Circuit for measuring the peak of sinewave in range from 0 to 5 v with high accuracy

[mohmd10]

I need an electronic circuits to measure the peak of any sine wave in range from 0
to 5 v with high accuracy

thank you

 

[pauloynski]

how to design a peak detector

It depends on the accuracy you need but a good starting point is **broken link removed**
Regards

 

[mohmd10]

Re: peak detector design

It depends on the accuracy you need but a good starting point is **broken link removed**
Regards

thank you alot for your reply ,
i tried this circuit but when i entered input as example 3v it measures it as 2.9 and it takes alot of time to settle .
my application has a sine wave input at 100khz and i need senstivity as 5 mv

can help me

 

[pauloynski]

Re: peak detector design

In this circuit the input offset voltage of IC2 is multiplied by 10 (or R2/R1). This may be the cause of the error you´ve reported. If this is right then if you measure using different input voltages they will show the same error (about 0.1V). If this is the case I would like suggesting your using a low offset voltage op-amp in replacement of the (old) LM358 or using a potentiometer to null the offset. Also, to achieve faster results, try to decrease the value of C1 (1uF).At high frequencies it can be very small.
Furthermore, if you´re using 5V supplies I recomend you´re replacing the LM358 by a rail-to-rail output op-amp, since the output voltage span of the LM358 is very limited (please see it´s manual). 3V is it´s limit and this can be the cause of the error you´ve seen. So, make some tests at a lower input voltage to confirm.
You may also find some interesting tips in AppNotes AN1097 and AN1309.0 from Intersil.
Regards

 

[mohmd10]

Re: peak detector design

thank you alot

I used opamp TL070 instead of LM358 and i decreased C1 to .5uf but it still inaccuracy.i need 5mv senstivity .can you help me?

 

[pauloynski]

Re: peak detector design

If you´re using 5V supplies please remember that the (old) TL072 is not a rail-to-rail output op-amp. Also, make some measurements to confirm you´re not having offset problems. Which type of comparator are you using at the input? Is the error the same for all inputs or does it change proportionally? You can further reduce the value of C1.

 

[mohmd10]

Re: peak detector design

thank you.
I put var resistor to solve the problem of voltage offset .The comparator is TLC372
.The error is changed randomly and it is about .3 v and i want senstivity 5mv

 

[pauloynski]

Re: peak detector design

What is the suplly voltage? Are you using a split or single voltage?

 

[mohmd10]

Re: peak detector design

my supply is 5 v and single

 

[pauloynski]

Re: peak detector design

Please read carefully the specs of the LF358 and TL072. These are NOT rail-to-rail output op-amps. In fact, 5V is out of the specs of the TL and probably you won´t get nothing at its output. The LF358 can swing to 0V but the TL072 can´t. Furthermore, both can´t swing to 5V.
Regards

 

[Petre Petrov]

Re: peak detector design

".... has a sine wave input at 100khz and i need senstivity as 5 mv ...
"... TL070 instead of LM358 ..."


Hi
You need some much more faster than LM358 and TL07X for good accuracy at 100 kHz.
If you make several simple calculations you will have the results.
BR
Petre Petrov

 

[pauloynski]

Re: peak detector design

Dear Petre: even the LM358 is able to operate at 100kHz with low gain. The circircuit really needs a faster stage at the input: the comparator.

 

[Petre Petrov]

Re: peak detector design

Dear pauloynski,
Thanks for the answer.
You are right.
The gain i really low (approx 10).
But the results are not good from my point of view.
e.g.
**broken link removed**
If you are satisfied - that OK for you.
BR
Petre Petrov

 

[pauloynski]

Re: peak detector design

Dear Petre: I agree with you. But the LM358 should not be so bad as the results are showing. I believe that it should work well for amplitudes below 3Vpp when using a single suplly voltage of 5V. The TL072 is out of question: it can´t work with 5V. To achieve full voltage span a rail-to-rail output op-amp should be used. Also, it is not clear how the input voltage is apllied to the input of the comparator. The input voltage can´t go negative.

 

[FvM]

peak detector design

When reviewing the edn article, a light shiver crossed my back, thinking how the comparator would behave in a feedback loop. But the presented error plot shows how...

I also noticed, that you neither specified attack nor decay time for the peak detector, respectively a frequency range with related accuracy.

 

[mohmd10]

Re: peak detector design

thank you alot for the replies.
Now what i should do ? i tried more with my circuit .it works well but i didnot achieve my gool (peak detector with 5mv senstivity).my application is RC meter and i used peak detector as a small (but its very important) step .can you help me?

 

[Petre Petrov]

Re: peak detector design

Hello!

mohmd10 :
“I need an electronic circuits:
* to measure the peak
* of any sine wave
* in range from 0 to 5 v with high accuracy”

Is that means:
1. To detect the positive peak?
2. Frequency from approx 0 to 100 kHz?
3. Tested at 5000mV p-p?
4. Accuracy in amplitude better than +-0.1%?
5. Temperature +25+-5C?
6. Sinusoidal signal not a “sine wave”?
Etc..
If you mean that I am not sure that you have it easy with that circuit.
Good luck!
BR
Petre Petrov

 

[mohmd10]

Re: peak detector design

thank you alot Petre Petrov

1. To detect the positive peak?
yes,i need to detect postive peak.
2. Frequency from approx 0 to 100 kHz?
frequency is fixed 100khz but i can change it to 1khz .
3. Tested at 5000mV p-p?
yes,from 0 to 5 v
4. Accuracy in amplitude better than +-0.1%?
if it be .05% it will be excellant
5. Temperature +25+-5C?
yes
6. Sinusoidal signal not a “sine wave”?
I need for sine wave only

 

[Petre Petrov]

Re: peak detector design

Hi mohmd10
May be that will help you:

Please note that you have to develop clear requirements and test procedures.

Several notes, questions and proposals:
1. Fix at the lowest possible frequency, e.g. 1000Hz
2. Fix the test levels, e.g. 5000mV.
3. You cannot have the same error at 5.00mV, 50.0mV, 500.0mV, 2500 mV and 5000mV.
4. Fix the error levels at each test level.
5. Forget accuracy better that approx +-1% at 2500mV with analogous circuit. (There are exceptions but I do not think you have access to them).
6.What about an ADC solution?
7. If you choose ADC you could use:
https://www.ieindia.org/pdf/88/88ET104.pdf

Good luck
BR
Petre Petrov