Multi-Pole filter with opamp

[ethan169]

Ok guys,

Heres my question. Is it possible to configure an opamp in a multipole filtering? Im basically trying to filter a 5Mhz square wave into a sine wave, and would like to basically cut everything off from 5Mhz and up.

I cant remember ever actually seeing an opamp filter design with multiple poles but im sure there someone who has tried. I have a quad package to work with for now.

Thanks for any help!

Ethan

 

[Hero999]

Yes everything above 5MHz will be lost.

The 5MHz squarewave contains a 5MHz sinewave plus a load of higher frequency harmonics, all you need to do is remove the harmonics and you'll have your sinewave.

Why are you using an op-amp? A passive LC resonant circuit is all you need, simple use a 1nF capacitor in parallel with a 1µH inductor.

 

[ethan169]

Yes everything above 5MHz will be lost.

The 5MHz squarewave contains a 5MHz sinewave plus a load of higher frequency harmonics, all you need to do is remove the harmonics and you'll have your sinewave.

Why are you using an op-amp? A passive LC resonant circuit is all you need, simple use a 1nF capacitor in parallel with a 1µH inductor.

Thats a great idea! I was using an op amp because I am building the sine wave gen out of an exisitng PCB that has an OP462 on it and was going to use that for filtering. Ill try the LC resonant circuit though.

I let you know how it works out

Thanks again!

-EThan

 

[ethan169]

Just a quick question. How did you get those values for the L and the C?

-Ethan

 

[Hero999]

[latex]F_0 = \frac{1}{2 \pi sqrt{LC}}[/latex]

 

[MikeMl]

You can easily get two poles per OpAmp when building active filters; three poles if you are building low-pass or high-pass.

To make an active LPF with an OpAmp with a cutoff at 5Mhz, you would have to use an OpAmp that has a Unity Gain Bandwidth of 50+Mhz, not the sort of thing you likely have laying around.

Since you are starting with a squarewave, how much suppression do you need of the third harmonic? A two-pole filter will only have ~20db suppression of the third harmonic. If you are generating an RF signal which ultimately reaches an antenna, you should have like 60db suppression of the third harmonic, which takes a much more elaborate filter, beit LC or active, more like 6 to 8 poles.

 

[ethan169]

[latex]F_0 = \frac{1}{2 \pi sqrt{LC}}[/latex]

Right this formula i am familiar with. But after thinking about a parallel LC it didnt make sense to me. Nonetheless i tried it and it didn't really work properly. Wich makes sense if you think about how an L behaives at DC and AC and the same with a C, this is creating a notch filter wich is not what i was looking for. So connected in series like an RC circuit this works beautifuly! Super crisp stable sine wave.

Thanks again for the ideas im glad i was able to do this without using the on board opamp!


-Ethan

 

[Hero999]

So you used a series LC circuit not a parallel LC circuit; is that right?

That's a perfectly acceptable way of doing it.