Filter design

[audioguru]

Your 8-pole highpass filter is biased correctly and should work. Except the TL074 opamp does not go high enough.

EDIT:
Filter Pro says to use high speed opamps, especially the 4th one. The TL074 is too slow for this filter.

 

[Electronman]

Your 8-pole highpass filter is biased correctly and should work. Except the TL074 opamp does not go high enough.

EDIT:
Filter Pro says to use high speed opamps, especially the 4th one. The TL074 is too slow for this filter.

Thanks,

Can you let me know if I will have a good carrier at the output of the above filter?
I do not know if that 8pole filter is good for 10kHz of bandwidth?
I fear of phase shift at the output of the above filter?
I do not know any high speed opamp. what are they? Do you think LF411 and LF356 are high speed opamps?

P.s did you care of the single power supply and the 1k biasing resistor?? No need to capacitors for them?

 

[audioguru]

Thanks,

Can you let me know if I will have a good carrier at the output of the above filter?

The filter will reduce the amplitude of the carrier a little as shown on the frequency response in the simulation.

I do not know if that 8pole filter is good for 10kHz of bandwidth?

The filter passes frequencies above 40kHz like a piece of wire (the TL074 goes up to 100kHz) so does not affect the bandwidth.

I fear of phase shift at the output of the above filter?

It will work a lot better with the high speed opamps it needs.

I do not know any high speed opamp. what are they? Do you think LF411 and LF356 are high speed opamps?

The LF411 and LF356 are not high speed. Look at what is needed on the Filter Pro simulation then look at their datasheets. I have never used high speed opamps.

P.s did you care of the single power supply and the 1k biasing resistor?? No need to capacitors for them?

Of course a capacitor is needed at the junction of the two 1k resistors to ground. Try 10uF.

 

[Electronman]

The filter will reduce the amplitude of the carrier a little as shown on the frequency response in the simulation..

I do not know how to read the values on the software graph.

The filter passes frequencies above 40kHz like a piece of wire (the TL074 goes up to 100kHz) so does not affect the bandwidth..

If so maybe a one pole filter does work? So why we use the higher order filters?

Of course a capacitor is needed at the junction of the two 1k resistors to ground. Try 10uF.

Why we have to use that capacitor there?
Do I need to put that caps on any voltage divider based upon resistors?

By the way I do not know if the tolerance of the resistors in a voltage divider changes the operation of any opamp or not??

 

[audioguru]

You want a very sharp filter to cut the lower sideband. It has no effect on the upper sideband so has no effect on audio bandwidth.
A one pole filter will not remove much of the lower sideband and will also have no effect on the upper sideband.

The two 1k resistors bias the opamps. The bias voltage must be a low impedance so a capacitor to ground is used.

5% resistors are usually used so their small tolerance has a small effect on the operation of the opamps but only when their output level approaches max.

 

[Electronman]

You want a very sharp filter to cut the lower sideband. It has no effect on the upper sideband so has no effect on audio bandwidth.
A one pole filter will not remove much of the lower sideband and will also have no effect on the upper sideband.

The two 1k resistors bias the opamps. The bias voltage must be a low impedance so a capacitor to ground is used.

5% resistors are usually used so their small tolerance has a small effect on the operation of the opamps but only when their output level approaches max.

Sorry, I am a beginner so need more help ans more explanation to learn things

Why a one pole filter behaves so?

Why the voltage bias must be low and how a caps makes it lowe?

 

[audioguru]

A one pole filter is very simple so it filters poorly. For a highpass filter the amplitude is 0.707 at the cutoff frequency and is only 0.5 at half the cutoff frequency.
A filter with 8 poles has its amplitude at 0.02 at half the cutoff frequency so it reduces the low frequencies much better.

The capacitor at the bias voltage filters any signal away.

 

[Electronman]

Ok I made the above 'Sallen –Key' filter in real with TL074.
Unfortunately it did not work at all (I do not know the reason).
So I went to design one filter based upon 'MFB Single-Ended'.
Now it does work but the output is gain seem to be less than one?

The bellow is the 'MFB Single-Ended' circuit designed by Filterpro and simulated by Pspice at 2kHz and then at 10kHz.
Because I am not able to put more than 3 pictures in a post I will post the FFT of the Pspice at 2kHz And 10kHz modulating signal in the bellow post.

 

[Electronman]

Here's the FFT of the signal at 2kHz, 10kHz.

How to improve the output signal of the filter?
Can I change the values of the filter I do some change in filterpro to get more gain from the filter?

The lower sideband as you can see exist at 2kHz modulating signal so what do do?

 

[Electronman]

Is the waver method a better way to go for instead of filtering method?

 

[audioguru]

Is the waver method a better way to go for instead of filtering method?

Ham radio operators use crystals to cut a sideband. Then the bandwidth is so narrow that you can't understand the chipmunk voices.

I don't know why your circuit is biasing capacitors at the bottom. Capacitors do not need a DC bias voltage.

 

[crutschow]

Your 39KHz fillter is only 1KHz about the lower sideband with a 2KHz modulation frequency. That's why so much of the lower 2KHz sideband gets through. You should go to a 40KHz filter. That will reduce the lower 2KHz sideband and still leave significant carrier feedthrough as you desired.

You could also go to a filter with a faster rolloff such as a Chebyshev. That will give you some ripple in the passband response, but that should be tolerable.

What is the waver method?

 

[Electronman]

Ham radio operators use crystals to cut a sideband. Then the bandwidth is so narrow that you can't understand the chipmunk voices.

I don't know why your circuit is biasing capacitors at the bottom. Capacitors do not need a DC bias voltage.

I have several 40kHz crystals but I do not know if they can be used to create a high pass filter with 10kHz of bandwidth?

Which part are you meaning? Anyway the above circuit is exactly suggested by the Filterpro software.

 

[Electronman]

Your 39KHz fillter is only 1KHz about the lower sideband with a 2KHz modulation frequency. That's why so much of the lower 2KHz sideband gets through. You should go to a 40KHz filter. That will reduce the lower 2KHz sideband and still leave significant carrier feedthrough as you desired.

You could also go to a filter with a faster rolloff such as a Chebyshev. That will give you some ripple in the passband response, but that should be tolerable.

What is the waver method?

It does not matter using 39kHz or 40kHz anyway. I want to transmit my voice or music so the modulating freq changes frequentely.

Go here for the waver method:
Single-sideband modulation - Wikipedia, the free encyclopedia

 

[audioguru]

Don't you know that old fashioned Amplitude Modulation (AM) is usually very narrow band, is full of static interference and interference from anything that operates near its frequency like compact fluorescent light bulbs and anything that makes sparks?

Wideband FM modulation is used for voice and music today.

 

[crutschow]

It does not matter using 39kHz or 40kHz anyway. I want to transmit my voice or music so the modulating freq changes frequentely.

Go here for the waver method:
Single-sideband modulation - Wikipedia, the free encyclopedia

I does matter whether the filter is 39KHz or 40KHz since it affects the suppression of modulation sidebands that are close to the carrier.

The Weaver (not waver) Modulator is a rather complex method of generating a SSB signal, generally implemented with digital circuits. It's much more difficult to implement then the simple filter method you are attempting.

 

[JimB]

After 3 pages of this thread there is one question I must ask...

Why on earth are you wanting to create a 40khz single sideband, full carrier signal?
Not only that, but with a modulation frequency response extending to 10khz?

What possible application does it have? I seems like pure madness.

Please enlighten me.

JimB

 

[Electronman]

After 3 pages of this thread there is one question I must ask...

Why on earth are you wanting to create a 40khz single sideband, full carrier signal?
Not only that, but with a modulation frequency response extending to 10khz?

What possible application does it have? I seems like pure madness.

Please enlighten me.

JimB

Well several days ago I read a paper about creating audio from ultrasound. So I want to design a circuit to do a similar task.
I have to filter out the output to create SSB or VSB somehow.

You guys, please let me know If those high pass filters which I designed (the above pictures) cause the output to be VSB if they do not remove the toltal lower side band?
If I am can not to create a good filter for SSB then I want to create a filter to truncated the lower sideband of the dsb with carrier signal. Any suggetion?

 

[Electronman]

JimB,

If you are interested, This night a met a Bat in my room while implementing the ultrasonic device.

It was a nice night. I was able to invite a bat inside my room with ultrasonic, lol

What is the Echo freq of Bats really?
Is there any circuit to call bats? ( I do not talk about bat detectors)

 

[Sceadwian]

Bat echo location is upwards of 80k or so I think.