frequency clamp

[yoyoman]

I am looking for a circuit to help me clamp off a 5 volt square wave at 1800hz. I have tried a few different things but I cant get it to work properly. Any help would be greatly appreciated.Yes I am a little new to this.. LOL

 

[electricpete]

clamp? Do you mean that you want a low-pass filter with cutoff in the neighborhood of 1800hz?

 

[yoyoman]

yes, I just need to allow everything up to 1800hz through and nothing past that frequency

 

[IRstuff]

There's a serious flaw in your problem. A square wave is mathematically an odd function whose Fourier expansion involves the odd harmonics of the fundamental. This means that if you really want an 1800 Hz square wave, you need to pass frequencies up to at least 9000 Hz, otherwise you'd wind up with an 1800 Hz sine wave.

This, of course, means that you're passing lots of other sine waves up to 9000 Hz.

Otherwise, you're looking for a low-pass filter, but bear in mind that the cut-off frequency for a single pole filter is simply the frequency at which the response drops to 50% of that of the passband. A more complicated filter will introduce phase delays, which means that your square wave will get distorted in any case.

You need to review what you're trying to do. Perhaps a phase-locked loop might make more sense?

TTFN

 

[yoyoman]

I truly appreciate the replies. I must admit though that I am getting a little lost. Do you guys know where I can find a starting point for this thing. Basically this a mass air sensor varying freq. from 25 to 1800hz and I am just trying to build a circuit to clamp it off at this 1800hz freq. It is a 5 volt square wave. I was told it was as simple as a resist and capacitor but I am beginning to think that my co worker was wrong.

 

[IRstuff]

Your co-worker was correct in that a resistor/capacitor divider provides a low-pass filter function, but the response function falls at only 6dB per octave, so there would be a considerable amount of bleed through. More complicated filters can be devised, but it could mess up your square wave.

Is the varying frequency due to the actual measurement, e.g., is it a frequency modulation proportional to the measured quantity? Does it need to stay a square wave? If you have some sort of demodulator than converts the frequency to a voltage, it may be easier to "clamp" the frequency after the conversion, since that could be as simple as a zener diode clamp.

TTFN

 

[yoyoman]

this sensor varies frequency based on air passed by so it starts at 25-50hz and as engine speed increases so does the frequency, up to 1800hz. anything above that and it hits cutoff so i need to keep it at 1800hz.it does need to stay a square wave though.

 

[pebe]

One way would be to monitor your sensor with a frequency-sensitive device (eg. a 555 wired as a monostable triggered on succesive pulses). When this device sensed 1800Hz or higher, it would switch in a 1800Hz oscillator in place of the sensor output.

 

[melone]

Why would you want to modify the Mass Air Flow sensor? That will have serious impact on your engine performace, not to mention totally destroy your emissions. Intentionally causing increased emmissions can get you in a lot of trouble.

 

[StanStep3]

Hi,

Here's an idea. Use a low pass filter (see Maxim for a good assortment of filter types and pole numbers) to make your 1800 Hz cutoff frequency. Then use further amplification and a comparator to re-create your square wave pulses after the filter.

Good luck,
Stan

 

[nbucska]

Measure the frequency digitally: If you just need one,
you can make discrete TTL or CMOS logic or you can use
a microcomputer e.g. a PIC or a DOS STAMP.

If U need many pieces, use a cheaper micro. You can be
under $10 excluding PCB . <nbucska@pcperipherals.com>