calculate the Q factor of LC bandpass

filter,

Cf,

lowpass

Sorry? Where did I claim that even once, let alone habitually?

Butterworth, Bessel, and so on are specific examples requiring specific pole placements. That has nothing to do with what we were arguing about, except that they're both examples of one-hump responses from a

Cheers

Phil Hobbs

(Who has been building and tuning actual physical non-simulated LC filters for about 35 years at this point, and who owns 2 grid dip meters for the purpose)

filter,

Cf,

lowpass

My real point is that you have the remarkable ability to be very arrogant, and very wrong, simultaneously.

out

in

filter,

Cf,

lowpass

I generally look up a lowpass prototype in Williams, do the resonance thing if it's a bandpass, then normalize to the actual cutoff frequency. Lately, we have been using the remarkable Nuhertz software, which does a huge range of LC filter designs using finite-Q, standard-value parts.

I'm sure that Jim always designs LC filters from scratch, and does all his multiplication and long division by hand.

out

in

filter,

Cf,

lowpass

You disclaimed "rickman's" finding of two peaks... did you not?

I've been building production quantity filters before there was ever CAD to do the calculations >:-}

And I also own a grid-dip meter, except it's FET, not a toob-type ;-) ...Jim Thompson

through

out

in

filter,

Cf,

lowpass

Show me. My first post in the thread--March 8, 9:59 PM:

|The source and load are left out, but otherwise, that's your basic |4-pole bandpass ladder filter. To make a 6-pole, he could add another |series LC at the output (wye) or another parallel one at the |input(delta).

When the two-peak thing came up, you started confusing the pole positions with the peak positions, and made the rather ridiculous claim that a one-peak response was "physically impossible".

Now you're just trying to muddy the water and sniping from the sidelines, apparently because you can't stand the loss of face. You got something wrong. Just like we all do, often. No biggie.

Get over it.

Cheers

Phil Hobbs

the

through

out

in

filter,

Cf,

lowpass

Really? Who is it that is confused?

As posted! You know, I know it, why the CYA BS?

You're becoming more and more Larkin-like every day :-(

You forgot the "cluck-cluck".

...Jim Thompson

certainly

the

and

through

out

realizable in

came

two-hump

filter,

frequency Cf,

parallel

lowpass

Sad old fool. Junkyard dogs don't age well.

Larkin can't find the pole locations without a CAD program. I can. ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

certainly

load,

the

and

through

brains out

realizable in

achievable.

came

two-hump

filter,

frequency Cf,

parallel

lowpass

I can see I'm going to have to separate Phil and Jim, but this isn't even *your* discussion! You will just have to stay after class in detention.

Rick

certainly

load,

the

and

through

brains out

realizable in

achievable.

apart? /

came

two-hump

lowpass filter,

frequency Cf,

parallel

the lowpass

this

He started it! Pbbbbbbbbt!

I assume, from his posts, that he designs LC filters from scratch, using pole-zero math with pencil and paper. That's an interesting claim.

Berkeley

This

Oh well. If it is not secret perhaps you could send me a copy of the file and let me tinker it a bit. I might figure it out.

?-)

two-hump

component

Also

lowpass filter,

frequency Cf,

parallel

of the lowpass

prototype.

be used

with

used

do with

before

s and

forgotten.

warp is

beautiful

;)

ever

you

just

harmonics

restore

low

phase is

but

Hate to disappoint but my book is like FilterPro. But i will take a look at the book you pointed to. Would have been a great help say 30 years ago.

?-)

two-hump

filter,

frequency Cf,

parallel

lowpass

I think I bought it in 1970 ;-) ...Jim Thompson

Another Polish joke?

For what it is worth, i remember it as rickman claiming only two peaks is possible. Please recheck the thread.

specific

meters

As the original schematic was posted. Add a source impedance and a load impedance, d'Q-ing, then you can make a smooth top. I vaguely recall some theorem saying you can't have COINCIDENT poles in a PASSIVE network. I think they can be on the same circle, but I haven't been able to verify that yet either.

'Twasn't secret, but it was a while ago. If I can't find it, I'll dream up a network that 3f will do a .pz analysis on, and let you have it. LTspice netlists need quite a bit of massage to run in Berkeley 3f, which doesn't support .param, and other things.

LTspice just barfs at a .pz statement.

Looks like HSpice can do it (Chapter 24), but must have transfer function in the form H(s)=N(s)/D(s) ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at