Lookinf for options for an Absolute Magnitude" or precision rectifier circuit.

[Clarence_A]

Must work over substantial range, say .01 to .6 Volts RMS.

Serious responses only. Please.

I have a working circuit now, but there is some variation between
positive and negative half cycles which seems to be inherent.
Must be some easy way to eliminate this. Yes?

 

[Jim Thompson]

Must work over substantial range, say .01 to .6 Volts RMS.

Serious responses only. Please.

I have a working circuit now, but there is some variation between
positive and negative half cycles which seems to be inherent.
Must be some easy way to eliminate this. Yes?

Frequency?

...Jim Thompson

 

[Jim Thompson]

Must work over substantial range, say .01 to .6 Volts RMS.

Serious responses only. Please.

I have a working circuit now, but there is some variation between
positive and negative half cycles which seems to be inherent.
Must be some easy way to eliminate this. Yes?

See.....

Newsgroups: alt.binaries.schematics.electronic
Subject: S.E.D: Looking for Options for an "Absolute Magnitude" or
Precision Rectifier Circuit - FullWaveRectifier.pdf
Message-ID: <[email protected]>

...Jim Thompson

 

[Clarence_A]

Frequency?

380 to 800 HZ

 

[Clarence_A]

See.....

Newsgroups: alt.binaries.schematics.electronic
Subject: S.E.D: Looking for Options for an "Absolute Magnitude" or
Precision Rectifier Circuit - FullWaveRectifier.pdf
Message-ID: <[email protected]>

...Jim Thompson

Thanks Jim. I'll see if it fixes it.
The circuit I used is the one in the SO291 Specification as a
suggested application. Works, except as noted.

 

[Clarence_A]

Magnitude"


Thanks Jim. I'll see if it fixes it.
The circuit I used is the one in the SO291 Specification as a
suggested application. Works, except as noted.

Opps! OP291

 

[John Larkin]

Must work over substantial range, say .01 to .6 Volts RMS.

Serious responses only. Please.

I did a mechanical version once. No matter which direction you turned
the input shaft, the output shaft went clockwise. The MEs were stumped
so I showed them how to do a rotary full-wave rectifier.

I have a working circuit now, but there is some variation between
positive and negative half cycles which seems to be inherent.
Must be some easy way to eliminate this. Yes?

Can you post your circuit?

John

 

[Clarence_A]

"John Larkin" <[email protected]>
wrote in message

I did a mechanical version once. No matter which direction you turned
the input shaft, the output shaft went clockwise. The MEs were stumped
so I showed them how to do a rotary full-wave rectifier.


Can you post your circuit?

Sure, But it's easier to post a link!

http://www.analog.com/UploadedFiles/Data_Sheets/813995969OP191_291_491_c.pdf

On page seventeen!

 

[John Woodgate]

I read in sci.electronics.design that John Larkin <jjSNIPlarkin@highTHIS
landPLEASEtechnology.XXX> wrote (in <ltea015gvuthb414lq6kqc1kv806prgjef@
4ax.com>) about 'Lookinf for options for an Absolute Magnitude" or
precision rectifier circuit.', on Sat, 5 Feb 2005:

I did a mechanical version once. No matter which direction you turned
the input shaft, the output shaft went clockwise. The MEs were stumped
so I showed them how to do a rotary full-wave rectifier.

With ratchet wheels or the spring-loaded orbital idler (which looks like
magic to me)?

 

[John Larkin]

"John Larkin" <[email protected]>
wrote in message

Sure, But it's easier to post a link!

http://www.analog.com/UploadedFiles/Data_Sheets/813995969OP191_291_491_c.pdf

On page seventeen!

Crazy! It pulls the opamp input below ground, in violation of their
own abs-max ratings! That 2 volt p-p input must be on the very edge of
blowing the opamp up.

John

 

[John Larkin]

I read in sci.electronics.design that John Larkin <jjSNIPlarkin@highTHIS
landPLEASEtechnology.XXX> wrote (in <ltea015gvuthb414lq6kqc1kv806prgjef@
4ax.com>) about 'Lookinf for options for an Absolute Magnitude" or
precision rectifier circuit.', on Sat, 5 Feb 2005:


With ratchet wheels or the spring-loaded orbital idler (which looks like
magic to me)?

Planetaries and orbitals and differentials are too complicated for me,
too. I used notched pulleys/belts to get counter-rotations and a pair
of one-way roller clutches to rectify them into the output shaft. It
actually drove a synchro for a shaft-rotaton counter on a ship. It was
really silly, since ships spend so little time going backwards you'd
never really miss those turns.

John

 

[Clarence_A]

"John Larkin" <[email protected]>
wrote in message

"John Larkin" <[email protected]>
wrote in message

Crazy! It pulls the opamp input below ground, in violation of their
own abs-max ratings! That 2 volt p-p input must be on the very edge of
blowing the opamp up.

So you see why I wanted a replacement!

 

[John Woodgate]

I read in sci.electronics.design that John Larkin <jjSNIPlarkin@highTHIS
landPLEASEtechnology.XXX> wrote (in <f1ja01d165rcirkpdo1o77q4mr793v2iic@
4ax.com>) about 'Lookinf for options for an Absolute Magnitude" or
precision rectifier circuit.', on Sat, 5 Feb 2005:

Planetaries and orbitals and differentials are too complicated for me,
too. I used notched pulleys/belts to get counter-rotations and a pair
of one-way roller clutches to rectify them into the output shaft. It
actually drove a synchro for a shaft-rotaton counter on a ship. It was
really silly, since ships spend so little time going backwards you'd
never really miss those turns.

Your roller clutches are equivalent to ratchet wheels, and diodes. I saw
the orbital idler solution in a Hong Kong clockwork car. It reversed the
car whenever it wouldn't go further, so it was actually an inverter, but
used the other way round, it's a rectifier. Unfortunately, the kid
wouldn't let me steal it to see in detail how it works. (;-)

 

[Glen Walpert]

"John Larkin" <[email protected]>
wrote in message


Crazy! It pulls the opamp input below ground, in violation of their
own abs-max ratings! That 2 volt p-p input must be on the very edge of
blowing the opamp up.

John

Their abs-max input voltage rating of "GND to Vs 10V" is not clear to
me; perhaps they mean you can go outside the PS rails by 10V max, or
perhaps they mean the input is limited to 10V max even if the power
supply is at the specified 16V max? Either way seems more than a
little bit odd to me, if I were using this part I would call my AD FAE
for an interpretation. Data sheet typos are not unheard of.

I would however recommend the more common precision rectifier circuit
using diodes in the feedback loop of an op-amp if better performance
is required. AOE discusses the half wave version in some detail, full
wave is left as an exercise for the student.

Glen

 

[Jim Thompson]

"John Larkin" <[email protected]>
wrote in message

Sure, But it's easier to post a link!

http://www.analog.com/UploadedFiles/Data_Sheets/813995969OP191_291_491_c.pdf

On page seventeen!

What a crock! Relies on rail-to-rail OpAmp with negative rail of
first stage tied to ground :-(

...Jim Thompson

 

[Fred Bloggs]

Crazy! It pulls the opamp input below ground, in violation of their
own abs-max ratings! That 2 volt p-p input must be on the very edge of
blowing the opamp up.

In addition to knowing nothing about the 291, you can't read- the AM
Rating was obviously intent on saying GND+/-10V, this design is not only
ALL ABOUT input overdrive but also overdrive w/o phase inversion. There
are 5K resistors in series with each input, diode clamps to V+ from
there, and antiparallel diodes shunting the input PNP-NPN differential
pairs. These resistors are designed to handle 5mA continuous current.
The OP said:"Must work over substantial range, say .01 to .6 Volts RMS"
and "but there is some variation between positive and negative half
cycles" -DUHHUHHHH. Looks like Vout=Vbe-Vbc of the saturated Q33
pulldown- the Voh and Vol specs on Vce for the output transistors are
something like 5mV at 50uA loading- but with loading to voltage below
GND the pulldown saturation is inverted. Then 0.6RMS is something on the
order of 1V peak negative, making for (1V-0.6)/10K=40uA max negative
loading, so that a 100K pullup from the input follower output to Vs+
(=5V?) should keep the Q33 saturation from inversion and Vout~5mV or
less during negative cycles- but the output follower will amplify this
by x2 for +10mV output offset for the input negative half cycle- and
this is comparable to the low end of the signal amplitude range. I
suppose you could try something like this:

Never mind- will not waste time on that individual...

 

[Fred Bloggs]

What a crock! Relies on rail-to-rail OpAmp with negative rail of
first stage tied to ground :-(

Heheh- the reference level about which rectification is performed is
defined by the clamp level of the first amplifier buffer. Ok- make that
"rectalication" for the alien fascist, Clarence-

 

[Glen Walpert]

"John Larkin" <[email protected]>
wrote in message

So you see why I wanted a replacement!

Take a look at figure 2.42(d) in the classic 1966 Philbrich
Applications Manual for Computing Amplifiers at:
<http://www.analog.com/library/analogDialogue/archives/philbrick/computing_amplifiers.html>
it is in Part 2 discontinuities, but I recomment reading the entire
book esp the discussions of opamp error sources.

Glen

 

[Clarence_A]

"Glen Walpert" wrote

"Clarence_A" wrote

Take a look at figure 2.42(d) in the classic 1966 Philbrich
Applications Manual for Computing Amplifiers at:
<http://www.analog.com/library/analogDialogue/archives/philbrick/c
omputing_amplifiers.html>
it is in Part 2 discontinuities, but I recommend reading the entire
book esp the discussions of opamp error sources.
Glen

This is the kind of authoritative references I sought.

Background:
I was contracted to do a design, so I was provided with the design
of a previous unit they wanted to move into the new application.

For various reasons I simply started to design a new circuit block
by block (as I modeled the damn thing) finding many problems with
the previous design as I progressed.

Since I need to change what the "app note paste artist" put
together, I need to refute his premise that "it wouldn't be
published if it wasn't correct."

I was prepared to do so, however more ammunition is always a good
strategy when dealing with 'entrenched' management.

Jim, Win, and Philbric, (TNX Glen) should be an army of expert
opinion to break down the walls!

Besides lives are at stake (Not just the job!) these units go on
commercial aircraft! Scary!

 

[richard mullens]

In addition to knowing nothing about the 291, you can't read- the AM
Rating was obviously intent on saying GND+/-10V, this design is not only
ALL ABOUT input overdrive but also overdrive w/o phase inversion. There
are 5K resistors in series with each input, diode clamps to V+ from
there, and antiparallel diodes shunting the input PNP-NPN differential
pairs. These resistors are designed to handle 5mA continuous current.
The OP said:"Must work over substantial range, say .01 to .6 Volts RMS"
and "but there is some variation between positive and negative half
cycles" -DUHHUHHHH. Looks like Vout=Vbe-Vbc of the saturated Q33
pulldown- the Voh and Vol specs on Vce for the output transistors are
something like 5mV at 50uA loading- but with loading to voltage below
GND the pulldown saturation is inverted. Then 0.6RMS is something on the
order of 1V peak negative, making for (1V-0.6)/10K=40uA max negative
loading, so that a 100K pullup from the input follower output to Vs+
(=5V?) should keep the Q33 saturation from inversion and Vout~5mV or
less during negative cycles- but the output follower will amplify this
by x2 for +10mV output offset for the input negative half cycle- and
this is comparable to the low end of the signal amplitude range. I
suppose you could try something like this:

Never mind- will not waste time on that individual...

It seems that Clarence_A may unwittingly be Osama's secret weapon.