100KHz low pass with constant 50R input impedance

[Clifford Heath]

On 18/6/19 5:20 am, makolber@yahoo.com wrote:

It seems it should be possible to do away with or reduce some of the
resistors (at the expense of greater dependence on the 50R nature of the
load) and in the process to reduce the loss... but I can't see how to
calculate that.

Any thoughts on reducing the loss while staying close to 50R in and out
of the passband?



A common method at Microdyne was to use a 3dB pad between the filter and a mismatch source or load.

yes,

the key concept is that an ideal LC filter alone creates loss in the stop band and transition band by REFLECTING energy back to the input and thus CANNOT present broadband 50 Ohms. It has no other option. Ideal L C components cannot dissipate energy. So to have a wideband 50 Ohm input you NEED to use a diplexer configuration of HPF with LPF with a dummy load or a pad.

If you are telling Terrell that his approach doesn't work, you are
wrong. A 3dB pad attenuates reflected energy by 6dB, and it's a common
approach that doesn't require a diplexer. It only works if the SNR
effect is allowable. The diplexer works when you need to remove more of
the reflections.

 

yes,

the key concept is that an ideal LC filter alone creates loss in the stop band and transition band by REFLECTING energy back to the input and thus CANNOT present broadband 50 Ohms. It has no other option. Ideal L C components cannot dissipate energy. So to have a wideband 50 Ohm input you NEED to use a diplexer configuration of HPF with LPF with a dummy load or a pad.

If you are telling Terrell that his approach doesn't work, you are
wrong. A 3dB pad attenuates reflected energy by 6dB, and it's a common
approach that doesn't require a diplexer. It only works if the SNR
effect is allowable. The diplexer works when you need to remove more of
the reflections.

you must have missed the last 3 words of my post

"or a pad"

m

 

[Simon S Aysdie]

On Saturday, June 15, 2019 at 6:24:45 AM UTC-7, John Larkin wrote:

On Sat, 15 Jun 2019 20:24:10 +1000, Clifford Heath
no.spam@please.net> wrote:

On 15/6/19 12:44 pm, John Larkin wrote:
On Sat, 15 Jun 2019 12:20:39 +1000, Clifford Heath
no.spam@please.net> wrote:

A down-conversion mixer from up to 200MHz down to 50KHz low IF likes to
see a solid 50R load, and not see signals reflected from a following
low-IF filter. It should not be necessary to use a really fast op-amp
(that can deal with image frequencies) in the LPF, if a passive LPF is
used first. So that's what this filter is intended for.

https://www.dropbox.com/s/ixzbczde1bqo91m/100KHz50R_LPF.jpg

The filter sketched has 100KHz corner frequency, and a very constant 50R
input impedance. I got this working nicely in LTSpice, but can't find
the ASC file now.

It seems it should be possible to do away with or reduce some of the
resistors (at the expense of greater dependence on the 50R nature of the
load) and in the process to reduce the loss... but I can't see how to
calculate that.

Any thoughts on reducing the loss while staying close to 50R in and out
of the passband?

Clifford Heath.

Jeroen here posted some designs for constant-input-impedance filters.
I did some simplified versions, not as constant as his.

The idea is to make an LC filter, and hang a network across the input
to absorb the stopband drive that the filter wants to reflect. The
simplest such network is a series RC to ground.

To stay resistive, it has to match the order of the-band.

Your filter looks pretty lossy to me.

It's necessary to "lose" the stop-band signal somewhere, so that has to
be resistive. The question is how not to lose so much pass-band.

Here are Jeroen's notes:

Filed away for reference thanks, but these are all band-pass filters,
and >=3rd order, which I don't need.

Jeroen's filters are all lowpass and start at 3rd order.


Do you need wideband termination on both ends of the filter?

I don't think the kind of LPF I will follow this with (100KHz) will show
much impedance variation from 1MHz up, so no, I don't think I do. If I
needed matching at the passive filter output I would use a buffer stage.

I just want to stop most of the RF from reaching the steep op-amp filter
after the passive filter.

Clifford Heath.

I was thinking that an ideal LC lowpass can be paralleled at its input
with an ideal LC highpass, both filters load-end terminated. The
paralleled input end must look like a perfect wideband 50 ohms.

(derived by the maximum power transfer theorem and conservation of
energy)

So the best front-end correction network for a lowpass might be a
terminated highpass.

For the first-order case, the lowpass is just series L into the 50r
load, and the absorber is C+50r to ground. If L/50 = C*50, the input
looks like 50 ohms wideband.

We call it a "diplexer."

It is standard fare for implementing a wideband load for a mixer.

 

[Jeroen Belleman]

On 2019-06-15 18:49, John Larkin wrote:

On Sat, 15 Jun 2019 17:33:39 +0200, Jeroen Belleman
jeroen@nospam.please> wrote:


There are also constant-resistance bridged T or L sections. These
have first-order frequency responses. Let me know if you're
interested.


I'd like to see that, if it's not too much trouble. We are doing some
work with distributed amplifiers and it may relate.

I've been a little short of time these last few days and I haven't
the opportunity to put together something polished. Then again, these
little circuits are very simple.
See <http://cern.ch/jeroen/CRnetworks>.

I've been using them to hide impedance excursions of matched-input
impedance LNAs and to simulate the frequency responses of gadgets
that do not fit on my lab bench.

Jeroen Belleman

 

[Simon S Aysdie]

On Saturday, June 15, 2019 at 12:36:29 AM UTC-7, Clifford Heath wrote:

On 15/6/19 4:40 pm, upsidedown@downunder.com wrote:
On Sat, 15 Jun 2019 12:20:39 +1000, Clifford Heath
no.spam@please.net> wrote:

A down-conversion mixer from up to 200MHz down to 50KHz low IF likes to
see a solid 50R load, and not see signals reflected from a following
low-IF filter. It should not be necessary to use a really fast op-amp
(that can deal with image frequencies) in the LPF, if a passive LPF is
used first. So that's what this filter is intended for.

https://www.dropbox.com/s/ixzbczde1bqo91m/100KHz50R_LPF.jpg

Do you need a constant impedance up to a few gigahertz or just some
nice 50 ohm load around the 400 MHz (RF+LO image frequency) ?

Well, up to GHz really, since this is for instrumentation use I cannot
know what antenna might be used, just that I don't want rectified spuria
upsetting things. We plan to digitise I & Q from the output of the LPF,
so can reduce the resolution BW below 200KHz in DSP. Can't get rid of
something odd that aliases into that band though, so the LPF should be
clean and flat-looking.

What is the sampling rate? How well will the high frequency stuff go through the amp, or will it generate spurious stuff in the amp? While terminating (not reflecting back) all frequencies is a goal for the mixer, some level of rejection of the unwanted "high" frequencies is also implied. What is the rejection mask?

The "normal-simple" way is to use a straight 3 dB point overlap LPF-HPF diplexer where the HPF leg dumps into a 50 ohm term. 200 kHz and 200 MHz are 3 orders of magnitude apart, so low order filters (even 1st order?) should suffice minus additional requirements.

A trouble you may have if the LPF corner is really low, is that the LPF inductor (at the diplexer common) won't be an inductor at 1 GHz, and the HPF cap (at the diplexer common) won't be a cap. Your desired return loss may not pan out. Of course, a conical inductor is great, but very expensive, fragile, and Production hates them. There are some tricks, but the higher you can push your corner-overlap up, then the easier it is to make the high pass leg look like a good term at 1 GHz. In principle, you can use multiple diplexers. (Split once, split again.) Use "real" parts in your sims.

 

[Clifford Heath]

On 19/6/19 11:37 am, Simon S Aysdie wrote:

On Saturday, June 15, 2019 at 12:36:29 AM UTC-7, Clifford Heath wrote:
On 15/6/19 4:40 pm, upsidedown@downunder.com wrote:
On Sat, 15 Jun 2019 12:20:39 +1000, Clifford Heath
no.spam@please.net> wrote:

A down-conversion mixer from up to 200MHz down to 50KHz low IF likes to
see a solid 50R load, and not see signals reflected from a following
low-IF filter. It should not be necessary to use a really fast op-amp
(that can deal with image frequencies) in the LPF, if a passive LPF is
used first. So that's what this filter is intended for.

https://www.dropbox.com/s/ixzbczde1bqo91m/100KHz50R_LPF.jpg

Do you need a constant impedance up to a few gigahertz or just some
nice 50 ohm load around the 400 MHz (RF+LO image frequency) ?

Well, up to GHz really, since this is for instrumentation use I cannot
know what antenna might be used, just that I don't want rectified spuria
upsetting things. We plan to digitise I & Q from the output of the LPF,
so can reduce the resolution BW below 200KHz in DSP. Can't get rid of
something odd that aliases into that band though, so the LPF should be
clean and flat-looking.

What is the sampling rate? How well will the high frequency stuff go through the amp, or will it generate spurious stuff in the amp? While terminating (not reflecting back) all frequencies is a goal for the mixer, some level of rejection of the unwanted "high" frequencies is also implied. What is the rejection mask?

The "normal-simple" way is to use a straight 3 dB point overlap LPF-HPF diplexer where the HPF leg dumps into a 50 ohm term. 200 kHz and 200 MHz are 3 orders of magnitude apart, so low order filters (even 1st order?) should suffice minus additional requirements.

A trouble you may have if the LPF corner is really low, is that the LPF inductor (at the diplexer common) won't be an inductor at 1 GHz, and the HPF cap (at the diplexer common) won't be a cap. Your desired return loss may not pan out. Of course, a conical inductor is great, but very expensive, fragile, and Production hates them. There are some tricks, but the higher you can push your corner-overlap up, then the easier it is to make the high pass leg look like a good term at 1 GHz. In principle, you can use multiple diplexers. (Split once, split again.) Use "real" parts in your sims.

All good points. Using two diplexers is pretty easy and a good idea.
Especially since I'd like to increase the upper frequency limit, if I
can work out how to easily produce two phase-locked signals, f0 and
(f0+IF). The first is the analyser stimulus, the second used to
demodulate signals from the port couplers, preserving RF phase in the IF
phase. I can do that to 200MHz using an AD9959 DDS.

The "normal-simple" 3dB way is what I ended up with, more or less; the
following active filter can deal with 2MHz but I don't want it to see
200MHz.

The goal is to get out of the RF domain as quickly and accurately as
possible, into a low-IF signal processing chain I can accurately
digitise and process in software.

Clifford Heath.

 

[Simon S Aysdie]

On Tuesday, June 18, 2019 at 6:37:06 PM UTC-7, Simon S Aysdie wrote:

On Saturday, June 15, 2019 at 12:36:29 AM UTC-7, Clifford Heath wrote:
On 15/6/19 4:40 pm, upsidedown@downunder.com wrote:
On Sat, 15 Jun 2019 12:20:39 +1000, Clifford Heath
no.spam@please.net> wrote:

A down-conversion mixer from up to 200MHz down to 50KHz low IF likes to
see a solid 50R load, and not see signals reflected from a following
low-IF filter. It should not be necessary to use a really fast op-amp
(that can deal with image frequencies) in the LPF, if a passive LPF is
used first. So that's what this filter is intended for.

https://www.dropbox.com/s/ixzbczde1bqo91m/100KHz50R_LPF.jpg

Do you need a constant impedance up to a few gigahertz or just some
nice 50 ohm load around the 400 MHz (RF+LO image frequency) ?

Well, up to GHz really, since this is for instrumentation use I cannot
know what antenna might be used, just that I don't want rectified spuria
upsetting things. We plan to digitise I & Q from the output of the LPF,
so can reduce the resolution BW below 200KHz in DSP. Can't get rid of
something odd that aliases into that band though, so the LPF should be
clean and flat-looking.

What is the sampling rate? How well will the high frequency stuff go through the amp, or will it generate spurious stuff in the amp? While terminating (not reflecting back) all frequencies is a goal for the mixer, some level of rejection of the unwanted "high" frequencies is also implied. What is the rejection mask?

The "normal-simple" way is to use a straight 3 dB point overlap LPF-HPF diplexer where the HPF leg dumps into a 50 ohm term. 200 kHz and 200 MHz are 3 orders of magnitude apart, so low order filters (even 1st order?) should suffice minus additional requirements.

A trouble you may have if the LPF corner is really low, is that the LPF inductor (at the diplexer common) won't be an inductor at 1 GHz, and the HPF cap (at the diplexer common) won't be a cap. Your desired return loss may not pan out. Of course, a conical inductor is great, but very expensive, fragile, and Production hates them. There are some tricks, but the higher you can push your corner-overlap up, then the easier it is to make the high pass leg look like a good term at 1 GHz. In principle, you can use multiple diplexers. (Split once, split again.) Use "real" parts in your sims.

"I cannot know what antenna might be used"

Yeah, RF isolation isn't infinite, there will be harmonic mixing possible, and other indeterminate products sneaking in for an "open to the world antenna." A filter somewhere prior to the mixer might be nice too. Not sure if it is practical. So you have a single conversion cascade?

 

On Tue, 18 Jun 2019 18:49:14 -0700 (PDT), Simon S Aysdie
<gwhite@ti.com> wrote:

On Tuesday, June 18, 2019 at 6:37:06 PM UTC-7, Simon S Aysdie wrote:
On Saturday, June 15, 2019 at 12:36:29 AM UTC-7, Clifford Heath wrote:
On 15/6/19 4:40 pm, upsidedown@downunder.com wrote:
On Sat, 15 Jun 2019 12:20:39 +1000, Clifford Heath
no.spam@please.net> wrote:

A down-conversion mixer from up to 200MHz down to 50KHz low IF likes to
see a solid 50R load, and not see signals reflected from a following
low-IF filter. It should not be necessary to use a really fast op-amp
(that can deal with image frequencies) in the LPF, if a passive LPF is
used first. So that's what this filter is intended for.

https://www.dropbox.com/s/ixzbczde1bqo91m/100KHz50R_LPF.jpg

Do you need a constant impedance up to a few gigahertz or just some
nice 50 ohm load around the 400 MHz (RF+LO image frequency) ?

Well, up to GHz really, since this is for instrumentation use I cannot
know what antenna might be used, just that I don't want rectified spuria
upsetting things. We plan to digitise I & Q from the output of the LPF,
so can reduce the resolution BW below 200KHz in DSP. Can't get rid of
something odd that aliases into that band though, so the LPF should be
clean and flat-looking.

What is the sampling rate? How well will the high frequency stuff go through the amp, or will it generate spurious stuff in the amp? While terminating (not reflecting back) all frequencies is a goal for the mixer, some level of rejection of the unwanted "high" frequencies is also implied. What is the rejection mask?

The "normal-simple" way is to use a straight 3 dB point overlap LPF-HPF diplexer where the HPF leg dumps into a 50 ohm term. 200 kHz and 200 MHz are 3 orders of magnitude apart, so low order filters (even 1st order?) should suffice minus additional requirements.

A trouble you may have if the LPF corner is really low, is that the LPF inductor (at the diplexer common) won't be an inductor at 1 GHz, and the HPF cap (at the diplexer common) won't be a cap. Your desired return loss may not pan out. Of course, a conical inductor is great, but very expensive, fragile, and Production hates them. There are some tricks, but the higher you can push your corner-overlap up, then the easier it is to make the high pass leg look like a good term at 1 GHz. In principle, you can use multiple diplexers. (Split once, split again.) Use "real" parts in your sims.

"I cannot know what antenna might be used"

Is the antenna directly connected to a passive mixer ? In that case, a
-3 dB post mixer pad will degrade the noise figure even further. With
an RF amplifier prior to the mixer, the pad losses can be compensated
by the RF amplifier.


>Yeah, RF isolation isn't infinite, there will be harmonic mixing possible, and other indeterminate products sneaking in for an "open to the world antenna." A filter somewhere prior to the mixer might be nice too. Not sure if it is practical. So you have a single conversion cascade?

When a mixer is driven by a square wave, there will be mixing products
with +/-3LO, +/- 5LO etc. An octave filter in front of the mixer will
help eliminate spurious responses like 2RF-3LO and so on. Also there
are images at mRF+nLO which also needs to be terminated. Thus a 200
MHz receiver might have mixing products over 1 GHz.

If e.g. a 1/4 wave antenna is used, it has also nice resistive
impedances at 3RF, 5RF, in which the antenna works as 3/4 resp. 5/4
wave antenna resonances. A LPF above highest needed RF frequency
helps a lot.

A 1/4 wave antenna doesn't have any resonances below resonance and at
those frequencies the antenna is highly reactive with a resistive
component of only a few ohms, so it doesn't match well with a 50 ohm
input.

For this reason a LPF filter above wanted frequency is important, but
in order to avoid harmonic mixing it is a good idea to filter out
frequencies below RF/2, if they are strong for some reason.

 

[amdx]

On 6/18/2019 8:49 PM, Simon S Aysdie wrote:

On Tuesday, June 18, 2019 at 6:37:06 PM UTC-7, Simon S Aysdie wrote:
On Saturday, June 15, 2019 at 12:36:29 AM UTC-7, Clifford Heath wrote:
On 15/6/19 4:40 pm, upsidedown@downunder.com wrote:
On Sat, 15 Jun 2019 12:20:39 +1000, Clifford Heath
no.spam@please.net> wrote:

A down-conversion mixer from up to 200MHz down to 50KHz low IF likes to
see a solid 50R load, and not see signals reflected from a following
low-IF filter. It should not be necessary to use a really fast op-amp
(that can deal with image frequencies) in the LPF, if a passive LPF is
used first. So that's what this filter is intended for.

https://www.dropbox.com/s/ixzbczde1bqo91m/100KHz50R_LPF.jpg

Do you need a constant impedance up to a few gigahertz or just some
nice 50 ohm load around the 400 MHz (RF+LO image frequency) ?

Well, up to GHz really, since this is for instrumentation use I cannot
know what antenna might be used, just that I don't want rectified spuria
upsetting things. We plan to digitise I & Q from the output of the LPF,
so can reduce the resolution BW below 200KHz in DSP. Can't get rid of
something odd that aliases into that band though, so the LPF should be
clean and flat-looking.

What is the sampling rate? How well will the high frequency stuff go through the amp, or will it generate spurious stuff in the amp? While terminating (not reflecting back) all frequencies is a goal for the mixer, some level of rejection of the unwanted "high" frequencies is also implied. What is the rejection mask?

The "normal-simple" way is to use a straight 3 dB point overlap LPF-HPF diplexer where the HPF leg dumps into a 50 ohm term. 200 kHz and 200 MHz are 3 orders of magnitude apart, so low order filters (even 1st order?) should suffice minus additional requirements.

A trouble you may have if the LPF corner is really low, is that the LPF inductor (at the diplexer common) won't be an inductor at 1 GHz, and the HPF cap (at the diplexer common) won't be a cap. Your desired return loss may not pan out. Of course, a conical inductor is great, but very expensive, fragile, and Production hates them. There are some tricks, but the higher you can push your corner-overlap up, then the easier it is to make the high pass leg look like a good term at 1 GHz. In principle, you can use multiple diplexers. (Split once, split again.) Use "real" parts in your sims.

"I cannot know what antenna might be used"

Yeah, RF isolation isn't infinite, there will be harmonic mixing possible, and other indeterminate products sneaking in for an "open to the world antenna." A filter somewhere prior to the mixer might be nice too. Not sure if it is practical. So you have a single conversion cascade?

I'm not really following this, I thought a Chebychev's were 50ohms in
and out. (or what ever impedance you build them for)
That said this is the only Diplexer I know, it was called DC to Daylight
and was for a direct conversion receiver. (50 ohm)
Pay attention to inductor resistance and adjust R accordingly.
Mikek

 

[Tim Williams]

"amdx" <nojunk@knology.net> wrote in message
news:qp5i9s$vma$1@dont-email.me...

I'm not really following this, I thought a Chebychev's were 50ohms in and
out. (or what ever impedance you build them for)
That said this is the only Diplexer I know, it was called DC to Daylight
and was for a direct conversion receiver. (50 ohm)
Pay attention to inductor resistance and adjust R accordingly.

Characteristic impedance is the number that keeps coming up time and again
in such circuits, but it takes a very special filter to have it all the
time!

In particular, the impedance of such a filter varies up and down around the
[geometric] mean, by a ~constant amount across frequencies (how much and how
often, depends on the filter type).

Think of the impedance of a transmission line stub, it has peaks and troughs
but the mean value is Zo.

Butterworth is interesting for many reasons, not just maximal flatness but
also for the sharpest(?) filter with a one-to-one impedance curve (i.e., the
choke-input kind, the impedance keeps going up and up into cutoff; the
cap-input kind, the impedance keeps going down and down).

Cheby doesn't have this property, and so can't be used to make (perfectly
flat) diplexers IIRC (if nothing else, the circuit is necessarily more
complex).

Tim

--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Design
Website: https://www.seventransistorlabs.com/

 

[amdx]

On 10/27/2019 11:29 PM, Tim Williams wrote:

"amdx" <nojunk@knology.net> wrote in message
news:qp5i9s$vma$1@dont-email.me...
I'm not really following this, I thought a Chebychev's were 50ohms in
and out. (or what ever impedance you build them for)
That said this is the only Diplexer I know, it was called DC to
Daylight and was for a direct conversion receiver. (50 ohm)
 Pay attention to inductor resistance and adjust R accordingly.


Characteristic impedance is the number that keeps coming up time and
again in such circuits, but it takes a very special filter to have it
all the time!

In particular, the impedance of such a filter varies up and down around
the [geometric] mean, by a ~constant amount across frequencies (how much
and how often, depends on the filter type).

Think of the impedance of a transmission line stub, it has peaks and
troughs but the mean value is Zo.

Butterworth is interesting for many reasons, not just maximal flatness
but also for the sharpest(?) filter with a one-to-one impedance curve
(i.e., the choke-input kind, the impedance keeps going up and up into
cutoff; the cap-input kind, the impedance keeps going down and down).

Cheby doesn't have this property, and so can't be used to make
(perfectly flat) diplexers IIRC (if nothing else, the circuit is
necessarily more complex).

Tim

Seems I forgot the link for the DC to Daylight diplexer.
https://www.dropbox.com/s/k0yn667awrxuwg7/DC%20to%20daylight%20Diplexer.jpg?dl=0

Mikek

 

[Jeroen Belleman]

amdx wrote:

On 10/27/2019 11:29 PM, Tim Williams wrote:
"amdx" <nojunk@knology.net> wrote in message
news:qp5i9s$vma$1@dont-email.me...
I'm not really following this, I thought a Chebychev's were 50ohms in
and out. (or what ever impedance you build them for)
That said this is the only Diplexer I know, it was called DC to
Daylight and was for a direct conversion receiver. (50 ohm)
Pay attention to inductor resistance and adjust R accordingly.


Characteristic impedance is the number that keeps coming up time and
again in such circuits, but it takes a very special filter to have it
all the time!

In particular, the impedance of such a filter varies up and down
around the [geometric] mean, by a ~constant amount across frequencies
(how much and how often, depends on the filter type).

Think of the impedance of a transmission line stub, it has peaks and
troughs but the mean value is Zo.

Butterworth is interesting for many reasons, not just maximal flatness
but also for the sharpest(?) filter with a one-to-one impedance curve
(i.e., the choke-input kind, the impedance keeps going up and up into
cutoff; the cap-input kind, the impedance keeps going down and down).

Cheby doesn't have this property, and so can't be used to make
(perfectly flat) diplexers IIRC (if nothing else, the circuit is
necessarily more complex).

Tim

Seems I forgot the link for the DC to Daylight diplexer.
https://www.dropbox.com/s/k0yn667awrxuwg7/DC%20to%20daylight%20Diplexer.jpg?dl=0


Mikek

For Butterworth filters, it's easy. You'd use a filter for
a zero-impedance source and put the dual filter in parallel
at its input. ('Dual' means a filter with reciprocal element
values in the normalized prototype.)

For 'tame' filters, such as Bessel, Gaussian or linear
phase with small ripple, a small adaptation as described
here <http://cern.ch/jeroen/crfilter/crfilter.html> can
do the job. For more aggressive filters, Chebychev, Cauer
and whatnot, I have no good solution. I worked out a
procedure for perfectly matched Bessel filters, but I
realized that their practical realization wouldn't be any
better than the approximate solution described in my web
page.

Component parasitics limit the useful bandwidth over which
a diplexer will work. That also goes for your DC-to-daylight
diplexer. You can extend the bandwidth of the match with
one or more constant-impedance bridged-T low-pass stages.

Jeroen Belleman