FM monolithic IC??

[Learner]

Hi guys,
I am looking to build a frequency modulation synthesizer and was wondering if anything could suggest any IC for doing this?

What I would like to do is actually to use this as a pitch shifter eg. input an audio signal from one end and a frequency of 2 hz, 4 hz etc... on the other in order to shift up the frequency.

I have found signal generator chip such as the Max038 and the XR2206, does anyone have experience with these chips, thoughts?

Are there any other alternatives besides these?

Thanks!

 

[Nigel Goodwin]

As usual, you need to give more details - the 2-4Hz frequency shift sounds like you are looking for a feedback reducer?. Oscillator chips would be no good for that!.

 

[Learner]

Hi Nigel,
I am looking to modulate/shift the frequency by a number of octaves which is still within low(audio) frequency spectrum < 20Khz, I am looking for an IC that can multiply the frequency of 2 AC signals 1 would be the audio and another coming from a OSC generating 2hz, 4 hz and 8 hz etc....
Hoping to get a shifted audio signal at the output in higher octave as a pitch shifter, but I donno what I can use to achive this.......

That's why I am thinking if there is a FM IC that can do this, where you can input 2 AC signals and multiply both frequency to generate the shifted pitch at the output.

Can I use a 4 quadrant multiplier?

 

[Nigel Goodwin]

Things don't work that easily :cry:

One way you could do it is to convert the audio signal (voice, whatever) to an RF single sideband signal. Then convert it back to audio using a slightly different RF carrier insertion oscillator.

By varying the frequency of the carrier insertion oscillator you can vary the pitch of the resultant audio signal.

 

[Learner]

Hi Nigel,
I dont quite understand why it is necessary to modulate the audio signal then change the carrie frequency to get the pitch shift, if I use a 4 quadrant multiplier eg. 2hz AC x 2 hz AC wouldn't I get 4 hz AC at the output of the multiplier?

Thanks again for your help!

 

[Nigel Goodwin]

Hi Nigel,
I dont quite understand why it is necessary to modulate the audio signal then change the carrie frequency to get the pitch shift, if I use a 4 quadrant multiplier eg. 2hz AC x 2 hz AC wouldn't I get 4 hz AC at the output of the multiplier?

You will get the sum and the difference, plus both original signals (although reduced due to the double balanced mixing). The reason for using a much higher carrier is so you can filter it to get just the part you want - rather like image reception in a radio, a higher IF frequency gives much better image rejection.

If it was simple to do it would be commonplace and cheap!.

 

[Learner]

Hi Nigel,
In order to implement this method, would I be on the right track to use 2 x 4 quadrant multiplier in series. First one for modulating the signal eg. signal modulated by a 2 hz carrier then run the output into the next 4 quadrant multiplier and then modulate it with an invereted carrier of 4 Hz to get the pitch shift?

 

[Nigel Goodwin]

Hi Nigel,
In order to implement this method, would I be on the right track to use 2 x 4 quadrant multiplier in series. First one for modulating the signal eg. signal modulated by a 2 hz carrier then run the output into the next 4 quadrant multiplier and then modulate it with an invereted carrier of 4 Hz to get the pitch shift?

You could certainly try it, but I'm doubtful it would work?.

 

[Learner]

Hi Nigel,
Any ideas on how I should implement the carrier method? :?

 

[Nigel Goodwin]

Hi Nigel,
Any ideas on how I should implement the carrier method? :?

The biggest problem is the filter (to remove the unwanted sideband), these are expensive - and generally only wide enough for speech use.

 

[Learner]

Do you know what the Q size would be required for this task? For a wide bandpass filter with low a Q I could probably use a hi pass in series with a low pass filter, and place them between the first and the second 4 quad multiplier?

 

[Nigel Goodwin]

Do you know what the Q size would be required for this task? For a wide bandpass filter with low a Q I could probably use a hi pass in series with a low pass filter, and place them between the first and the second 4 quad multiplier?

The Q depends on the bandwidth required, but the slope needs to be VERY steep, expensive crystal filters are commonly used.

 

[Learner]

mmmm.............. sounds like a job for dedicated state variable filter for each hi pass and low pass filter, it is capable of implementing Q of up to 500. Time to put theory to the test, I just hope I can get down to the bottom of this and learn my lesson on why it would or wouldn't work.


Wish me luck Nigel!


BTW, if I was to try to do this with a uController. How would I go about doing this?

AD-> fill register -> DA at different clock rate?

I think latency will be a big problem.....

 

[Learner]

Hi Nigel,
Having spent some time thinking about the modulation method, I came to the conclusion that this method would not work at all..... Since if I was to use a different frequency for demodulation, wouldn't I miss the carrier completely and get no output?

Otherwise we would be getting multiple channels on the same dial on the FM radio, except that each signal would have different pitch as oppose to the one that's occupying the broadcasting frequency?

Thoughts?

 

[Nigel Goodwin]

Hi Nigel,
Having spent some time thinking about the modulation method, I came to the conclusion that this method would not work at all..... Since if I was to use a different frequency for demodulation, wouldn't I miss the carrier completely and get no output?

Otherwise we would be getting multiple channels on the same dial on the FM radio, except that each signal would have different pitch as oppose to the one that's occupying the broadcasting frequency?

That's because it's FM, not SSB (Single Side Band) - if you've ever used an SSB receiver you will have heard the effect you're looking for frequently :lol:

 

[Learner]

Hi Nigel,
THANK YOU SO MUCH!!!! For setting me straight!!!!!

I really really appreciated it!!!!

Thank god there are people like you around!!!

Donno what I would do with out people like you, and the fact that you have actually experienced such an effect with AM. I am definitely going to try and build a prototype circuit and see how it works out, may be even tune into some AM radio to see if I can try to hear the pitch shift.

Thanks again!!!!!

regards,
Scott

 

[Nigel Goodwin]

Donno what I would do with out people like you, and the fact that you have actually experienced such an effect with AM. I am definitely going to try and build a prototype circuit and see how it works out, may be even tune into some AM radio to see if I can try to hear the pitch shift.

Not AM, SSB, you can demodulate SSB on an AM radio by adding a CIO (Carrier Insertion Oscillator) - and by varying it's frequency (or altering the radio tuning) alter the pitch of the demodulated audio.

BTW, SSB can be either upper sideband, or lower sideband, if you demodulate using a carrier the wrong side you get totally unintelligible audio (it's a crude way of scrambling a phone transmission).