How to convert sinusoidal signal other than RMS converter or rectifier techniques?

[syamin0712]

Hi,

I have my system that consists of sinusoidal signal and I would like to use it for ADC part.
Thus, how can I convert the signal so that I can use it for ADC part?
What I know is by converting the sinusoidal into full wave rectifier or by using RMS-to-DC converter, is there any other techniques that I can applied so that I can use it for ADC purpose.

regards,
Yasmin

 

[KlausST]

Hi,

You problem is not clear, it needs a more detailed description.

Specify: amplitude, frequency.
Where does the signal come from?
What do you want to do with the signal?
Why can't you use the sinusoidal signal?
Why can't you use the RMS signal?

A schematic could help also. (Sketch, photo...)

Klaus

 

[syamin0712]

Hi Klaus,

My applications is actually for electrical tomography with 16 sensors.
I give the 10 Vp-p, with operating frequency is 2 MHz sinusoidal waveform at transmitter sensor by using PIC18F4580 to trigger the DDS module,AD9833 and amplify using signal generator circuit,LT1226.
Then, the signal that received at the receiver is need to do the I-V converter by using LT1360 and other process before I can use the signal in the MATLAB software to get some image.
Also,the output from LT1360 is around 1Vpp maximum.

Actually, I'm trying to find the best method to convert the signal so that I can use it for software part. Maybe the term of 'ADC' that I used was wrong. It was actually to convert the sinusoidal signal (analog signal) into dc signal.
What I know is I have to convert the sinusoidal signal to make the signal becomes positive value, so that in my case Im using microcontroller ; it can do the sampling much easier.

I have tried the full wave rectifier but because the output that I got is not symmetry due to the high frequency, i want to try other method.
Now, Im trying to use the rms-to-dc converter by implementing LTC1968IMS8. I got the simulation output from Multisim that is the Vdc of LTC1968 is equal to the VRMS value of my I-V output.
But maybe because Im just trying the circuit on the brade board, it contains a lot of noise and I need to trouble shoot it.

Thus, I just want to know if there is any other methods that we can implement to convert the sinusoidal signal so that it becomes much more simpler for the microcontroller to sample the signal.

- - - Updated - - -

Also,

For your information, the signal that sampled by the microcontroller will be used in the matlab software to get some image.

 

[FvM]

RMS and rectified value are different parameters, so the first question would be which parameter you actually want to measure. There's a fixed relation between these parameters for a known waveform, e.g. a sine, but can you be sure that the input signal is always a pure sine?

Each parameter can be either determined by an analog signal converter circuit, or by digital signal processing means if you are sampling the input signal with sufficient high rate.

- - - Updated - - -

After your latest clarification. 2 MHz isn't well suited for digital signal processing. There's also no advantage of a true RMS converter for small band signals. An active rectifier circuit made with sufficient fast OPs (at least 100 MHz bandwidth) should work well. Also a simple schottky diode rectifier might be O.K.

 

[KlausST]

Hi,

a tomography....
I doubt the signal is a pure sinusoidal signal. I assume it is more a pulsed sinusiodal signal.
Can you show us a scope picture of the waveform?

I assume you want to measure the amplitude of the receiving signal. And i also assume the amplitude changes with time, wihtin milliseconds.
If so, then you need to do high speed signal sampling of the 2MHz signal. I assume at least 20 or 50 MSmpl/s, store the data in an SRAM and process it after the measurement, because the PIC may be too slow to do this in real time.

Klaus

 

[syamin0712]

After your latest clarification. 2 MHz isn't well suited for digital signal processing. There's also no advantage of a true RMS converter for small band signals. An active rectifier circuit made with sufficient fast OPs (at least 100 MHz bandwidth) should work well. Also a simple schottky diode rectifier might be O.K.

I had post my thread regarding this rectifier before this as the link below:

https://www.edaboard.com/threads/342365/


Should I change into Lt1226 for example?

Can you show us a scope picture of the waveform?

See the attachment...As what I got, i believed that te signal is still a sine waveform.
Does means that I can't apply the RMS-to-DC converter?
https://obrazki.elektroda.pl/1125920100_1448358141.jpg

If so, then you need to do high speed signal sampling of the 2MHz signal. I assume at least 20 or 50 MSmpl/s, store the data in an SRAM and process it after the measurement, because the PIC may be too slow to do this in real time.

Ok, I will noted it

- - - Updated - - -

yellow line was the signal from signal generator

blue line was the signal from LT1360

- - - Updated - - -

I had post my thread regarding this rectifier before this as the link below:

absolute value circuits for high frequency (2MHz)


Should I change into Lt1226 for example?

From the discussion I have changed the schottky diodes to BAT81 and the output it much better.
https://obrazki.elektroda.pl/6127508300_1448358818.png

Before this, for testing...this rectifier circuit I've done by using the transmitter signal direct fromfunction generator. then, you can just ignore the yellow line of the signal

But, I think it quite difficult to sample the signal if compared to the RMS value.
is it correct?

 

[KlausST]

Hi,

do you know i good RMS-to-DC converter for 2MHz signal?
I do not, but maybe there is one.

But still it depends on your input waveform.
Your scope picture shows a sinal generator as source.
Is the continous signal from your signal generator really comparable to your real "received" waveform?
I doubt this.

Do you have acces to the 2MHz source? (I don´t mean the "received" signal).
If yes, you could use a lock_in_amplifier.

Is there a variable phase shift between source and reveived signal?
If yes, are you interested in this phase shift?

Klaus

 

[syamin0712]

Hi ,

1) I found only one ic RMS-to-DC converter for high freq purpose at element14 website. The ic is LTC1968IMS8 with gain bandwith product 15MHz.
Thus, i decided to try it. Or this ic is not enough good for 2MHz system?

2) The shape of 'received' signal should be similar that is sinusoidal either I used direct source sinewaveform 2MHz from function generator or from my signal generator circuit.

3)To access of 2MHz sinewaveform; im was using DDS module ic that is AD9833 and amplify the signal using high speed and high gain op-amp,LT1226.

4)yes, there has a slight phase shift between the source and received signal. But, im only just concern on the amplitude of the received signal.

Thus, what is the best method to convert the sinusoidal signal in my case? Is it by using rectifier or rms or other method?

 

[KlausST]

Hi,

Did you look for "lock in amplifier"?

Klaus

 

[syamin0712]

Hi,

Not yet. What is 'lock in amplifier' and what does the function? Could you suggest example model of 'lock in amplifier'?