RF notch filter design (help)

Go To Last Post
36 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi there,

I would like to buid a notch filter for any of the CATV channels 31 - 39. The band frequencies for these channels are from 573.25 to 621.25 MHz. Each channel center frequency abstains 6 MHz from the previous one center frequency and 6 MHz from the next one center frequency. In other words Each channel has a 6 MHz bandwidth.

I have found that I cannot use coils beacause of their low Q for this notch filter. I have also found that I cannot use quartz filters because they cannot made for so high frequencies.

Next week I have to answer if I can design this circuit or not and I would like to know if there is any way to do it.

Please Help,

Thank you.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Guess this is a school assignment?

One way to go could be to use high impedance transmission lines (implemented as microstrip lines) to replace the inductors.

Google a bit for that, I remeber philips(now NXP) had a good discussion of this technique for matching power amplifiers. I can try to find the specific appnote tonight.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks CountZero,

I am waiting for your news.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Search cavity resonantors. How much do you have to reject the adjacent channels by? First impression is that this is a VERY serious filtering requirement.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

this might be a better alternative: http://www.skyworksinc.com/produ...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Each channel usefull bandwidth is 5MHz +/-2,5MHz from each center frequency. So each channel abstain 1MHz from it's previous one and 1MHz from it's next one.

In other words the notch filter has to let pass frequencies not more than 5 MHz.

I will inform you about my research.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

well, you can't get a filter that blocks everything! You surely have some specification like -20dB at this frequency offset and -40 at this and so on..

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

This is a very difficult project. High Q resonators are required and, at those frequencies, it is not likely to be achieved with transmission line equivalents.

One possibility is a comb filter for each channel. Another is the use of multiple helical resonators for each channel. Dishal tuning method would help you get everthing tuned up.

As CountZero suggests, you also do need some proper specifications, like bandwidth, skirt steepness, etc. Also, is your primary goal to attenuate the video signal, or do you need sound, also? That might change the structure of the filter(s).

But, in any case, this is NOT a trivial undertaking! Further, I hope you have a GOOD network analyzer to work with this. Otherwise, the pain level will be VERY high.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Err...
Wouldnt you just do a Down Conversion to an Intermediate Frequency where you would then apply the Filter. This way you only need a single filter, and it can be done at a 'nicer' frequency.

This is assuming that you want to select a channel and thus the filter is a band pass type. Ive only ever considered a notch filter as a band stop type or wave trap but I guess they can all be transposed.

Steve

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Is there any other way to do this?
I mean is there any other way (not notch filters) to exclude a TV channel without effecting its previous and its next one?
My mind focuses on RF amplifier used as attenuator.

Thanks.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Can you describe WHAT you are trying to do?

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The thing about the tuner, Icarus, is that they exclusively (at UHF frequencies) use a local oscillator to mix the original signal and produce an image at a frequency which can be filtered much more efficiently. They don't do very much at the 600MHz frequencies you're mentioning.

One trick that *might* have sufficient resolution - though frankly I doubt it - is to use carefully tuned stubs of coax cable 'teed' off your signal path; half a wavelength at 600MHz is about seventeen centimetres (depending on the transmission speed of the coax; I've assumed 200m/us). Tuning it would be a pig, and isolation if you need to filter more than one frequency would need to be considered... and basically, I suspect it would only act as a first order filter; you'd probably effectively lose the lot.

Neil

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Of course Jim;

I want to make a product using an ATMega48 micro and dip switches used for exclude one to eight channels (selectable) from UHF super band CATV 31-39.

The product will powered from th public network and will have as input the output from the video signal amp. and as output the coaxial cable going to the TV tuner.

My problem is that the frequencies are very high and each channels usefull frequency band is 5MHz and its center frequency abstains 6MHz from its previous and next one center frequency.

Thanks.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

But how much attenuation do you need?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Are you are trying to prevent certain channels from being viewed?

There are other ways of doing this that don't require notch filters! For example, you can inject a modulated narrow-band signal into your amplifier output such that it prevents the receiver from locking onto the video signal.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ok,

But if I would like to have more than one TVs connected on the same amplyfier output, but with different configuration on the permissive for viewing channels, then this would be a bad idea.
I can say that this is a good idea Jim and I have to admit that I never had a thought from the amp's side. Unfortunately this is not what I am trying to do.

Also: Nice photo in your site's home page.

Thanks.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think that you need to redirect your question so that you get some CATV experts in here. It has been 40 years since I was in that business, however I believe that there have been techniques developed for what you are trying to achieve.

Maybe keywords: encryption, encrypted video, scramble, pay to view, pay TV, etc.

Good luck

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Right with DaveF.

I know a fair amount about RF but what you are asking would be a major project. Yet, the CATV folks have been doing that for years. Even the cable installers have little "blocks" that can be put in the coax to prevent certain channels. Those can't be much. You MAY only need to block the FM carrier and the paritally suppressed video carrier to be fully effective without blocking the entire channel. That would only be two narrow-band notches that don't interact (much) with each other. Those could be done with helical resonators (Q = 1000 pretty easy).

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yhat seems good.

Please, could you be more analytical, because I don't have enough experience on this kind of issues?

Thank you.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

icarus1 wrote:
Yhat seems good.

Please, could you be more analytical, because I don't have enough experience on this kind of issues?

Thank you.

Though helical filters have high enough Q to filter one channel, they won't help you much to filter out an arbitrary combination of 9 channels. You would need 9 filters each separately bypass-able. Would be hard to make them not detune when switching the configuration.

BTW, I don't see how is this idea better than a jamming oscillator. Each TV output needs a separate set of filters. In fact, it would be sufficient to have a single oscillator per output.

IMO this project is nearly impossible for anyone asking how to do it. Pretty much like asking how to do a brain surgery.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I agree with gizmo. While your idea seems simple when you first describe it, it is factually NOT so simple. I would not quite say it is as complex as brain surgery, it is not far behind.

(1) The design of just one filter is challenging.

(2) To make it so it can be switched "on" or "off" adds even more challenge.

(3) To make it so any arbitrary conbination of filters can be used or not is a really big challenge. This is because each "on" filter interacts with the other "on" filters and each one shifts the frequency of the others. I might put a 0db )buffer) amplifier between each filter to reduce the interaction.

(4) It sounds like you want to manufacture this device. This device is going to require manual tuning by a skilled person. and with good test equipment. You really need the same or better equipment just to prototype this.

(5) From your messages, I could provide you with the names of a number of reference texts (Zvarev is my favorite and Dishal tells how to tune them) but I doubt that you have the background knowledge to even understand them.

As a result, I think you are trying to do something which is beyond your capabilities. Waaaaaay beyond. It may be a good idea, but having the idea does not automatically mean that it is possible to achieve that idea.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Jim, Do you mean Zverev?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ooops, yes, Zverev!

I've always had difficulty with that name, along with Chebychev (Tchebchev, Chebyshef, ???)

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Here's a thought; absolutely no idea how the *practical* side of this would work; consult your local RF specialist... but...

You have a small band of frequencies which are both narrow in terms of octaves, close in terms of octaves, and at a high enough frequency that sacrifices of small animals are involved in the design stage of suitable equipment.

If you mix the incoming signal with a suitable local oscillator - the same way a normal tuner works - you could recreate the image in the - say - 1 to 50MHz region, rather than 575 to 625MHz. Now your signals represent perhaps a tenth of the band, rather than one percent - a much easier design for the filter stage. (Though it still won't be easy). Then mix again to get the image back where you need it, sans your unwanted channel.

In fact, I suspect that you could design a *single* good filter and by changing the local oscillator, ensure that any of the channels could be placed on the critical frequency of that filter, simplifying the filter design at the expense of the oscillator design.

But frankly, I'd be tempted to use a bank of demodulators, take all the signals to baseband video and audio, and remodulate only the ones you want to pass through. That's going to work - perhaps with a little extra noise - and will use off the shelf components. Won't be too cheap though, a few hundred bucks perhaps.

Of course, this will all be academic when digital terrestrial TV hits your part of the world.

Neil

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Actually, there is a germ of a good idea in Neil's post.

Why not down-convert those UHF channels to a much lower channel? Simply disallow the converter to tune to the forbidden channels! This is pretty straight forward technology. No fancy filters, etc.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Can't help with the spelling for Chebyshev, I'm totally confused regarding this. Perhaps Svofski can help us out :)

Here is at least(what I believe to be) a picture with russian spelling: http://steiner.math.nthu.edu.tw/...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I have another idea.

I think that if there would be an amplifier for each channel image carrier frequency and an amplifier for each channel voice carrier frequency, then the TV tuner would be blocked up when the each time "forbidden" channel would be selected.

Am I right on this?

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Building an amplifier for each channel has exactly the same issues as building filters, only more so since you now have to consider noise issues too.

Unless you demodulate to baseband and remodulate, as I suggested above.

But I still don't fully understand what you're trying to do, other than in general terms; is this - for example - a group of receivers which must all have the same channel removed, or different channels, or just one receiver?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Neil,

I would like to know the way how to do this:

A group of receivers connected to the same antenna amplifier output, programmable via dip switches for viewing (or not) channels from 31 to 39.
Each receiver may be programmed to have different settings from the others.

I have already been late (4 days), to answer the question, if I could take the project. The deadline is been moved on Saturday morning. I know that if I catch the way I can do it, because I have all the tools I need to do my work.

But still the way is undefined for all of us, having this web convertsation and handshakeing ideas.

Thanks.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Does anybody knows any video signal UHF encoder/decoder chips?

If there was a chip doing this then there may be an encoder after the antenna amplifier converting the channels from 31 to 39 to lower frequencies. Then, each receiver will have a decoder for each channel. Each decoder will be chip selected by the microcontroller.

Just another idea...

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you can accept a baseband video output, then you could simply use one of the commercial tuner chips used in domestic TV's.

These are pretty much all processor driven these days... simply use the dip switch to select allowable channels. Piece of cake... You could even use a commercial tuner *module* with all the RF stuff done for you.

http://www.globalsources.com/man... - second google hit of around two million.

Suddenly, your project has gone from almost impossible without specialist equipment and major research/development effort, to remarkably simple... just from knowing what it is that you want.

Neil

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ok,

I said yes, which means that I assumed the project. There are some free channels channels near the 150MHz (much lower than 600MHz), that's why I said yes.

May God help me...

There is a time of 8 months from now to finish the research.

Also, I would like to thank you all for your priceless help on my mysterious questions.

See you.

Michael.

User of:
IAR Embedded Workbench C/C++ Compiler
Altium Designer

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

CountZero wrote:
Guess this is a school assignment?

One way to go could be to use high impedance transmission lines (implemented as microstrip lines) to replace the inductors.

Google a bit for that, I remeber philips(now NXP) had a good discussion of this technique for matching power amplifiers. I can try to find the specific appnote tonight.

Great advice Count. I would definitely look into different impedance factors you could also look into different RF filters to see if there are other specs that might be an easier solution. here are some specs that might help http://www.oscilent.com/catalog/Category/rf_saw_filter.htm

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

After 4 years it seemed like spam and almost got zapped but it's a useful site.

John Samperi

Ampertronics Pty. Ltd.

www.ampertronics.com.au

* Electronic Design * Custom Products * Contract Assembly

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think if I had to do this, I would try a 1/4 wavelength open circuit stub. A single stub should give at least 10 dB attenuation over that bandwidth. Of course the OP has not specified how much attenuation is required, despite several requests. A specan would be essential, as while you can cut coax off, it is damn impossible to cut back on.

Charles Darwin, Lord Kelvin & Murphy are always lurking about!
Lee -.-
Riddle me this...How did the serpent move around before the fall?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I used o make these at Austenna in Melbourne 30 +year ago.

Helical resonators top loaded and coupled through 1/4 wave lines with adjustable coupling capacitor.
Did n't have a spec an.. didnt have a network analyser.. had to do it with a swept frequency test bed.