Question for the Oscilloscope Experts

[rsfoto]

Its TINA TI, Ver 9.3, the NOT free version from Texas Intstruments: . I've used the free version (https://tina-ti.software.informer.com/9.2/) but it lacks the bells and whistles. Scroll until you see the gray "Regular Version" button to download the free version. The "Premium version" is not free.

I like TINA because it has motors that turn, LEDS that light up, relays that move, etc.. Just suits me better when I run a sim.

Most members here, though, prefer LTSpice (https://www.linear.com/designtools/software/), also free and very powerful but, IMHO, with a considerably longer learning curve.

Thank a lot. Will take a look into it

regards Rainer

You need to hit the piezo VERY HARD with a hammer for it to produce a signal into the very low value 280 ohm resistor. I think you said it will be used to sense (small) vibration from a telescope turntable?
Most piezo vibration circuits have the piezo loaded with a high value resistor that biases the opamp input at 0V, then a diode prevents the piezo output from damaging the opamp input with a negative voltage. A zener diode is used to prevent a strong knock on the piezo from damaging the opamp input with a positive voltage that is too high.

Hi audioguru,

Yes I did setup that circuit and I did not get any reasonable response in the output.

What values would you recommend using fot that R C combination. I guess the resistor has to be big and the capacitor very small as my frequency is quite fast. I guess there must be somewhere a formula for that.

I think I will have to build a test vibration platform with a little motor and an excentric placed drive hammer in order to get a constant frequency so I am able to simulate some vibration.

Remember I am not an electronics expert.

regards Rainer

 

[Nigel Goodwin]

Yes I did setup that circuit and I did not get any reasonable response in the output.

What values would you recommend using fot that R C combination. I guess the resistor has to be big and the capacitor very small as my frequency is quite fast. I guess there must be somewhere a formula for that.

You need a very high input impedance. 4.7M or 10M should be OK - and the LM358 is totally unsuitable, one of the Texas TL07x/08x series would be a better bet.

A low impedance will cripple sensitivity, and also drastically reduce bass response.

 

[rsfoto]

Hi Nigel,

Thanks. I can get here in Mexico the following ones and will get a few of them. I will read the datasheet and try to set up some circuit.

TL084CN
TL081CP
TL082CP

I was using the LM358 because from other projects I had then laying around here

Please explain me the difference of the LM358 to a TL082CP if you have time

OK about the resistance and what about the capacitor ? I guess the capacitor does just flatten out my wave signal of the vibration ¿ correct ?

Thanks again and regards Rainer

Just looked at the datasheet and I see that there is a R/C component at the output connected to earth with a capacitor value but resistance there says only RL si my next question is, hoe do I define that Load resistor ?

 

[Nigel Goodwin]

The TL08x is a JFET input device, so has a high input impedance - the LM358 is bipolar, so has a relatively low input impedance.

 

[audioguru]

The LM358 has inputs that work all the way down to 0V when there is no negative supply voltage and its output goes down very close to 0V.
The TL07x or TL08x have inputs that do not work if they are a closer than a few volts from 0V and their output does not go closer than a couple of volts to 0V when there is no negative supply voltage.
You do not need a capacitor at the input, but since the piezo sensor has a resonance at about 80Hz then you might want to filter it out.

 

[rsfoto]

The LM358 has inputs that work all the way down to 0V when there is no negative supply voltage and its output goes down very close to 0V.
The TL07x or TL08x have inputs that do not work if they are a closer than a few volts from 0V and their output does not go closer than a couple of volts to 0V when there is no negative supply voltage.
You do not need a capacitor at the input, but since the piezo sensor has a resonance at about 80Hz then you might want to filter it out.

Hi,

So that means that if I kill the negative voltage of the piezo sensor via a rectifier the TLXXX are useless for me ? As if I understood correctly what you wrote the MiniSense 100 from MEAS when I do slightly tip on it does deliverup to 100mV and if I really give him a hard hit it goes up to perhaps 2 or 2.5V ...

Now I am a bit confused Sorry

regards Rainer

 

[audioguru]

You simply read the Absolute Maximum Ratings of an opamp on its datasheet and follow them.
The input of almost any opamp is destroyed if its input voltage exceeds the supply voltage. For an LM358 with only a single positive supply and its input biased at 0V then its absolute maximum negative input is -0.3V so you need to connect a Schottky (low forward voltage) diode at its input to prevent the piezo voltage destroying it. You also need a zener diode to prevent the positive input voltage from going higher than the supply voltage.
A TL07x or TL08x will not work in that circuit unless you add a negative supply voltage in addition to the positive supply voltage. They still need the diodes.

 

[cowboybob]

Now I am a bit confused Sorry

So am I...

Could you explain to us a little more information about the actual motion of the telescope pier you are trying to monitor, how it moves, why it moves, at what general range of motion/frequency, etc.?

I think you might try attaching the sensor to a pier and monitor, with an oscilloscope, the typical peak outputs and frequencies.

 

[rsfoto]

So am I...

Could you explain to us a little more information about the actual motion of the telescope pier you are trying to monitor, how it moves, why it moves, at what general range of motion/frequency, etc.?

I think you might try attaching the sensor to a pier and monitor, with an oscilloscope, the typical peak outputs and frequencies.

Hi cowboybob,

That is exactly what I want to do. Measure the vibration of the pier on which the telescope mount is mounted in order to know when it is vibrating too much. More or less a simple seismograph

A pier is where on the top a telescope is mounted with its mounts be it an equatorial or an ALT / AZImuth mount (also known as fork mount).

Frequency is not the same as every pier vibrates diferrently. The idea of this vibration sensor is not to measure the flexion of the pier but only detect when it vibrates. We can then make a test about how much is too much and therefore we use the images of the Night Sky which we make looking at the roundness of the stars.

That is why I am calculating with very low mV outputs from the piezo film because we will never have strong impact except if a meteorite hit our telescope but after that the whole Observatory is gone

That is the idea of this circuit.

BTW I also have an Acelerometer which I will test later on but with the Acelerometer I need to bring the base voltage of 1.66V down to zero and then just measure the differences of the 3 channels. This I will develop later when I get to a result with this piezo film sensors.

Yes the sensors do show a nice curve in the Oscilloscope but only when sampling over 500ms to 1s. I connected the sensor to the probe of the Oscilloscope (cheap Hantek 6022BE, which for my purposes at the moment is more then enough) ans tipped slightly on the sensor ans there is a nice shaped curve which is very high at the beginning and has a slow downwards. The period is very very fast. I will try to catch via screen shot such a curve and post it here.

 

[rsfoto]

Here are some samples with the Oscilloscope

Picture with the nice even curve is base noise = +-16.3mV



Below a curve when I really tip strong with the finger. Every rise is a tipping on the sensor.

 

[rsfoto]

EUREKA

it is working. The trick was to use a 10mΩ resistor parallel to the outputs of the sensor. The chip is the LM358

Circuit schematics



Input and Output at the Oscilloscope



Thanks for all your help. Learned a lot. Position correctly a Voltage Divider. Learning what is High Impedance.

Use an Oscilloscope for seeing what happens at high speeds in a circuit.

Now after trying so many different ideas the simplest solution is the one which works.

Now the next step is to integrate a LM3419 circuit as a visual meter.

 

[audioguru]

It looks like we are seeing the artifacts of a digital oscilloscope instead of seeing the sensor signal. What is causing the repetitive signal and what frequency is it (I cannot see the 'scope setting)? Is it mains hum picked up?

The maximum input bias current of an LM358 is 0.25uA which in your 1M input resistor will cause the input voltage with no signal to be up to +0.25V. The opamp amplifies it and its output can be 21.3 x 0.25V= +5.33V. An LM158A has a maximum input bias current of 0.05uA. Or if you do not want this offset you can reduce the 1M resistor to 100k ohms or less.

 

[cowboybob]

Good job, rsfoto.

I suspected that the max expected output(s) of the piezo (in this case) would be low enough that the negative swing would not be a problem for the input of the LM358.

 

[rsfoto]

It looks like we are seeing the artifacts of a digital oscilloscope instead of seeing the sensor signal. What is causing the repetitive signal and what frequency is it (I cannot see the 'scope setting)? Is it mains hum picked up?

The maximum input bias current of an LM358 is 0.25uA which in your 1M input resistor will cause the input voltage with no signal to be up to +0.25V. The opamp amplifies it and its output can be 21.3 x 0.25V= +5.33V. An LM158A has a maximum input bias current of 0.05uA. Or if you do not want this offset you can reduce the 1M resistor to 100k ohms or less.

Hi audioguru,

The signal is from me drumming on the breadboard with a finger. Look at the attached image. I drummed with the finger 2 times quickly with a pause inbetween.

 

[rsfoto]

Good job, rsfoto.

I suspected that the max expected output(s) of the piezo (in this case) would be low enough that the negative swing would not be a problem for the input of the LM358.

Hi cowboybob,

Thanks for the info. If you look at the Oscilloscope diagramm in message #30 you can see that I got -1.06V but there the piezo was not connected to the LM358. After connecting it and hitting hard on it in this configuration I get a max negative voltage of about -600mV.

I have kicked the piezo up to 1V and still working. It looks like the LM358 does not care about the negative voltage. So far it is still working

I guess I could add a rectifier to the piezo sensor so I get only positive voltage and also add a capacitor to smooth down the waves ?

What do you think ?

TIA

 

[rsfoto]

It looks like we are seeing the artifacts of a digital oscilloscope instead of seeing the sensor signal. What is causing the repetitive signal and what frequency is it (I cannot see the 'scope setting)? Is it mains hum picked up?

The maximum input bias current of an LM358 is 0.25uA which in your 1M input resistor will cause the input voltage with no signal to be up to +0.25V. The opamp amplifies it and its output can be 21.3 x 0.25V= +5.33V. An LM158A has a maximum input bias current of 0.05uA. Or if you do not want this offset you can reduce the 1M resistor to 100k ohms or less.

Hi audioguru,

On the attached image you can see the output when the piezo sensor is not being excited.

The yellow line is the LM358 output and the green line is the input signal into the LM358.

 

[audioguru]

Here is a digital trace with high resolution for a good 'scope and low resolution like yours.

 

[rsfoto]

Here is a digital trace with high resolution for a good 'scope and low resolution like yours.

Did you forget to attach anything ?

 

[audioguru]

Did you forget to attach anything ?

OOps, here it is:

 

[cowboybob]

To assist in calibrating your circuit, I'd suggest replacing resistor R1 (1MΩ) with a 1M linear potentiometer, wired as below:

This will allow you to more easily adjust your gain as you find the "sweet spot" for your system where average pier vibration generates an average indication level with your (eventual) LED light bar.