that surprised me... testing the effectiveness of passive filtering is next.
 
Maybe the wirewound pot has some inductance which is causing the ADC input to effectively be a higher impedance at the short duration where the ADC samples the signal?
 
Status
Not open for further replies.
Back
Top