Question for the Oscilloscope Experts

audioguru said:

I forgot that the datasheet for the LM3915 shows a vibration meter circuit and a piezo transducer feeding a 1M resistor that works fine with the lM3915. Also, the LM3915 does not become damaged by negative inputs like the LM358 opamp. Then is the opamp needed anymore?

Click to expand...

Hi audioguru,

Let me explain the development of this story. From other projects I had already made I knew how the LM358 works more or less and therefore I started with the LM358. Then I want to implement the optical messaging about vibration remembered that I had done already somthing with the LM3914 chip. OK. as searching for the datasheet of the LM3914 again I saw that there also exist the LM3915 as well as the LM3916. So I started to read the datasheet of the LM3915 and suddenly I do see the vibration meter circuit.

So I went to buy some LM3915 in order to experiment.

Now after having this interesting conversation with all of you here I have to decide which way I go.

The circuit with the LM358 gives me a linear output and applied to a LM3914 would give me a linear messaging of the LEDs on the output.

The output of the LM3915 is Logarithmic output instead and so I see a problem that if my vibration signal is very low I would not get that much info as I would get with the LM358 and the optical signal at the output of the LM3914.

Having now circuits I will analyze the best output and go then for that one.

I = 6 V/650 Megs is not a lot of current. Protecting from negative voltages might be a good idea as AG says. The 650 is spec'd at 1V, then there's a blub somewhere are referring to 6V.

If you set an LM3914 and an LM3915 as sensitive as they can go then the 10th LED lights when the input is 1.25V.
The first LED on an LM3914 lights when the input is 0.125V and on the LM3915 the input is 0.06V. Therefore the LM3915 is more sensitive and has a wider range than the LM3914.

I doubt the source resistance of the piezo sensor is very high at 650M ohms because if it is then its signal feeding a 1M resistor would be reduced to almost nothing.

audioguru said:

If you set an LM3914 and an LM3915 as sensitive as they can go then the 10th LED lights when the input is 1.25V.
The first LED on an LM3914 lights when the input is 0.125V and on the LM3915 the input is 0.06V. Therefore the LM3915 is more sensitive and has a wider range than the LM3914.

I doubt the source resistance of the piezo sensor is very high at 650M ohms because if it is then its signal feeding a 1M resistor would be reduced to almost nothing.

Click to expand...

Hi,

Thanks. Well according to the LM3914 calculator https://www.electro-tech-online.com/tools/LM3914V2.php the first LED lights up at 0.06V also on the LM3914.

I am just testing the LM3914 with direct input from the piezo and it does exactly what I am looking for. So the next step will be to set up the LM358 amplifier and use a high gain so I can register at the minimum voltage of 1.25V at the amplifier output very tiny vibrations of the Piezo, eg. 25mv from the Piezo and a gain of 50 would end up as 1.25V into the LM3914. So my 25mv would have steps of 2.5mV.

I will try that now that I have the necessary potentiometers at home in order to finely adjust the gain ...

Will see what happens

The LM3914 calculator is wrong.
It turns on the 10th LED when the input is 0.95V instead of when its input is the same as its reference voltage which is 1.25V.
It turns on its 1st LED when the input is less than 0.06V but more than 0.05V when it should be 1.25V/10= 0.125V.
It turns on its 2nd LED when its input voltage is a little higher than 1.6V which is not double the voltage that turns on its 1st LED but 1.5V would be correct.
It turns on its 3rd LED when the input voltage is 0.25V when it should be 0.375V, etc.

The LM358 opamp is not good enough for your very high sensitivity because it is noisy (rumble and hiss) and it has a maximum input offset voltage of 7mV.

EDIT: The crossover distortion from the LM358 might cause errors with your measurements.

audioguru said:

The LM3914 calculator is wrong.
It turns on the 10th LED when the input is 0.95V instead of when its input is the same as its reference voltage which is 1.25V.
It turns on its 1st LED when the input is less than 0.06V but more than 0.05V when it should be 1.25V/10= 0.125V.
It turns on its 2nd LED when its input voltage is a little higher than 1.6V which is not double the voltage that turns on its 1st LED but 1.5V would be correct.
It turns on its 3rd LED when the input voltage is 0.25V when it should be 0.375V, etc.

The LM358 opamp is not good enough for your very high sensitivity because it is noisy (rumble and hiss) and it has a maximum input offset voltage of 7mV.

EDIT: The crossover distortion from the LM358 might cause errors with your measurements.

Click to expand...

WOW. That is not good ...

Thanks. Will use the LT08X which I have here ...

There is one more thing and that is the Minisense 100. Sometimes it delivers between 0.0V and 8.14mV and sometimes it goes up to around 25mV or values around that and this of course is amplified too ...

You wrote LT08x but I think you mean TL08x?
The maximum input offset voltage of the TL08x is 15mV which is more than double the input offset of the LM358.
The TL08x will not work in the circuit unless it has a an additional negative supply then its output swings below 0V which is not what you want.

How do you attach the vibration sensor to the telescope? If it is not a rigid mount then its output level will be less.

audioguru said:

You wrote LT08x but I think you mean TL08x?
The maximum input offset voltage of the TL08x is 15mV which is more than double the input offset of the LM358.
The TL08x will not work in the circuit unless it has a an additional negative supply then its output swings below 0V which is not what you want.

How do you attach the vibration sensor to the telescope? If it is not a rigid mount then its output level will be less.

Click to expand...

Hi,

Yes that was a typo, meant TL08x

OK, KeepItSimpleStupid tells me I need a TL07x or TL08x and now you tell me it is useless message #22

So that menas I can dump all the TL08x I bought ?

What Operational Amplifier would you recommend at a reasonable price not over US $ 5.00 ...

As the device is meant to be portable the question is how to get a negative Voltage input ... ¿ 2 separate batteries and wired how ?

you wrote

... then its output swings below 0V which is not what you want.

Click to expand...

and I guess yes because when going into the LM3914 ¿ a negative input is dangerous ?

My tests yesterday with the LM358 showed me that the LM3914 seems to ignore the negative input of my amplified sensor signal.

¿ How I attach the sensor to the scope ? not yet decided but I was thinking in using a magnet . ¿ How would this influence the signal if I use a long cable from the sensor to the circuit ?

I know that mounting the sensor with some double sided tape the vibration could be damped. I am thinking in making a little PCB and glue the magnet with epoxy resin onto it. That should be a non damping fixing.

Questions over questions

Hi to Anyone reading this thread

Who would be willing to develop for me a working circuit for my Idea, of course against a reasonable payment for your working hours, and you get the credit for developping the circuit ?

Also I would like to learn together with you.

You could buy the necessary parts and I pay them via PayPal would be the easiest way. Shipping them from Mexico to USA is nonsense. Shipping would cost more then the parts cost ... On the other side the possibilities to get the appropiate parts in USA is by far much higher

I am open to hear offers.

Thanks in advance ...

rsphoto said:

OK, KeepItSimpleStupid tells me I need a TL07x or TL08x and now you tell me it is useless message #22

Click to expand...

In post #22 https://www.electro-tech-online.com...cilloscope-experts.145961/page-2#post-1235029 Nigel Goodwin, not me, recommended the TL082

BTW: KISS works as an alternate name.

In fact, last night I briefly looked for a better OP amp.

Where does the 12 V DC power come to play? Is that available?

KeepItSimpleStupid said:

In post #22 https://www.electro-tech-online.com...cilloscope-experts.145961/page-2#post-1235029 Nigel Goodwin, not me, recommended the TL082

BTW: KISS works as an alternate name.

In fact, last night I briefly looked for a better OP amp.

Where does the 12 V DC power come to play? Is that available?

Click to expand...

OH, KISS I am very sorry and lots of apologies. I am already confused with so many messages, Again Sorry

The 12V DC at the moment comes from a regulated power supply but has NO -12V output signal. As said I am thinking of a portable device but easily could then use 2 LiPo batteries wired in that way that I get +12V, -12V and Ground, But I prefer to make it KISS and use only +12V. If that is a problem here well then I would take the bullet and use 2 small LiPo batteries as the power consumption here is ridiculous I guess and they are rechargeable.

Hi KISS,

I have been searching the net for amplifiers and found the OPA XXXX Op Amps which are not cheap

I am also looking at the LT1xxx from Linear technology ...

The TL08x is a "general purpose" opamp. The TL07x is a TL08x selected for low noise and a lower input offset voltage then it has been used in millions of audio products.
The datasheet for the LM3914 says and shows that it has an input that is protected from damaging positive and negative voltages as high as 35V but ignores the negative and displays the positive voltages.

Now that we hear that you need extremely high voltage gain I mentioned that the circuit will also amplify the input offset voltage of the opamp. But offset voltage is DC and the vibration you want to measure is AC so a simple coupling capacitor will block the DC offset voltage but pass the AC vibration signal. Then since you are amplifying AC and not DC you can use an audio low noise opamp that works from a low supply voltage and have it biased at half the supply voltage so its inputs and output do not need to work down to 0V. I will select a low noise opamp that uses a low supply voltage and low supply current.

But vibration is an AC signal and has positive AND negative peaks. The LM3914 will display only positive peaks. So why not make the full-wave average or peak detector circuit shown in the datasheet of the LM3915, then the LM3914 will show the levels of positive or negative peaks. I have a full-wave detector circuit that works from a single low voltage supply.

Now I am thinking about cancelling the nasty vibration, but that will be your next project.

Hi,

Can I use the LM358 instead of the LF353 ?

Now I am thinking about cancelling the nasty vibration, but that will be your next project.

Click to expand...

What do you mean by nasty vibration or did you mean noise ?

The reason why I asked is:
1) There's a fair amount of <=5 V stuff developed for the automotive environment and sensor interfacing.
2) There is an IC called a "rail splitter" which "develops" an artificial ground at 1/2 the supply voltage when you need a little bit of +- power,
3) Your right that the power consumption is high.

Aside: While we're at it, it's IMPORTANT to have bypass capacitors. These are USUALLY small ceramic caps located really close to the power supply pins.
Recommendations can sometime be found in the datasheets. There is a whole science around bypassing and sometimes different types of capacitors are paralleled.

ag said:

So why not make the full-wave average or peak detector circuit shown in the datasheet of the LM3915, then the LM3914 will show the levels of positive or negative peaks.

Click to expand...

Why not a precision rectifier to preserve the peaks?

Something like this and if I apply 12V I would get +6V, GND, -6V ¿ correct ?

Like this ?

https://www.ti.com/lit/ug/tidu030/tidu030.pdf

Rail Splitter: More like these: https://www.ti.com/sitesearch/docs/universalsearch.tsp?searchTerm=rail splitter#linkId=1&src=top

I just realized that your piezo sensor does not show how much vibration, instead it shows that anything that starts it moving causes it to ring like a bell for a while at its resonant frequency of 75Hz.
Its datasheet says its sensitivity is significantly higher at its resonant frequency and says that operation should be limited by your circuit to below 40Hz. That is why your 'scope showed an oscillation, the sensor was resonating each time you touched it.
Then your circuit needs a lowpass filter of 2nd or 3rd order with its cutoff frequency at 40Hz or 42Hz.

Since it resonates, a full-wave detector is useless and a half-wave detector is fine. A half-wave detector can use an opamp that does not need a negative supply.

Here is what you need:
1) A low noise AC preamp with a high input impedance. It does not need a negative supply or a rail splitter. Maybe an OPAx134. Its minimum supply is 5V so a 9V battery is fine.
2) An opamp maybe the same as the preamp opamp for a lowpass filter. An OPA2134 has two opamps in it or an OPA4134 has 4 opamps in it.
3) An opamp that works down to 0V as a half-wave detector. Use an LM358 or an LM324 that has 4 of the same opamps in it.
4) An LM3914 to drive the LED display.

You asked about an LF353. It is a dual opamp with a high input impedance and it needs a dual polarity supply or a rail splitter. Its minimum supply is 8V or 9V and its noise level and input offset voltage are high. It is almost the same as a TL082.