Hi All,
I am measuring the resistance of a semiconductor bar using 4-probe method with a power source (6221) and nano-voltmeter (2182A).I am applying 1 microA DC pulse, and tried with different pulse frequency, pulse width and currents. I am observing huge sinusoidal noise during each resistance measurement. When I decrease the temperature of the sample, resistance increases; the sinusoidal noise amplitude also increases with resistance value. Again the noise frequency do not change with any change in the dc pulse.
I would highly appreciate it if you could provide any suggestions to reduce the noise.
Thanks.
Find the frequency of the interfering signal and filter it out. Usually this kind of signal is called interference rather than noise, and because it is just at one frequency, or close to it, it can be mostly removed with a filter either in the measuring equipment, of by filtering the data, by averaging over one cycle, for instance.
The sinusoidal interference is not noise, but a sinusoidal signal with an oscillation duration of 29 seconds, which can be masked out with a notch filter that is tuned to this frequency. This works very well with MATLAB, you just have to guess the frequency. This is also possible with EXCEL: You program a sine function and subtract it from the measurement signal. If the estimated values f = 35 mHz, A = 3 and the phase of the sine function are gradually optimized, the measurement curve remains.
Dear Malcolm White and Herbert Weidner, Thank you very much for your comments and suggestions.
Dear Malcolm White and Herbert Weidner , Could you please comment on the sources of this interference signal? Is it from the sample holder or due to improper electrical connection to the sample or any other source? I am looking forward to hearing from you.
I don't know where a frequency of about 35 mHz would come from. Something that changes about every 29 seconds - is there something in your lab that does that? Otherwise, if you are sampling, it could be a beat frequency between your sampling frequency and the interference frequency. Mains is often the problem, with interference at 50 and 100 Hz, in the UK. Is your sampling frequency close to your mains frequency? One way to remove that would be to trigger the sampling from the mains.
The frequency of the disturbance is noticeably small, the cause could be a slow temperature regulation. Is a thermostat used in any device?
- Badges
- Science topic
- Similar topics
- Entomology
- Resistance