High Frequency Full Wave Rectifier

Thread Starter

synws93

Joined Nov 18, 2016
25
<Moderator's note: This post was moved from http://forum.allaboutcircuits.com/threads/high-frequency-full-wave-rectifier.116931/#post-1062554 to create a new thread. It is not homework.>

Somewhere on the web there's a neat precision full wave rectifier that uses 2 single supply opamps and no diodes. Designed by a couple Spanish students IIRC...

I'd post a schematic, but this is homework.:rolleyes:
Hi @dl324 ,

I've found out the schematic that you have been talking about. Is it possible to use MCP6292 for this application? Sorry I'm quite new to electronics, and I have to do a project with precision full wave rectifier.
MCP629x specification sheet: http://ww1.microchip.com/downloads/en/DeviceDoc/21812e.pdf

This is the schematic that I've found out.



Thanks in advance.
 
Last edited by a moderator:

crutschow

Joined Mar 14, 2008
34,470
Here's the LTspice simulation of a simplified version of the no-diode precision rectifier.
I used a different op amp but it should also work with the Microchip device.

upload_2016-11-20_0-49-27.png
 

Thread Starter

synws93

Joined Nov 18, 2016
25
What are the performance requirements for your circuit?
In my project, the input of the signal will be around 400 - 500kHz. And I want to rectify my signal to a precision full wave.

Your post title says "high frequency".
What is high?
It will be around 400 - 500 kHz signal.

Sorry for the late replies, as I was not aware that my post was moved to a new thread.
 

dannyf

Joined Sep 13, 2015
2,197
you are looking for small signal germanium or shottky diodes. There are tons of them and they are very inexpensive.

If you cannot find any, use small signal transistors. for your frequency range, 2n2222 would be more than enough. or 2n390x for higher frequency, or s901x into 100Mhz or over. or jfets.

choices are literally endless.
 

Thread Starter

synws93

Joined Nov 18, 2016
25
Here's the LTspice simulation of a simplified version of the no-diode precision rectifier.
I used a different op amp but it should also work with the Microchip device.

View attachment 115638
you are looking for small signal germanium or shottky diodes. There are tons of them and they are very inexpensive.

If you cannot find any, use small signal transistors. for your frequency range, 2n2222 would be more than enough. or 2n390x for higher frequency, or s901x into 100Mhz or over. or jfets.

choices are literally endless.
Let's say I'm using crutschow's circuit, where do I place these germanium diodes?

Thanks in advance. PS: I'm quite new to electronics.
 

bertus

Joined Apr 5, 2008
22,278
Hello,

The circuit of crutschow does not use diodes.
There are two opamps. U2 determines if the signal is positive or negative,
U3 amplifies the signal by +1 or -1 depending on the signal from U2.

@crutschow , please correct me if I amwrong.

Bertus
 

Thread Starter

synws93

Joined Nov 18, 2016
25
Hello,

The circuit of crutschow does not use diodes.
There are two opamps. U2 determines if the signal is positive or negative,
U3 amplifies the signal by +1 or -1 depending on the signal from U2.

@crutschow , please correct me if I amwrong.

Bertus
Hi Bertus, yes I know that crutschow's circuit does not use diodes.

But dannyf suggested that if I don't mind non-linearity, I can put in germanium diodes. I'm just not sure where can I put the suggested diodes into my circuit.
 

crutschow

Joined Mar 14, 2008
34,470
dannyf suggested that if I don't mind non-linearity, I can put in germanium diodes. I'm just not sure where can I put the suggested diodes into my circuit.
If you use diodes and don't mind some non-linearity, then you don't need the op amps.
Note that a diode will only give half-wave rectification, not the full-wave of the op amp circuit.
 

Thread Starter

synws93

Joined Nov 18, 2016
25
If you use diodes and don't mind some non-linearity, then you don't need the op amps.
Note that a diode will only give half-wave rectification, not the full-wave of the op amp circuit.
Thanks for sharing this information.

Means if i need linearity in my signal, op amps are needed.

If I need non-linearity in my signal, I can basically use diodes for my precision full wave rectifier.
 

crutschow

Joined Mar 14, 2008
34,470
There are two opamps. U2 determines if the signal is positive or negative,
U3 amplifies the signal by +1 or -1 depending on the signal from U2.
Basically yes.
The circuit requires a single-supply opamp with its negative power supply being ground, and thus U2 simply is unable to generate a negative output.
Thus for the positive polarity input U2 and U3 amplify the signal, both as followers.
For negative input, U2's output must stay at zero and U3 then amplifies the signal as an inverter.
 

dannyf

Joined Sep 13, 2015
2,197
Let's say I'm using crutschow's circuit, where do I place these germanium diodes?
I don't see every post so cannot help you there.

I think if all you want is a full-wave rectifier, a few diodes / transistors will do just fine, as I laid out earlier.
 

Thread Starter

synws93

Joined Nov 18, 2016
25
Here is a similar circuit with a easier-to-follow description.
http://electronicdesign.com/analog/rectifier-has-no-diodes

A little more information about your application might trigger some better solutions than a full wave rectifier. What is your application?
Hi DickCappels, I tried to use your circuit. I have input 500kHz sine wave and I got back a 500kHz sine wave too but not rectified. Not sure where is the problem. I've attached to you the images of my set-up.



 
Top