Author Topic: Fast(er) op-amp for a precision full-wave rectifier  (Read 5622 times)

0 Members and 1 Guest are viewing this topic.

Offline ltz2000

  • Regular Contributor
  • *
  • Posts: 103
Fast(er) op-amp for a precision full-wave rectifier
« on: April 02, 2015, 09:51:12 am »

I happened to have a large quantity surplus MC33079 op-amps which I thought would do the job. Very low offset voltage (but it is bipolar, so need to be careful with the impedance balancing). The prototype worked fine down to a few mV.

Except the frequency response. Perfect up to a few kHz, but I measured -4dB at 20 kHz for low level signals.

The resistance values are as low as they can be and the diodes (1N4148) can't be the problem. So I need a faster op-amp.


Last time the solution was an LM301 (?) with a special compensation scheme making it faster. But that was more than 30 years ago.

I know that there must be a number of suitable op-amp available and new products introduced every year.

But what would be the "industry standard fast op-amp"? Low cost, available almost everywhere.


 

Offline dannyf

  • Super Contributor
  • ***
  • Posts: 8229
  • Country: 00
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #1 on: April 02, 2015, 10:49:41 am »
Quote
But what would be the "industry standard fast op-amp"? Low cost, available almost everywhere.

NE5532, good for a few Khz. Not exactly your circuit but works just as well.
================================
https://dannyelectronics.wordpress.com/
 

Offline ltz2000

  • Regular Contributor
  • *
  • Posts: 103
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #2 on: April 02, 2015, 11:23:55 am »
Quote
But what would be the "industry standard fast op-amp"? Low cost, available almost everywhere.

NE5532, good for a few Khz. Not exactly your circuit but works just as well.

The 5532 is actually worse than the 33079 that I used. No improvement in speed, but higher offset and known to be unstable in non-linear circuits.
 

Offline blackdog

  • Frequent Contributor
  • **
  • Posts: 583
  • Country: nl
  • Please stop pushing bullshit...
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #3 on: April 02, 2015, 12:21:06 pm »
Hi LTZ2000,

This....
But what would be the "industry standard fast op-amp"? Low cost, available almost everywhere.
Is looking for the holy grail...  :-DD

Faster opamp you can buy, are all relatively bad DC performers.
You can use somthinge like the old LM318 and use some feedforeward, but, not so good DC performance.

Why not use a AD736 or a LTC1968, but be shure to use the right level (dynamic range)

To make your schematic better, use low power schottky diode's for a lower voltage drop and more speed.
Keep the resistance value is relatively low, use bias compensation resistors on the + inputs of the opamp.

The are there are plenty of fast opamps available from Analog, Linear and TI, but no yelly beans...


Kind regarts,
Blackdog
“Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe.”
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 13200
  • Country: gb
  • 0999
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #4 on: April 02, 2015, 12:31:13 pm »
Try the circuit on page 17.

http://www.ing.unlp.edu.ar/electrotecnia/circelec/CA3140.pdf

It has to drive a 10k load and has a gain of 0.5 but a buffer or gain stage can be added if needs be.
 

Offline blackdog

  • Frequent Contributor
  • **
  • Posts: 583
  • Country: nl
  • Please stop pushing bullshit...
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #5 on: April 02, 2015, 12:40:46 pm »
Hi Hero999,


The problem whit this schematic is, try 10mv, is wil not be -3dB @ 290Khz :-)
A fast schottky diode can help...

Kind regarts,
Blackdog
“Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe.”
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 13200
  • Country: gb
  • 0999
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #6 on: April 02, 2015, 12:47:56 pm »
Hi Hero999,


The problem whit this schematic is, try 10mv, is wil not be -3dB @ 290Khz :-)
A fast schottky diode can help...

Kind regarts,
Blackdog
What do you mean? I don't understand.
 

Offline blackdog

  • Frequent Contributor
  • **
  • Posts: 583
  • Country: nl
  • Please stop pushing bullshit...
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #7 on: April 02, 2015, 01:05:35 pm »
Hi Hero999,


With low input voltages, the opamp must work verry hard to overcome the diode threshold.
Thats why you need a fast opamp, and low threshold diodes(schottky) helps.

Kind regarts,
Blackodg
“Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe.”
 

Offline ltz2000

  • Regular Contributor
  • *
  • Posts: 103
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #8 on: April 02, 2015, 01:10:08 pm »
use low power schottky diode's for a lower voltage drop and more speed.

Makes the "dead zone" smaller. I thought that the difference would be minimal, but definitely worth trying. Any recommendations for a holy grail small signal fast schottky diode?

Keep the resistance value relatively low

Already done, as mentioned in the first post.

use bias compensation resistors on the + inputs of the opamp.

Already done, as mentioned in the first post.
 

Offline dannyf

  • Super Contributor
  • ***
  • Posts: 8229
  • Country: 00
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #9 on: April 02, 2015, 01:37:29 pm »
Only slew rate matters here. So you zero bias diodes or germanium diodes would help.

Also, a current sensor is a uni-direction rectifier. With two back to back current sensors, you have a rectifier, especially good for small signals.
================================
https://dannyelectronics.wordpress.com/
 

Offline dannyf

  • Super Contributor
  • ***
  • Posts: 8229
  • Country: 00
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #10 on: April 02, 2015, 01:38:13 pm »
Also think of current transmission vs. voltage transmission. The ne5532 rectifier I built went all the way to 120Khz.
================================
https://dannyelectronics.wordpress.com/
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 13200
  • Country: gb
  • 0999
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #11 on: April 02, 2015, 02:12:13 pm »
Hi Hero999,


With low input voltages, the opamp must work verry hard to overcome the diode threshold.
Thats why you need a fast opamp, and low threshold diodes(schottky) helps.

Kind regarts,
Blackodg
That's true, but that circuit is still easier on the op--amp than other circuits. Try changing the diode for a Schottky.

Another thing you could do is amplify the signal before the rectifier but that will limit the maximum input voltage.
 

Offline Liv

  • Regular Contributor
  • *
  • Posts: 133
  • Country: by
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #12 on: April 02, 2015, 04:08:06 pm »
I used the AD8512 opamp and Schottky diode BAS70-04 for detector in frequency response meter up to 500 kHz.
PSU PSL-3604 Pulse gen. PG-872 Freq. cnt. FC-510
 

Offline ltz2000

  • Regular Contributor
  • *
  • Posts: 103
Re: Fast(er) op-amp for a precision full-wave rectifier
« Reply #13 on: April 02, 2015, 08:06:02 pm »

use low power schottky diode's for a lower voltage drop and more speed.

Makes the "dead zone" smaller. I thought that the difference would be minimal, but definitely worth trying.

Tested at 5mV signal level:

1N4148: -4dB at 20 kHz

1N5711 (Schottky): -3dB at 20 kHz

1 dB improvement. Still need the faster op-amp...
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf