Question for the Oscilloscope Experts

[rsfoto]

Your circuit has many parts that are not needed.

Capacitor C2 passes AC but not DC so the second opamp amplifies vibrations with a gain of 23 times but does not amplify its DC input offset voltage.
The first opamp has a gain of only 2 so it is not needed and the gain of the second opamp can be doubled instead.
You are using a polarized capacitor for C2 which has AC across it that is wrong. If the values of R4 and R5 are increased then capacitor C2 can be a small film (non-polarized) capacitor.
C3 is SHORTING the output of the second opamp so it should be removed. If you want to cut high frequencies then add an RC filter at the output of the opamp, my R6 and C3 cut frequencies above 97Hz.
You do not need two rectifying opamps.
My added diode prevents the rectifier opamp from saturating as low as it can go.

Thanks AG for the circuit. Will breadboard your circuit today afternoon. I do not have a Film 1uF capcaitor. Can I take an Elko ?

 

[rsfoto]

Hi AG,

Question from a Dummy or better said ¿ DO I understand your full wave rectifier correctly ?

The AC wave goes into the NonInverting input and Diode D2 letpass the Positive part. The negative wave is rerouted through Diode 1 into the Inverting Input and inverted to a positive Wave which fill up the space between the positive wave ¿ Is my thinking correct ?

Thanks for the 5MΩ hint. Yes that is a 4.7MΩ but I measured it and it gives me 4.93MΩ and that is why I put there 5MΩ. Next time I will stick to the correct denomination of the resistor. Sorry

regards Rainer

 

[KeepItSimpleStupid]

My bigger comments is to short C1 and short D1 (easy to test)

Glad you used a bipolar supply.

Possibly add full wave precision diode.

Really, nice work for a novice! Just in time for Christmas.

 

[rsfoto]

My bigger comments is to short C1 and short D1 (easy to test)

Glad you used a bipolar supply.

Possibly add full wave precision diode.

Really, nice work for a novice! Just in time for Christmas.

Hi KISS,

Thanks will try and see what is the outcome on the Oscilloscope.

The second OpAmp TL082 is according to me, a Precision Full Wave rectifier or " Full Wave Precision Diode " as you call it.

 

[rsfoto]

Your circuit has many parts that are not needed.

Capacitor C2 passes AC but not DC so the second opamp amplifies vibrations with a gain of 23 times but does not amplify its DC input offset voltage.
The first opamp has a gain of only 2 so it is not needed and the gain of the second opamp can be doubled instead.
You are using a polarized capacitor for C2 which has AC across it that is wrong. If the values of R4 and R5 are increased then capacitor C2 can be a small film (non-polarized) capacitor.
C3 is SHORTING the output of the second opamp so it should be removed. If you want to cut high frequencies then add an RC filter at the output of the opamp, my R6 and C3 cut frequencies above 97Hz.
You do not need two rectifying opamps.
My added diode prevents the rectifier opamp from saturating as low as it can go.

Just breadboarded your circuit with components I have here.

Is it supposed to give me a rectified output ? Because what I saw is a variable Sine Wave at the output (typical for a vibration curve eg. Car shock absorber) in the same way as I get when I use my circuit

 

[cowboybob]

Is it supposed to give me a rectified output ? Because what I saw is a variable Sine Wave at the output (typical for a vibration curve eg. Car shock absorber) in the same way as I get when I use my circuit

That's what I got with a SIM as well, rsfoto:

Which makes sense since first one diode conducts and then the other does and so, presto - a sine wave out. Not sure what this part of the circuit is supposed to do but rectification, best I can tell, is not what its doing.

 

[audioguru]

Use three 0.33uF (330nF) film capacitors in parallel to make 1uF. An elco might conduct like a diode when it has reversed polarity every cycle of the input signal.

Yes, my halfwave rectifier circuit was wrong. You almost had it correctly.
Here is a good halfwave active rectifier and a good fullwave active rectifier:

 

[rsfoto]

Use three 0.33uF (330nF) film capacitors in parallel to make 1uF. An elco might conduct like a diode when it has reversed polarity every cycle of the input signal.

Yes, my halfwave rectifier circuit was wrong. You almost had it correctly.
Here is a good halfwave active rectifier and a good fullwave active rectifier:

Hi AG,

Thanks. I have seen those active rectifier circuits and many many more. So in any case if I use an active full wave rectifier circuit I will need a double Op Amp chip

My second Op Amp is an active full wave rectifier I will think about an active rectifier and see if I can live with a 4 diode bridge rectifier, which would make the output by far more simple

Thanks

 

[rsfoto]

Hi AG, KISS and cowboybob,

How can I avoid that the 1MΩ resistance of my oscilloscope probe does not influence the measuring value ? Using a capacitor and if yes any value would do it ?

Another problem is that I have a slight negative amount of voltage = -11.9mV at the output of my active FW rectifier.

 

[KeepItSimpleStupid]

If you use a x10 or greater "real oscilloscope probe", the effective input Z become 10 Meg ohms.

The scope's 9 M is in series with the 1 Meg in the scope.

This https://www.radio-electronics.com/info/t_and_m/oscilloscope/oscilloscope-probes.php is a better explanation.

Very short version:
The scope is typically 9 M in parallel with 22 pf. A 10 x probe is typically 1 M with a variable capacitor. The coax of the probe adds capacitance and you WANT the scope to see something purely resistive.

The artifacts show up at high frequencies and the edge of a square wave is composed of many frequencies (sorry, short answer says I won't discuss further). So, you get artifacts like shown in the photos. These ARTIFACTS are adjusted out by using the "calibrate" signal of the scope which is typically a kHz and a small amplitude.

So you adjust the variable capacitor on the probe, until you gate a square wave. On the x1 scale , if the probe has one, the adjustment is covered on most probes.

You could easily breadboard a X1 buffer and set the input Z to say 10 M with a resistor. Definitely not the best idea.

==

You still have to remember these few rules

1) The voltage across a capacitor can't change instantaneously
2) The current through an inductor can't change instantaneously.

3) Twisting attenuates EMI (Electromagnetic interference)
4) Shielding attenuates RFI (Radio frequency Interference)
5) Earth Shields should be attached at one end (preferably the source of the signal).

6) Long distance transmission should be done differentially (think ethernet) or current (4-20 mA)

7) Avoid ground loops by design.
8) Simple rule: Types references can be kept together BUT only tied together at "one point". It's hard to achieve.

9) Follow the manufacturers recommendation for "bypassing". They are capacitors VERY CLOSE to an IC's power pins and elsewhere as required. They are "REQUIRED".
10 Bypassing can be a parallel combination of different types of capacitors. (reasons I won't go into) except the capacitor construction gives the capacitors different characteristics.
11) Stuff that's seemingly unimportant can matter. e.g. The foil side of a non-polarized mylar capacitor can matter what side it's connected too in a circuit. Yep, a non-polarized cap being effectively polarized. e.g. less hum when connected when the foil goes to the low impedance side.

12) Don;t try to make pico-amps of current flow into a load of nearly 0 ohms. I did once,

13) Read the datasheet.

14) An unpowered circuit has Vss and Vdd = 0. So Vdd +0.3 means 0.3V

15) the elecron flow and "conventional current flow" are in opposite directions. Usually it does not matter.

16) I'l tell you that I have measured the resistance of a sheet of paper. 99.99999% of the time most people cant.

17) Numbers used to analyze can fool you. A 14 MW power plant is minus 14 MW. Power being positive is power dissipated. Generated power is negative.

18) Sometimes wierd stuff matters: 3 1K resistors in series is not 3K. SMT resistor mounted sideways have a different component value.A fingerprint matters. Conduction of electricity may be better in hollow silver plated tubes.
I challenged answers in college and won. One time I was told, your not supposed to know that yet. A solder (type of solder) copper joint may have more offset than another. A clean joint has less offset than one made form an oxidized copper. As I said, sometimes wierd stuff matters.

So, there's a few things that may help.

 

[audioguru]

A TL084 has four of the same opamps that are in a TL082 that has two opamps. A TL081 has only one opamp.
The 4 diodes bridge does not work in your simple circuit. Your second rectifier opamp is not an active fullwave rectifier because it simply duplicated the same halfwave from the first opamp rectifier.

 

[rsfoto]

A TL084 has four of the same opamps that are in a TL082 that has two opamps. A TL081 has only one opamp.
The 4 diodes bridge does not work in your simple circuit. Your second rectifier opamp is not an active fullwave rectifier because it simply duplicated the same halfwave from the first opamp rectifier.

Hi AG,

Thanks, Yes the numbering is known to me

By the way you have much more experience in Electronics then me but I do not agree with you about the statement

quote: " ... it simply duplicated the same halfwave from the first opamp rectifier ... "

Look at the analysis of one of the frames of my video I posted before.

My logic tells me that if that would be the case, then I should be seeing same 2 peaks repeated which is not the case. Look at the attached image

I will do a test and measure my FW rectifier.

 

[rsfoto]

Hi KISS,

Somewhere you mentioned the calibration of the Oscilloscope. I made a screenshot of probe 1 and so far I think the calibration is quite good

 

[KeepItSimpleStupid]

Yep, it's just that a 10x probe (common) increases the input Z from 1M to 10 M.

A half wave rectifier will keep the same frequency in, A full wave rectifier will double the ripple frequency.
e.g. 60 Hz half wave rectified --> 60 Hz ripple. Full wave rectified --> 120 Hz

 

[rsfoto]

Yep, it's just that a 10x probe (common) increases the input Z from 1M to 10 M.

A half wave rectifier will keep the same frequency in, A full wave rectifier will double the ripple frequency.
e.g. 60 Hz half wave rectified --> 60 Hz ripple. Full wave rectified --> 120 Hz

Hi KISS;

Thanks for confirming that I have a FW rectifier at the output look at the values at the pointing green arrows.



Quote: " Yep, it's just that a 10x probe (common) increases the input Z from 1M to 10 M. "

What do you want to explain me with that ? Measuring with 10X does it make more exactly or is it the same ?

I found this in the Internet https://info.tek.com/www-abcs-of-probes-primer.html

As you can see, I am trying to learn

 

[KeepItSimpleStupid]

FW: Looks like confirmation to me.

10 M: The tek full version is too complicated for now. Look here: https://electronicdesign.com/test-a...ive-and-active-oscilloscope-probes-your-tasks

A 10 X compensated probe on a 1 M || 22 pf scope makes your oscilloscope probe appear to the circuit as a10 Meg resistor.
If you "extended" the probe with a piece of coax, you may not be able to compensate it.

RG-174 is a thin 50 ohm coax. It has a capacitance of about 22 pf/foot. The whole point is the cable also has attenuations based on frequency and length. With a scope, you want the tip of the probe to be resistive. Hence, compensation. usually it's associated with high frequencies, but a 10 Hz square wave with a very fast rise time requires a bandwidth greater than 10 Hz to see it.

The probe is like a voltage divider, but it really is an impedance divider, hence the variable capacitor.
The scope is made intentionally complex 1M || 22 pf to accommodate probes that have capacitance.

 

[rsfoto]

Hi KISS,

This is the spec of the probes which came with the cheapo Hantek 6022BE



Would I get any advantage buying better probes or is the cheapo Hantek 6022BE not worth it ?

I have to say it works very stable with the original software.

 

[rsfoto]

I have a question.

I am working with my circuit as you can see below. I may sound stubborn, but that is the one I have found most useful so far in regard to sensitivity Gain of 100x eg, resistor R5 is 1MΩ



Now having glued the piezo on the wall I am getting a signal, red circles. From where does it come and how can I kill it. If my calculation is correct it is a 3000 to 4000Hz frequency




Thank you

 

[rsfoto]

I have a question.

I am working with my circuit as you can see below. I may sound stubborn, but that is the one I have found most useful so far in regard to sensitivity Gain of 100x eg, resistor R5 is 1MΩ

View attachment 96146

Now having glued the piezo on the wall I am getting a signal, red circles. From where does it come and how can I kill it. If my calculation is correct it is a 3000 to 4000Hz frequency


View attachment 96147

Thank you

Answering my own question = I have changed the Elko 10uF C1 for a 1uF Film capacitor and eliminated C3 so far and still experimenting.

 

[cowboybob]

Now having glued the piezo on the wall I am getting a signal, red circles. From where does it come and how can I kill it. If my calculation is correct it is a 3000 to 4000Hz frequency

First thought is a mechanical source, such as an HVAC fan or heat/AC motor. Sympathetic harmonic vibrations from the 50/60Hz powered hardware in your home/office/neighborhood.

Don't forget, your piezo sensor is very sensitive across a wide range of vibration frequencies. You may end up having to insert some filters into your circuit to "dump" (or block) the unwanted noise.

Just to muddy the waters even more, here's an interesting, recent ETO thread of note: https://www.electro-tech-online.com/threads/mysterious-sounds.146595/