Sweepable high pass filter schematic

[John Lyons]

> Is the impedance partially set with the 110K?

Blunt-hammer analysis:

Caps are open at DC and short at infinite frequency.

So at DC, and very low frequency, C1 is open, input impedance is infinite. (We may not care much, since this is the cut-band of the filter.)

At very high frequency, C1 C2 are dead short. The input must drive R2 110K to ground. Does it have to drive R1? This is a bit tricky. The amp is unity gain. With C1 C2 short, the signal at pin 6 (out) is the same as at pin 3 (+in). Therefore the signal voltage is identical on both ends of R1. No current flows. Effective R1 impedance is infinite. Effective signal input impedance is just R2 110K to ground.

So falling from infinity to some point, then 110K above that point.

Where that point be? Considering how we picked these Cs and Rs, probably around 100Hz. Probably slightly above 110K at 100Hz, rising at lower frequency.

Source loading is one issue. However this type of filter can be sensitive to source impedance. Generally they are optimum for zero-Z source, and sometimes have quite large errors if source resistance is not much-much-much lower than RC impedances.

The Idiot Assistant tells me 110.3K at 1KHz, 113K at 300Hz, 136K at 100Hz, 288K at 30Hz, 800K at 10Hz, and who-cares below that? (About 8Meg at 1Hz.) IA also shows that 50K of source drops the pass-band (obviously) and makes corner less-sharp, but nothing major, and not a problem for musical uses.

So impedance isn't an issue here then correct?
The values I have decided on sweep between 43hz and 1khz

[Superfly76]

I'm looking for a sweepable high pass filter from about 50hz to 500hz.
12db cut or more.
Anyone have a link for something like this?

I'm using this for acoustic guitar live.

I'm bringing this thread back from the dead because I'm looking for this exact thing. I see there was some progress in the thread, but not sure exactly how to do this.

Looking for a 2nd order HPF, sweepable from 50Hz to 500Hz. Using it for acoustic guitar. Really surprised that there isn't something like this on the market.

[PBE6]

Just goofing around with the Sallen-Key high pass numbers. Using C1 = C2 = 33nF and R1 = R2 = 15k with a 100k dual gang pot (audio taper preferably) in series with the resistors will get you in the ballpark. Add a buffer to the front end and you're good to go. The Q factor is about 0.5 which is low, so the roll off takes a few extra Hz to get down to -12dB/octave, but that might actually make for a more natural sound.

The alternative to a dual gang pot is to use a 4P6T rotary switch (or similar, a 3P6T might even work) to switch matched capacitors in and out of the circuit simultaneously. If you use a switch, you can sharpen the Q factor by adjusting the resistor ratio (making the feedback resistor smaller increases Q), but I doubt you want to add much resonance on the bottom end with an acoustic guitar.


Sent from my iPhone using Tapatalk

[Superfly76]

I like it! Thanks PBE6. What would be a good opamp to use in this circuit? Also, if my pickup is active, can I forgo the buffer?

I'd like to do a very hi impedance buffer if one is needed, as I'm dealing with piezo pickups here. 

[PBE6]

The Sallen-Key filter is a bit sensitive about the input impedance. You may be able to get away without a buffer if your guitar output is active and low impedance, but if it's higher than expected (or you decide to try using a passive guitar later) the filter may not work quite as designed. I would suggest adding one in.

A FET-based dual opamp like the TL072 would be a good choice, or you could try a slightly fancier version like an OPA2134. Either way, one dual opamp is all you need for this circuit. Just make sure the biasing resistor is large, something like 2.2M or higher, and you shouldn't experience any tone sucking.


Sent from my iPhone using Tapatalk

[Superfly76]

Very good info. Thanks.

I've brushed myself back up to speed on input buffers. I've found a handful of good ones that will work well for the front end of this little highpass project box. Still in research mode, I'll post the design here as it progresses.

I'd really like to do a jfet or mosfet based buffer, as I much prefer the sound of those over opamps. Less noise. I also like their response when they are driven hard.

I would also like to swap out the opamp on the sallen key filter for a jfet or mosfet based one. I realize I'll need a few more components to make it work, but I think that would be worth it.

I found a design here that uses jfets in the buffer and the filter section. personalpages.tds.net/~fdeck/bass Francis J Deck has designed an acoustic bass preamp and highpass filter that is built with jfets. It has an input impedance of 10M (great for piezos).

He has made his design open and has posted schematics. Will post my findings here when I figure out how to upload and insert an image.

I'm also thinking of adding a notch filter to this little project to remove feedback in a live situation.

[PBE6]

A discrete JFET or MOSFET input buffer would be a fine addition. I would still recommend using an opamp for the buffer within the filter though, it's use is strictly utilitarian here and doesn't add to (or take away from) any mojo. The other half could be used to buffer the power supply, so there's no wasted parts either.


Sent from my iPhone using Tapatalk

[Superfly76]

Cool. I was wondering that myself. Is there any particular opamp that would be best suited to this application? I have a few laying around, but would be willing to order something else if it were better for this build. Also, I'm thinking of adding a steep notch filter for anti feedback. Was wondering if the second half of the opamp could be used for that instead.

Out of curiosity, how would one use an opamp to filter power?

Edit:

Whoops, I see you already answered my "best opamp" question above. I have a bag of TL072s from earlier projects. I'll check out the other one.

[samhay]

>I'd really like to do a jfet or mosfet based buffer, as I much prefer the sound of those over opamps. Less noise.

If your op-amp buffer had more noise than a discrete buffer, then you have likely done something wrong.

>I'm thinking of adding a steep notch filter for anti feedback. Was wondering if the second half of the opamp could be used for that instead.

Yes. Look up 'bootstrapped twin T' filters, such as here:  http://geofex.com/Article_Folders/EQs/paramet.htm 

[Superfly76]

Awesome link! This is turning into a really fun project.

Noise was the wrong word to use for the opamp based buffers. I prefer the feel and sound of jfet based buffers over opamp ones. It might be more mental than anything. I've read that the response is more similar to the front end of a tube amp. I might just have some bias against opamps. Willing to change my tune if I'm off base or fell for some tone "snake oil".