When is a Butler not a Butler ?

[colin]

From the recent discusion I looked up butler oscillator to see exactly what
defines an oscillator as butler type,
but surprisingly wikipedia doesnt have it.

Ive always assumed it meant the tuning circuit was series resonant type is
this correct ?
or does it just mean the amplifier inputs and outputs are in phase ?

google shows up many references to butler type even a colpitz-butler
oscilator.

Colin =^.^=

 

[xray]

From the recent discusion I looked up butler oscillator to see exactly what
defines an oscillator as butler type,
but surprisingly wikipedia doesnt have it.

Ive always assumed it meant the tuning circuit was series resonant type is
this correct ?
or does it just mean the amplifier inputs and outputs are in phase ?

google shows up many references to butler type even a colpitz-butler
oscilator.

Colin =^.^=

As far as I know the Butler is only used with crystals and particularly
overtone crystals. It is essentially a Colpitts with a crystal added in
the emitter feedback. The LC of the Colpitts defines the overtone so
that the crystal operates at the correct overtone, but the high Q of the
crystal (operated in non-reactive series mode) determines the actual
operating frequency.

 

[Ken Smith]

From the recent discusion I looked up butler oscillator to see exactly what
defines an oscillator as butler type,
but surprisingly wikipedia doesnt have it.

Ive always assumed it meant the tuning circuit was series resonant type is
this correct ?
or does it just mean the amplifier inputs and outputs are in phase ?

I think the defining characteristic of the Butler circuit is that the
(series) tuned circuit is connected to the Cathode/Source/Emitter such
that it forces the amplifying device's current to be a sinewave.

 

[colin]

As far as I know the Butler is only used with crystals and particularly
overtone crystals. It is essentially a Colpitts with a crystal added in
the emitter feedback. The LC of the Colpitts defines the overtone so
that the crystal operates at the correct overtone, but the high Q of the
crystal (operated in non-reactive series mode) determines the actual
operating frequency.

Yes thats been my experience too, but I just heard it applied here to non
series resonant types too.

Although Ive found a 2 transistor butler circuit is prefereable to a coplits
at high overtones as you can use a much lower Q for the tank circuit so that
the tank does not affect the crystal frequency so much.

Colin =^.^=

 

[colin]

I think the defining characteristic of the Butler circuit is that the
(series) tuned circuit is connected to the Cathode/Source/Emitter such
that it forces the amplifying device's current to be a sinewave.

Yes that would make sense, as you need a low impedance input node with a
series tuned circuit.
Its sometimes between the emitters of a differential pair.

Colin =^.^=

 

[Jim Thompson]

As far as I know the Butler is only used with crystals and particularly
overtone crystals. It is essentially a Colpitts with a crystal added in
the emitter feedback.

That is correct. That's what I remembered from the last one I built,
back in the late '60's ;-) I Googled "Butler Oscillator" and found it
right away.

The LC of the Colpitts defines the overtone so
that the crystal operates at the correct overtone, but the high Q of the
crystal (operated in non-reactive series mode) determines the actual
operating frequency.

Yep. Quite nice and easy to connect to other circuits without
pulling.

...Jim Thompson

 

[Michael Black]

I think the defining characteristic of the Butler circuit is that the
(series) tuned circuit is connected to the Cathode/Source/Emitter such
that it forces the amplifying device's current to be a sinewave.

That's what comes to mind when I think of the Butler, but it seems
like every time someone came up with a new oscillator scheme (ie
a major change), someone else came along with a minor variant. Sometimes
it's almost like those old "spot the difference" challenges in children's
magazines, where the difference is pretty minor.

I don't know whether this is the case with the Butler, but with some
oscillators, two defining points are needed to differentiate it from
others.

Of course, older books tended to have a page or two of schematics for
the major types of oscillators, making it easier to discern the
differences, and people forget that people often put more effort
into books than people put into webpages.

Michael

 

[Barry Lennox]

From the recent discusion I looked up butler oscillator to see exactly what
defines an oscillator as butler type,
but surprisingly wikipedia doesnt have it.

"Butler" was a Frank Butler of the RAF/GCHQ who published his new
oscillator cct in "Wireless Engineer" ca 1944.

It was originally a cathode-coupled valve (tube) L-C oscillator and
his published circuit used a pair of 6V6's. The big attraction of it
was the simple parallel L-C circuit, so that freq band changes could
be made with a inductance change, no reaction windings or taps being
required.

However it's changed somewhat over the years, I first came across it
in the Collins 75 series receivers, it used an xtal in series feedback
between 2 halves of a 12AT7.

I think these days it applies to any oscillator that has a freq
sensitive element between an emitter follower and a grounded-base
amp.

Barry Lennox

 

[Andrew Holme]

I think the defining characteristic of the Butler circuit is that the
(series) tuned circuit is connected to the Cathode/Source/Emitter such
that it forces the amplifying device's current to be a sinewave.

Yes. The RSGB Radio Communication Handbook puts it like this: "the function
of amplitude limiting is performed by a second stage, leaving the function
of the first stage as an amplifier."

 

[Winfield Hill]

Andrew Holme wrote...

Ken Smith wrote ...

Yes. The RSGB Radio Communication Handbook puts it like this:
"the function of amplitude limiting is performed by a second
stage, leaving the function of the first stage as an amplifier."

Change of subject. I'd like to know why Analog Devices
dropped the elegant Butler JFET opamps from their lineup?

 

[Ken Smith]

Yes. The RSGB Radio Communication Handbook puts it like this: "the function
of amplitude limiting is performed by a second stage, leaving the function
of the first stage as an amplifier."

That would only apply if you aren't using AGC. They must be describing a
specific implementation of the Butler.

 

[Jim Thompson]

From the recent discusion I looked up butler oscillator to see exactly what
defines an oscillator as butler type,
but surprisingly wikipedia doesnt have it.

"Butler" was a Frank Butler of the RAF/GCHQ who published his new
oscillator cct in "Wireless Engineer" ca 1944.
[snip]

I think these days it applies to any oscillator that has a freq
sensitive element between an emitter follower and a grounded-base
amp.

Barry Lennox

I think it's been that definition since before the '60's when I ran
across it and used it extensively.

It's fits nicely into IC's... easy to control amplitude so you don't
saturate anything.

...Jim Thompson

 

[Jim Thompson]

That would only apply if you aren't using AGC. They must be describing a
specific implementation of the Butler.

You don't need AGC in a Butler, you can control amplitude with proper
selection of the values in the "harmonic" tank.

...Jim Thompson

 

[colin]

Andrew Holme wrote...

Change of subject. I'd like to know why Analog Devices
dropped the elegant Butler JFET opamps from their lineup?

The OP275 ? doesnt mention it on the info page, I wonder what the butler
part consists of as it doesnt show an internal schematic on the datasheet.

Colin =^.^=

 

[Bob Eld]

From the recent discusion I looked up butler oscillator to see exactly what
defines an oscillator as butler type,
but surprisingly wikipedia doesnt have it.

Ive always assumed it meant the tuning circuit was series resonant type is
this correct ?
or does it just mean the amplifier inputs and outputs are in phase ?

google shows up many references to butler type even a colpitz-butler
oscilator.

Colin =^.^=

I don't know...Could it be the combination of the Collpitts and Armstrong
oscillators, otherwise known as the ARM-PITS Circuit?

 

[Winfield Hill]

colin wrote...

Winfield Hill wrote ...

The OP275 ? doesnt mention it on the info page, I wonder what
the butler part consists of as it doesnt show an internal
schematic on the datasheet.

That's Butler, named after the inventor, James R Butler,
see his 1992 patent # 5101126 conveniently downloaded at
http://free.patentfetcher.com/Patent-Fetcher-Form.php

The OP176 and SSM2275 opamp datasheets have schematics.
He uses a bipolar input for low noise, plus a JFET input
for high slew rates for input error voltages over 120mV.
Butler was part of the Precision Monolithics team, SFAIK.

 

[Ken Smith]

[... Butler oscillator ...]

That would only apply if you aren't using AGC. They must be describing a
specific implementation of the Butler.

You don't need AGC in a Butler, you can control amplitude with proper
selection of the values in the "harmonic" tank.

Huh???? You either need a nonlinear effect, like bottoming or the gm
reduction with amplitude, or an AGC. If everything is linear and there is
enough gain, the oscillation will grow until something starts to limit it.

AGC has the advantage over limiting that it doesn't cause the noise near
harmonics to be aliased down to near the carrier.

 

[Jim Thompson]

[... Butler oscillator ...]

That would only apply if you aren't using AGC. They must be describing a
specific implementation of the Butler.

You don't need AGC in a Butler, you can control amplitude with proper
selection of the values in the "harmonic" tank.

Huh???? You either need a nonlinear effect, like bottoming or the gm
reduction with amplitude, or an AGC. If everything is linear and there is
enough gain, the oscillation will grow until something starts to limit it.

AGC has the advantage over limiting that it doesn't cause the noise near
harmonics to be aliased down to near the carrier.

The "Butlers" I've built have current sources in the bottom, not
resistors ;-)

...Jim Thompson

 

[Ken Smith]

Huh???? You either need a nonlinear effect, like bottoming or the gm
reduction with amplitude, or an AGC. If everything is linear and there is
enough gain, the oscillation will grow until something starts to limit it.

AGC has the advantage over limiting that it doesn't cause the noise near
harmonics to be aliased down to near the carrier.

The "Butlers" I've built have current sources in the bottom, not
resistors ;-)[/QUOTE]

Yes, and the transistors run from cut-off to conducting that much current.
The nonlinearity in this case is the "gm reduction" type. The further
into class C the stage goes, the lower its gm at the fundamental. It is
still a nonlinearity that is limiting the amplitude.

"gm reduction" is in some ways a very good way to regulate the amplitude
in an oscillator. Bottoming lowers the Q and because of saturation times
pulls the frequency. The problem with "gm reduction" is that the
amplitude varies from unit to unit more than an oscillator that uses
bottoming.

 

[Jim Thompson]

The "Butlers" I've built have current sources in the bottom, not
resistors ;-)

Yes, and the transistors run from cut-off to conducting that much current.
The nonlinearity in this case is the "gm reduction" type. The further
into class C the stage goes, the lower its gm at the fundamental. It is
still a nonlinearity that is limiting the amplitude.

"gm reduction" is in some ways a very good way to regulate the amplitude
in an oscillator. Bottoming lowers the Q and because of saturation times
pulls the frequency. The problem with "gm reduction" is that the
amplitude varies from unit to unit more than an oscillator that uses
bottoming.
[/QUOTE]

It doesn't work that way. The right-hand current source is smaller
than the left, so what is limited is drive current into a tank with
limited (and known) Q.

...Jim Thompson