On Tue, 16 Mar 2004 20:26:15 +0000, Paul Burridge
wrote:
If I'm not mistaken, "tuned amplification" IS "filtering".

An argument over semantics, then. AFAIC it's not filtering as such.
It introduces a high degree of selectivity, certainly. But when
someone says "filtering" I assume they're taking about a pi-network or
something of that sort, between stages or at the end of a chain of
stages.

Wow - the strange things you learn on this thread! So how many poles does a
circuit need for it to be called a "filter"?

Tony (remove the "_" to reply by email)

On Tue, 16 Mar 2004 20:26:15 +0000, Paul Burridge
wrote:


An argument over semantics, then. AFAIC it's not filtering as such.
It introduces a high degree of selectivity, certainly. But when
someone says "filtering" I assume they're taking about a pi-network or
something of that sort, between stages or at the end of a chain of
stages.

---
Any network which exhibits frequency selectivity is a filter, whether or
not you're concerned about whether or not it is or is not.

Think about it... from the lowly filter capacitor to the exalted
brickwall filter, they're all discrimating against a frequency or a set
of frequencies which we have told them we don't want them to let us see.

Filters, every one.

Just for grins, take a little trip over to a.b.s.e. (same subject
heading)and take a look at what John Larkin's series resonant filter
feeding a parallel resonant filter strategy looks like as far as
allowing you to get a fifth harmonic from a fundamental square wave
goes.

--
John Fields

On Tue, 16 Mar 2004 20:26:15 +0000, Paul Burridge
wrote:


An argument over semantics, then. AFAIC it's not filtering as such.
It introduces a high degree of selectivity, certainly. But when
someone says "filtering" I assume they're taking about a pi-network or
something of that sort, between stages or at the end of a chain of
stages.

---
Any network which exhibits frequency selectivity is a filter, whether or
not you're concerned about whether or not it is or is not.

Think about it... from the lowly filter capacitor to the exalted
brickwall filter, they're all discrimating against a frequency or a set
of frequencies which we have told them we don't want them to let us see.

Filters, every one.

Just for grins, take a little trip over to a.b.s.e. (same subject
heading)and take a look at what John Larkin's series resonant filter
feeding a parallel resonant filter strategy looks like as far as
allowing you to get a fifth harmonic from a fundamental square wave
goes.

--
John Fields

On Tue, 16 Mar 2004 19:54:27 -0600, John Fields
wrote:

On Tue, 16 Mar 2004 20:26:15 +0000, Paul Burridge
wrote:


An argument over semantics, then. AFAIC it's not filtering as such.
It introduces a high degree of selectivity, certainly. But when
someone says "filtering" I assume they're taking about a pi-network or
something of that sort, between stages or at the end of a chain of
stages.

---
Any network which exhibits frequency selectivity is a filter, whether or
not you're concerned about whether or not it is or is not.

Think about it... from the lowly filter capacitor to the exalted
brickwall filter, they're all discrimating against a frequency or a set
of frequencies which we have told them we don't want them to let us see.

Filters, every one.

Just for grins, take a little trip over to a.b.s.e. (same subject
heading)and take a look at what John Larkin's series resonant filter
feeding a parallel resonant filter strategy looks like as far as
allowing you to get a fifth harmonic from a fundamental square wave
goes.


That's just a standard bandpass. What you do is pick a normalized
lowpass filter that has the response shape you like, say a Tchebychev
(I know... various spellings) and scale it to the impedance Z' and
bandwidth W' you want. Then series resonate each L with a C, and
parallel resonate each C with an L, both at some desired center
frequency. Voila (pardon my French) a bandpass that's 2W' wide.

There's no real reason to cascade lossy Q-killing tuned transistor
stages when you can put all your Ls and Cs in one place.

John

On Tue, 16 Mar 2004 19:54:27 -0600, John Fields
wrote:

On Tue, 16 Mar 2004 20:26:15 +0000, Paul Burridge
wrote:


An argument over semantics, then. AFAIC it's not filtering as such.
It introduces a high degree of selectivity, certainly. But when
someone says "filtering" I assume they're taking about a pi-network or
something of that sort, between stages or at the end of a chain of
stages.

---
Any network which exhibits frequency selectivity is a filter, whether or
not you're concerned about whether or not it is or is not.

Think about it... from the lowly filter capacitor to the exalted
brickwall filter, they're all discrimating against a frequency or a set
of frequencies which we have told them we don't want them to let us see.

Filters, every one.

Just for grins, take a little trip over to a.b.s.e. (same subject
heading)and take a look at what John Larkin's series resonant filter
feeding a parallel resonant filter strategy looks like as far as
allowing you to get a fifth harmonic from a fundamental square wave
goes.


That's just a standard bandpass. What you do is pick a normalized
lowpass filter that has the response shape you like, say a Tchebychev
(I know... various spellings) and scale it to the impedance Z' and
bandwidth W' you want. Then series resonate each L with a C, and
parallel resonate each C with an L, both at some desired center
frequency. Voila (pardon my French) a bandpass that's 2W' wide.

There's no real reason to cascade lossy Q-killing tuned transistor
stages when you can put all your Ls and Cs in one place.

John

For a comparison of rectangular waveform on-times versus spectral
content, the following calculations were done on my WAVESPEC
program for a 0.50 to 0.25 times repetition period and with rise and
fall times equal to 0.02 times repetition period.

If the fundamental energy is the reference, then the harmonics
are down from that 0 db by the values shown:

/------------ width rel. to rep. period -==--------\
Harm 0.50 0.45 0.40 0.35 0.30 0.25
1 0 0 0 0 0 0
3 -6.4 -7.3 -8.4 -9.6 -11.1 -12.8
5 -28.0 -17.8 -15.1 -15.7 -20.3 *
7 -28.1 -20.1 -26.2 -27.7 -26.6 -29.1
9 -28.4 -24.6 -32.1 -44.2 -43.0 *
11 -28.7 -32.7 -42.2 -25.1 -35.8 -28.7
13 -29.1 * -29.1 * -29.1 *

* too far down to matter, not enough there

The above will hold true at any fundamental frequency provided the
rise and fall times are equal and each equal to 0.02 times the
repetition period. Those numbers will change given faster or slower
rise/fall times. All db calculated as 20 x Log (voltage). Width is
determined at the baseline, not the 50% amplitude point.

Len Anderson
retired (from regular hours) electronic engineer person

For a comparison of rectangular waveform on-times versus spectral
content, the following calculations were done on my WAVESPEC
program for a 0.50 to 0.25 times repetition period and with rise and
fall times equal to 0.02 times repetition period.

If the fundamental energy is the reference, then the harmonics
are down from that 0 db by the values shown:

/------------ width rel. to rep. period -==--------\
Harm 0.50 0.45 0.40 0.35 0.30 0.25
1 0 0 0 0 0 0
3 -6.4 -7.3 -8.4 -9.6 -11.1 -12.8
5 -28.0 -17.8 -15.1 -15.7 -20.3 *
7 -28.1 -20.1 -26.2 -27.7 -26.6 -29.1
9 -28.4 -24.6 -32.1 -44.2 -43.0 *
11 -28.7 -32.7 -42.2 -25.1 -35.8 -28.7
13 -29.1 * -29.1 * -29.1 *

* too far down to matter, not enough there

The above will hold true at any fundamental frequency provided the
rise and fall times are equal and each equal to 0.02 times the
repetition period. Those numbers will change given faster or slower
rise/fall times. All db calculated as 20 x Log (voltage). Width is
determined at the baseline, not the 50% amplitude point.

Len Anderson
retired (from regular hours) electronic engineer person

Avery Fineman wrote:

For a comparison of rectangular waveform on-times versus spectral
content, the following calculations were done on my WAVESPEC
program for a 0.50 to 0.25 times repetition period and with rise and
fall times equal to 0.02 times repetition period.

If the fundamental energy is the reference, then the harmonics
are down from that 0 db by the values shown:

/------------ width rel. to rep. period -==--------\
Harm 0.50 0.45 0.40 0.35 0.30 0.25
1 0 0 0 0 0 0
3 -6.4 -7.3 -8.4 -9.6 -11.1 -12.8
5 -28.0 -17.8 -15.1 -15.7 -20.3 *
7 -28.1 -20.1 -26.2 -27.7 -26.6 -29.1
9 -28.4 -24.6 -32.1 -44.2 -43.0 *
11 -28.7 -32.7 -42.2 -25.1 -35.8 -28.7
13 -29.1 * -29.1 * -29.1 *

* too far down to matter, not enough there

The above will hold true at any fundamental frequency provided the
rise and fall times are equal and each equal to 0.02 times the
repetition period. Those numbers will change given faster or slower
rise/fall times. All db calculated as 20 x Log (voltage). Width is
determined at the baseline, not the 50% amplitude point.

Len Anderson
retired (from regular hours) electronic engineer person

I wonder what happens to these numbers as the rise/fall time tends to
zero?

Peter Lawton

Avery Fineman wrote:

For a comparison of rectangular waveform on-times versus spectral
content, the following calculations were done on my WAVESPEC
program for a 0.50 to 0.25 times repetition period and with rise and
fall times equal to 0.02 times repetition period.

If the fundamental energy is the reference, then the harmonics
are down from that 0 db by the values shown:

/------------ width rel. to rep. period -==--------\
Harm 0.50 0.45 0.40 0.35 0.30 0.25
1 0 0 0 0 0 0
3 -6.4 -7.3 -8.4 -9.6 -11.1 -12.8
5 -28.0 -17.8 -15.1 -15.7 -20.3 *
7 -28.1 -20.1 -26.2 -27.7 -26.6 -29.1
9 -28.4 -24.6 -32.1 -44.2 -43.0 *
11 -28.7 -32.7 -42.2 -25.1 -35.8 -28.7
13 -29.1 * -29.1 * -29.1 *

* too far down to matter, not enough there

The above will hold true at any fundamental frequency provided the
rise and fall times are equal and each equal to 0.02 times the
repetition period. Those numbers will change given faster or slower
rise/fall times. All db calculated as 20 x Log (voltage). Width is
determined at the baseline, not the 50% amplitude point.

Len Anderson
retired (from regular hours) electronic engineer person

I wonder what happens to these numbers as the rise/fall time tends to
zero?

Peter Lawton

On Wed, 17 Mar 2004 11:06:42 +1000, Tony wrote:

On Tue, 16 Mar 2004 20:26:15 +0000, Paul Burridge
wrote:
If I'm not mistaken, "tuned amplification" IS "filtering".

An argument over semantics, then. AFAIC it's not filtering as such.
It introduces a high degree of selectivity, certainly. But when
someone says "filtering" I assume they're taking about a pi-network or
something of that sort, between stages or at the end of a chain of
stages.

Wow - the strange things you learn on this thread! So how many poles does a
circuit need for it to be called a "filter"?

"Words mean what I choose them to mean! No more; no less."
- the Red Queen
:-)
--

The BBC: Licensed at public expense to spread lies.