Thin and Thick film SMD resistors.

jdolecki

I have an Op Amp circuit for EMG control of my robot.
Since the voltages are so small, Micro and Milli volts I was wondering if the added precision of the thin film resistors is worth the added cost.
Does anyone have experience in specifying thick and thin film resitors in circuits?
Newark list a 1k thick film resistor at .021 cents
And a thin film resistor at .419 cents.
So we are talking a substantial price difference
thanks john

Phil Pilgrim (PhiPi)

Just to be clear, is that $0.00021 compared with $0.00419? Or $0.021 vs. $0.419?
In the former case, I'd say, "Who cares?" unless you're buying a million of them. In the latter case, yeah, it might make a difference.
-Phil

jdolecki

Just to be clear, is that $0.00021 compared with $0.00419? Or $0.021 vs. $0.419?
In the former case, I'd say, "Who cares?" unless you're buying a million of them. In the latter case, yeah, it might make a difference.


2 cents and 41 cents. 
I know you build and design circuit boards cause I have a few. when you spec SMC resistors do you normally use the thick film or thin film or do you always go for the lowes cost?

Yanomani

Hi jdolecki

As usual, a diagram is worth a thousand words, mainly when it comes to solve any problem related to electronic circuits.
Then, forgive any misconceptions of mine, in the following assumptions.

As a general rule, when working with signals of any level, introduced noise must be controled, at least to an order of magnitude of the expected measurement range.

As far as I can foresee the problem, I'll make some basic assumptions to analize the situation and attempt to preview the shortcomings:

- I'm assuming that the signals you are going to measure are AC coupled to the Op Amp's inputs;

- If the signals you are sampling are in the range of 1Hz to 1kHz, then current noise (uV/V) phenomena will rule.
In this case, voltage dividers, intended to bias the inputs to proper levels, will be the main concern and thin film resistors should be used to not introduce significant noise.

- If otherwise, the range is above 1kHz, thermal noise predominates.
Then, selecting resistor technologies displaying low positive temperature coefficients will help.

I have found many useful information at:

http://www.vishaypg.com/docs/49997/49997.pdf

Hope it helps a bit

Yanomani

Domanik

@jdolecki:
In the formula for shot noise, the value of the resistor is important; smaller is less noise.  Shot noise may not be an issue in a 1K ohm circuit.  Other sources of noise may be of greater concern.

EMG control of your robot sounds interesting.  How far along are you with a functional design? 

edit: For anyone interested here's a link:
http://www.instructables.com/id/Muscle-EMG-Sensor-for-a-Microcontroller/