[SOLVED] noise level of voltage source vs current source

[nanock]

Hi,

assume there is a simple connection of an ideal voltage source with a resistor. on the other hand, there is a current source connection with a resistor. I want to measure the resistance in each circuit. now,

1- for measuring resistance of a 10k or higher, it is preferred to use voltage source or current source? I mean you prefer to use a voltage source with a resistor or current source with resistor for measuring the resistance? why?

2- please focus on details, linearity, noise level, stability, responsiveness and what ever you can.

thanks a lot

 

[KlausST]

Hi,

to measure R you need to know U and I.

1) If you have a constant voltage, then you need to measure I somehow. Usually with a known series resistor and measure the voltage drop.
* if you want to measure very low resistors you need a lot of current. This generates heat in the DUT. Maybe it overheats and gets damaged.
* if you want to measure very high ohmic resistors, then the current is low. SNR becomes worse. Resolution becomes worse.

2) If you have a constant current source
* if you want to measure very low resistors, then the voltage drop is small. SNR becomes worse. Resolution becomes worse.
* with high ohmic resistors your current source has limited voltage capability. this limits the measurement range.

10kOhms:
..Is somehow inbetween low and high ohmic resistors. So I see no general beneifit in constant voltage or constant current.
For sure it depends what magnitude of voltage or current you choose.

examples:
You may use 3.000V (as constant voltage source ) and get 300uA
Or you may use 100.0uA and get a 1.00V voltage reading.

My opinion:
For low ohmic resistors (below 100mOhms) I´d use a high current source (maybe 1A) and use 4 wire measurement (kelvin connection)
The lower the expected resistor, the higher current source I´d use.
For even lower resistors (in microohms range) i´d use maybe 10A of current, but with alternated connection (very low freuqwncy square wave) to ged rid of thermocouple effects.

For high ohmic resistors I´d use higher voltage.

***
Linearity, stability (noise also):
Avoid temperature rise of the DUT because of dissipated power = current x voltage.

Klaus