I am trying to make an Instrumentation amplifier. My signal is in mV. I am getting my signal from shunt resistor. But problem is noise level. On oscilloscope I am getting 50 Hz noise and the amplitude of noise is greater than my desire signal.
I tried different techniques,
short probes,
battery supply,
shielding,
and I also used big value capacitor (1F, 2F). With a low pass filter plus high value capacitor I was able to remove the noise completely, but now problem is that small signal change across the shunt resistor is also filterd out.
Any suggestions?
See the picture below, well, ripple is not problem in this case, my concern is 50 Hz noise.
Hi Abid Hussain , May be you could use something equivalent to an isolation transformer for AC. i.e. something to physically separate output from input. Perhaps an optoisolator may work in your case.
Regards.
Hi Martin E.Baltazar-Lopaz, thanks for your answer.
I am already using isolation transformer.
I am getting my signal across a shunt resistor, and the resistor is a kind of floating. What I means is that I attached a series resistor across an electrode and electrode is attached to the positive terminal of the battery.
And with using differential amplifier I am getting this common-mode noise.
Have you tried using a thin film or wire wound shunt resistor? They have lower thermal noise than carbon composition resistors.
A while back, I had a similar problem and found that removing the grounding lead from a power strip that powered the oscilloscope lowered the AC noise that was being seen.
Thank Ira Chayut for suggestions, I will try wire wound shunt resistor.
For second part I think isolation transformer does isolate load from ground but anyhow I will remove the ground lead and will check the effect.
Ira Chayut I tired wire wound resistor and I also removed ground lead, but I did not see any difference.
I attached pictures of both test, you can see there is no difference. Well ignore these big spikes, its mechanical noise of motor.
Sorry that what helped me remove 60 Hz noise did not work for you. Please post if you find a solution. Thank you, Ira.
I solved my problem.
I want to point down few important things which I observed during my experiments, I think it will help others.
1. Proper shielding: Shielding, Amplifier should be shielded. I did several experiments, with and without shielding my circuit. Without shielding I noticed that even power supply increase the level of noise, this is true for short distance (less than 2 feet).
2. Shielded wire: Better to use shielded wire.
3. Battery Supply: Battery power supply always help in low level voltage measurement.
4. Short Ground leads: Use short ground leads for your oscilloscope. This is also an important factor.
5. Don't attach DMM: Digital voltmeter has long wires mostly. And if you are measuring low voltages in oscilloscope and you also attached DMM with your source, then you are in trouble.
6. By-pass capacitors: Using By-pass capacitor is a normal practice in circuit design.
7: Check your electrodes: If you are using electrodes better to use Ag not copper.
Note: In my case isolating AC power line of oscilloscope has no effect on noise. In FACT in some cases I noticed that isolation was creating noise. This is based on my own observation.
Hi Abid Hussain
If you have op-amp in your circuit, maybe using some capacitor in op-amp DC supply can improve noise level.
like 100nF + 100uF
I had this problem before. Try using an RF current probe (which is a clamp,🙃). This will remove your 50hz and then drop your floor, which will make it possible to visualise the actual emissions. It's a dynamic range problem.
Good luck.
- Badges
- Science topic