Tell me more ×
Electrical Engineering Stack Exchange is a question and answer site for electronics and electrical engineering professionals, students, and enthusiasts. It's 100% free, no registration required.
up vote 4 down vote favorite

I would like to design and build a tunable lowpass filter. I would like to adjust the cutoff frequency from 1 MHz to 500 MHz. Requirement on the resolution would be something like 10-50 MHz, obviously, it's not that critical.

I am going to send a programmable clock signal through this filter and want to reduce the high frequency harmonics. I am only using the clock signal as a simple way to get a programmable sinusoidal signal, hence the need for the filter. I was considering using the ADF4110, ADF4350, or one of the Silicon Labs Si570/571 as the programmable clock. I was told that a tunable lowpass filter over this range may be tricky, so I thought it best to consult the StackExchange. I am open to any suggestions.

Ultimately, what I'm trying to accomplish is get a sinusoidal signal that will sweep from 1-500 MHz.

share|improve this question
2  
The question is too broad. Why do you want to reduce the harmnonics? The clock shape will become closer to a sinusoid if you do and then it is not suitable as a clock. How much reduction to you want since this determines the complexity of the filter (number of poles, LC vs RC, etc)? Is the filter to be manually adjustable or by a digital control signal? At these frequencies, although the design is not tricky, the actual layout of the filter is since stray capacitance and improper grounding can negate the filter action. Please provide more details. – Barry Oct 2 '12 at 21:02
@Barry, thanks for the comments! I'll update the question. – macduff Oct 2 '12 at 21:21

2 Answers

up vote 5 down vote accepted

Tunable filters over this sort of range are really, really difficult to do. For reals.

In high end network analysers you will find a special type of filter called a YIG filter (low end ones often don't use them because of their price) or a sweep oscillator based on a YIG VCO. Your best bet would be to just make 10 lowpass filters and switch them in with relays (or use relays to change the capacitance). You'll note that older signal generators do a lot of clicking when you crank the frequency up. If all of the HP RF boffins couldn't work out a good way of making tunable filters then I think you may struggle.

But to be honest, I am not sure why you are doing this to generate a sine wave. Just go DDS! Any modern signal and/or function gen will use DDS for this frequency range. You can buy off the shelf chips that go way above what you need; the AD9914 does <0Hz to 1.4GHz sine waves.

share|improve this answer
thanks for the great answer!!! I would just send an email or something but you don't have one on your profile. Anyway, thanks for your input on this problem! – macduff Oct 3 '12 at 13:51
1  
Woah! That's an expensive chip. – Rocketmagnet Oct 3 '12 at 16:03
1  
In a spectrum analyzer or network analyzer, I'm pretty sure the YIG component is usually a VCO, and usually YIGs are used for higher frequencies like 500 - 50,000 MHz. The output of the VCO is used to up- or down-convert the input signal to a fixed IF range, where it can be filtered using a fixed filter, then finally down-converted again for detecting and measuring/displaying the response. Which is all good... – The Photon Oct 3 '12 at 16:33
But if OP had a 1-500 MHz VCO with clean output, he would already have his real problem solved and wouldn't need the filter he's asking for. – The Photon Oct 3 '12 at 16:34
Also, here's a look at why, even at $200-$300, the DDS solution is probably lower-priced than the YIG solution: microwaves101.com/encyclopedia/YIG.cfm – The Photon Oct 3 '12 at 16:40
show 2 more comments
up vote 1 down vote

Some companies like Pole/Zero make tunable filters for such applications in the bands you're working with. Cant speak to the effect on your square wave of cutting off the higher harmonics.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.