Unwanted noise in oscilloscope diaplay

[watsons]

Hi I'm a beginner on using an oscilloscope and I'm sorry if my question may be repeated in other posts.

I was working on a circuit with the following materials:

(1) an Al plate (5mm thick)

(2) a disk-shaped piezo transducer with special electrodes (product note)(http://www.piceramic.com/products/piezo-elements.html)

(3) two wires

(4) a coaxial cable

(5) a Tektronix DPO4104 oscilloscope

So the PZT was glued on the plate, and wires were soldered on the PZT to connect to the oscilloscope.

I tested my setup by knocking the plate, and ended up with sudden voltage rise/drop for the knocks. But I noticed the continuous milli-volt noise signals throughout the whole process.

Are they actually noise or something else? How should I reduce them to less than 0.1V as maximum amplitude, or even none? Because I would latter want it to detect 0.3-0.4 V.

I've tried scaling, reducing record length, and switching acquisition modes to SAMPLE / PEAK DEFECT / HI RES / ENVELOPE / AVERAGE(with max value possible). But they didn't work.
Fig 1. Output Graph

Fig 2. Screen picture

Fig 3. Parameters of oscilloscope
.

.. .

 

[berkeman]

Welcome to the PF.

In the Coupling menu for each channel, you should be able to set the Bandwidth (BW) Limit to a lower LPF frequency like 20MHz or lower. Do you see that setting?

You can also add your own lowpass filter (LPF) to your sensing circuit. do you know how to do that?

 

[watsons]

Welcome to the PF.

In the Coupling menu for each channel, you should be able to set the Bandwidth (BW) Limit to a lower LPF frequency like 20MHz or lower. Do you see that setting?

You can also add your own lowpass filter (LPF) to your sensing circuit. do you know how to do that?

The lowest possible bandwidth to set is 20MHz for this model.

If I use LPF for the circuit, I will also need constructing a amplifier circuit to distinguish the noise(0.2V) and my desired signal(0.3V).

 

[Svein]

Is there a radio station in the vicinity?

 

[tech99]

Welcome to the PF.

In the Coupling menu for each channel, you should be able to set the Bandwidth (BW) Limit to a lower LPF frequency like 20MHz or lower. Do you see that setting?

You can also add your own lowpass filter (LPF) to your sensing circuit. do you know how to do that?

If adding a LPF there is an issue that the PZ device has a capacitive source impedance which is very high and the LPF will shunt most of the energy. I suspect the noise is pick up on the electrodes themselves and it might be possible to ground the plate to the coax outer to obtain a degree of shielding.

 

[Baluncore]

The continuous small background wave looks to me like AC power frequency. It should not be stopped by a LPF because the signal has higher frequency components than the AC power. If anything it could be reduced by a HPF, but that is not the best way to go.
Reduce the area of any wire loops in your circuit layout, then you will get less magnetic field coupling from nearby AC power currents.
Is your aluminium plate grounded to the oscilloscope ground, or is it being an AC power voltage antenna ?
You might have a "hum loop" between the earth/ground of the oscilloscope and the transducer circuits. What else is grounded where ?
It could be audio noise from a cooling fan or air conditioner.

 

[tech99]

If adding a LPF there is an issue that the PZ device has a capacitive source impedance which is very high and the LPF will shunt most of the energy. I suspect the noise is pick up on the electrodes themselves and it might be possible to ground the plate to the coax outer to obtain a degree of shielding.

On a philosophical note, it appears that no filter can work without there being resistance somewhere. This also includes any circuit intended to produce a phase shift other than 180 degrees.

 

[jim hardy]

I can't make out from that oscillloscope display what is horizointal time base.
Your first question about the background should be "What's its frequency ? "

I'd hook an audio amplifier to it and listen then tap the plate, spin a coin on it
you'll get a feel for the sensitivity
i'd guess your 'noise' is mechanical motion
and it looks like a mix of at least two frequencies close to line frequency
is there any big equipment nearby, maybe an air conditioner compressor ?

Try suspending your plate on some very soft springs to lower its mechanical cutoff frequency
If that doesn't help, then my guess wasn't right.

 

[Baluncore]

I can't make out from that oscillloscope display what is horizointal time base.
Your first question about the background should be "What's its frequency ? "

Top window of screen shows ripple. Data was gathered for two seconds at 500 S/sec.
Main window is zoom of that. The square [a] and circle (b) time markers are 500 msec apart.
At the bottom of the screen it seems to be reporting an FFT frequency peak at 21.35 Hz.
That is a rumbling noise, but the ripple looks like the sum or beat of two or three frequencies.

 

[jim hardy]

here was my best estimate

something resonates after his tap at about 50msec period, 20 hz
looks like there's ~30hz background ?
i wondered if there's a 1750 rpm motor nearby

 

[watsons]

Just to update, I tried putting a foam plastic plate under the Al plate and half of the "noise" amplitude reduced (though I want to reduce more)

 

[Baluncore]

I tried putting a foam plastic plate under the Al plate and half of the "noise" amplitude reduced

It is almost certain that you have a source of vibration reaching the Al plate. It probably comes from the floor, up the legs of the table or bench.

Consider suspending the Al plate from elastic or rubber bands so that it does not rest on a vibrating surface. If it is still microphonic, try changing it's orientation to the airbourne sound.

You can trace the paths followed by vibratory noise to the transducer through the table by tapping different points on the table and observing the amplitude of the impulse detected. Look for places far from the transducer that generate a significant result, modify the mounting and repeat the process.

 

[jim hardy]

Just to update, I tried putting a foam plastic plate under the Al plate and half of the "noise" amplitude reduced (though I want to reduce more)

in the spirit of experimenting

plastic plate is a spring, if perhaps stiffer than i had in mind
Next, try adding some mass

F=MA, add mass and you lower the cutoff frequency..
can you set a dictionary between the plastic and aluminum plates? A nice heavy Webster's ?

Are you old enough to remember record players? Better ones mounted their turntable on soft springs for just this reason.