I am thinking clock or logic noise on the gnd. Possibly SMPS catch diode if frequency is between 70-100 MHz. Charge amps are usually inverting in design at the front end so the gnd ref really needs to be solid. A diff mode amp can reduce noise like this 30 dB, opto isolation can reduce it more. Also think linear power on an isolated island for the pre-amp. Test with a battery!

M

Hello Prasanna,

A first blind look at the images show that there are fine bursts of high frequencies repeating around 50Hz (20ms) rate. This could be your switching device carrying the SMPS noise into the half cycles rectified by the full-wave rectification.

If you look carefully with the time scale of your DSO, in both cases, the burst separation leads to approximately 10ms (Image 31 with 2ms/div scale) and similarly with two bursts in 20ms window (image30, 5ms/div scale) that supports my statement.

Try zooming on these bursts and check if any of your Switched Mode Power supplies are "NOT SITTING RIGHT" in the classroom!

Best luck!

-Prasanna Waichal

Dear Mr. Prasanna waichal, In my system there are no switching power supplies involved. Only linear supplies are being used.

Dear Mr.Rody, I have even tried powering the circuit with a battery suspecting that the 50 Hz transformer in the linear supply would induce such noise. Still this kind of noise is observed.

Dear Gandhiraj,

Please till us what is the signal at channel 1 and channel 2. Can you display the input of the amplifier?

The sources of noise may be the power supply, the ground, the amplifier itself or added to the input before entering the amplifier. It could be contaminated by the data acquisition process of the digital storage oscilloscope. The contamination by the sampling pulses is clear in your signal.

In order to avoid contamination of the signal by sampling you may better use analog oscilloscope.

It seems me that the hundred Herz pulses are superimposed on your signal.

Please show us the input signal.

wish you success

Could it be the florescent lights? I used to have problems when the new electronic ballasts were introduced. Not that this is a problem but your device may, or is photo-reactive and will introduce flicker noise from lab lighting. Also watch for crossover distortion in the pre-amp. Many designs get into problems going near zero and chirp as the control loop goes open.

good problem keep us posted!

Ok so post a schematic and we will get the resolved!

M

Dear Prasanna,

I would explore shielding the sensor cables with metalic mesh and earthing them to the data acquisition box.

If the noise comes from external eM fields you should see an improvement.

Good luck

Dear Prasanna,

In my experience, things like this are generated by switch mode power supply. It need not be yours, it can be another one on the grid, even to some equipment not related to your equipment. Some suggestions:

- Check the grid and the grid ground wires for the presence of this interference. Do you see it there too ?.

- You wrote that you powered the circuit with a battery, and still had it. Did you power also the measurement equipment from the battery ?. I mean, if it is on the grid, and you power the circuit from the battery, but the oscilloscope from the grid, you may still see it.

Best Regards,

Henri.

Hello Prasanna,

In my opinion this is either a classic example of Conducted and Radiated EMI or it is the ghost in your own house!!!

Sorry to use those funny words but I looked at your project activities under which you have asked this question.

I now strongly feel that it is the FAST SWITCHING PULSE that is bringing this HF noise into your signals.

Try to explore this more. As suggested by other mentors above, I would appreciate if you can share your actual schematic to comment about!

Best luck!

@Abdelhalim & @Prasanna waichal I herewith attach the schematics. The input signal is a very brief pulse of current from a photodetector.

Hi,

I primarily suspect the resistor R2 and the Capacitor C2. I am preparing a short document as response to convey what I am thinking. Also I would like to suggest to change the Diode (photodiode) powering. You have anode GROUNDed while CATHODE to a High Voltage (HV) supply through the resistor. Change the polarity is possible or atleast give a try by connecting Anode to -HV (Negative High Voltage) and directly connect Cathode to op-amp input. No resistor-capacitor (R2-C2) business.

If you want AC coupling do it at the output of your amplifier.

I shall provide a document to make it clear to you.

Thanks!

Without having component values, it is a bit difficult... But if R1 is high resistance, your circuit is lost likely catching some impulse(s) on the "-" leg between the capacitors resp. R1. Could well be from the driving circuits of fluorescent illumination. Sometimes the 'good old incandescent lamps' had their advantages.

As the noise seems not to originate from your grid-connected supply, you could try to shield the circuit - or switch off the illumination. Wouodn't be the first time...

Regards

Attached is the file but with schematic diagrams!

Best luck!

Hello Prasana Waichal, that is a nice way to convey your message, and an interesting approach to look at the cause of the issue. Definitely worth trying if the equipment  allows it (floating HV supply). However, I do not think that it is 'ringing' based on the input signal, as the bursts are also there when no input pulse is applied.

Prasana Gandhiraj, to eliminate some causes, these things could help, also for us to help you troubleshoot:

• Do you see the same signal if the HV supply is off / disconnected? 
• You could try to replace the photodetector with a normal diode with equivalent parameter (reverse voltage, reverse current, capacitance). That way you rule out any optical 'noise', although I do not suspect that as a source.
• Take your setup somewhere else! I've seen lots of interference due to table frames being grounded, HF transmitters (WiFi / cellular) being near to the setup, fluorescent lighting, etcetera. If possible, take your setup outside for a moment.
• Another thing, which is more suppression of the phenomenon than resolving it, is to see whether you can filter out the disturbance. What frequency range do you try to measure? In which frequency range is your useful signal? Can you filter the disturbance as close to the detector as possible? This -I must stress- is the least favorable option, because any change in the disturbing signal may render this useless. Better to solve issues at the source...

Good luck! If you find the cause, pleas post it here, that might really help others.

Dear P. Waichal,

There are many opinions from the colleagues. After the new data you supplied us i would expect that there is a direct interference from the control pulse generator to the amplifier input through air by electromagnetic coupling. This is because your amplifier input impedance is high. This is well known phenomenon. You can verify it by taking the control generator far away from the input of the amplifier or you can screen your amplifier by putting it into metallic box and grounding the metallic box. 

If you display your control pulse waveform one can see the direct correlation between it the interference pulses appearing in your circuit. 

the amplifiers normally used to amplify the photo current are called trans impedance amplifiers. You can find them as one integrated building unit.

In fact your circuit as a high input impedance circuit is sensitive to electromagnetic interference from any near by source specially the switching transients.

Shielding is one of the solutions to this problem.

Best wishes