High-pass filter for a beginner

[ahtinyk]

Hi all!

Let's start by mentioning that I'm all new to circuits and effects, as well as these forums. I tried searching here about this topic, but didn't really find answers for my questions.

I'm planning to buy a Micro POG octave pedal for my fender jazz bass. I'd love to use it for sub octave, but I'd need to filter off notes below B, meaning everything below 30Hz. I found some graphics about the circuit, but I'm wondering if it's as simple as it sounds? Let's say I follow the image below with a 115 Ohm resistor and a 47uF capacitor. Would I get the results I want, or do I need to take notice of anything else?




I'm sorry if there's a thread about this, that I didn't find. I hope I get some answers ^^ If it gets too complicated, I'll probably start with something simpler (if it can get any simpler).

[Granny Gremlin]

HPFs are not brick wall devices; they don't stop everything below the target freq; there is a rolloff.  That target freqq is actually what;s called ann F3 point, which means that it is not the point wherethe filter starts to work, but rather the point at which it has already cut the signal by 3db (the smallest anmount of gain/volume change detectable by the human ear).  The filter you are looking at attenuates at a rate of 6db per octave, so the filter kicks in at 37.5 Hz.  6db is actually very slow (1st order filter; they get steeper from here, at the cost of complexity), so if you are trying to make a rumble filter you need to either use a higher order filter or more the F3 point an octave upwards.

Another thing to consider is that a purely passive filter like this will be vulnerable to various interactions with the devices both before and after it (loading) thus altering the response. 

I was looking at a similar project (I have a Gibson EB3 bass with the infamous mudbucker pickup - love it but it is a handful in terms of low frequency output) and have been using an MXR 10 band eq pedal to cut 30 hz - waste of pedalboard space.  Haven't come up with anything I am happy with (have something on the go on my breadboard), or come across anything already figured out that suits my needs (was hoping to make this a rumble and noise filter; HPF and LPF)... kinda on the shelf lately as I work on a few other things.

[ahtinyk]

Thanks for the quick reply.

I guessed it wouldn't be a brick wall device, but if I understood what you wrote, then yeah, the effect is quite slow. I wouldn't mind if it wouldn't affect the lowest notes perfectly, but if it would reduce them even halfway it would be nice. Anyways, it's an effect I don't really want to spend hundreds to, so I thought it could be a cool project.

I guess I'll have to study the topic a bit more and see if it's worth trying. Thanks again!

[aron]

I thought the micro pog was digital?

[ahtinyk]

I thought the micro pog was digital?

Is that another point I should know about? 

[Granny Gremlin]

I guessed it wouldn't be a brick wall device, but if I understood what you wrote, then yeah, the effect is quite slow. I wouldn't mind if it wouldn't affect the lowest notes perfectly, but if it would reduce them even halfway it would be nice.

Nowhere close to halfway (unless you move the F3 point up or increaase filter order as mentioned).

[PRR]

> a 115 Ohm resistor and a 47uF capacitor

If you put a 115 Ohm load in the guitar chain, signal will be incredibly weak. Guitars (basses) don't drive anything under 10,000 Ohms at all well, and 115 is much-much-much smaller than 10,000.

If you put these values in your LoudSpeaker cable (which may be "too late", sub-bass already mudded-up the amplifier), the 8 Ohm speaker totally swamps the 115 Ohm nominal resistor, and your real bass-cut is abut 15 times (4 octaves) higher, 450, midrange.

> I'd need to filter off notes .... below 30Hz

Why?

Yeah, I know the racket of a 20Hz flap-flap-flap.

Knowing what the real issue may be might guide suggestions. (If it is speaker-flap, a vented box may need 3rd-order or sharper filter, which is a precision active filter or nasty coil+cap design.)

However if they are selling this POG to fools like me and you, maybe this POG has some in-built subsonic filter? (I really do not know.)

[bluebunny]

I think Aron might be right, judging by the SMT postage-stamps on the board:



I'm guessing the biggie is a DSP.  Hard to see the markings in this picture, but a gutshot of a regular POG clearly shows a similar chip which is a DSP.

[ahtinyk]

I think Aron might be right, judging by the SMT postage-stamps on the board:



I'm guessing the biggie is a DSP.  Hard to see the markings in this picture, but a gutshot of a regular POG clearly shows a similar chip which is a DSP.

So it being digital affects this whole thing how?

[GibsonGM]

It is being processed rather than 'shaped' by audio components, as we do here.   The audio is converted to 0s and 1s, made digital...and then MATH is done on the signal!   The signal has been broken up into many many levels, rather than one wave. 

The analog stuff we do deals with the actual signal, that wave, and rather than having discrete levels we work with the whole thing.  Usually (at least for us) that is a good thing, but in cases like this - where you really want to zero in on a small frequency range - that 'all inclusiveness' (for lack of a better term) makes it extremely hard to zero on the way you seem to want to.

Look up "HPF" in wikipedia or on an electronics tutorial site...you'll note this stuff is all curves, angles...no straight PASS NO PASS.   Digital can come close to a true slam-bam cut, but our components can't (not without extreme filter calculations, anyway!).   It is just because of the physics behind how capacitors, resistors and other components operate.    And, to get that kind of control over a signal is pretty complicated (but obviously CAN be done with work)...

[Granny Gremlin]

Guys, guys, the POG being digital is irrelevant - you can put a passive filter before or after it (read the OP: he never said he wanted to mod it). After it would make more sense if the concern here is to filter out excessive sub bass that the POG is capable of generating which WILL eat up headroom and cause mud (to answer PRR's question there); a rumble filter.  After also works well as the POG output is surely bufferred so loading won't be a concern on the input end (output, so far, is another matter) like it would be right out of the guitar.  If there are any other pedals in the chain after this, one thing to consider is that they may already have such filtering so this may be a redundant exercise.

Anyway, PRR makes a great point about the parts values - I was addressing other issues and didn't even pay attention to that.  Ahtinyk, go find a guitar HPF calculator online (there's plenty), put in 10k for the R and 60 Hz for the F (you won't miss that 6-9 DB of fundamantel low E to B - your bass doesn't actually produce very much of it anyway, tho the POG will) and see what you get for the cap.  If the cap value is insane try playing with the R value to see if you can bring the cap value down to earth. 

[GibsonGM]

Seemed like he asked how it operated, in a way, but maybe not    

If I were doing what you've now clarified the OP as wanting, I'd make it active, but that's just me...unless you make the R big, you'll have loss...

[Granny Gremlin]

It was ambiguous I'll give you that, but "I'd need to filter off ... everything below 30Hz"  is his only stated requirement and I inferred the rest (tragic love affairs with Mudbuckers help understand the need).  If all he wants is a rumble filter, passive is a workable option.  Tone shaping or multiband or boost as well as cut would be another matter.

That said, since the rumble filter would need to be on the output end of the POG (the superbass that needs to be tamed don't exist before the suboctave generator) it might be possible to hack into the traces near the output ( worst case, at the jack) and splice it in there.  Looks like there's easily enough room for a switch (stomp or otherwise) with a resistor and a cap hanging off it inside that thing. Though ideally the filter would disengage with the octave effect.  In that case it should be possible to splice it in off the stomp switch, depending how they wired it... though likely this wouldn't be very pretty and a royal PIA to figure out and then especially to execute. 

Without modding the POG at all it would be easy enough to put in an FX loop (no blend required because the POG has that already) in series ahead of the filter in the same enclosure; connect the POG to that loop and you have instant on/off (true bypass) for both the POG and the filter.  Still all passive. Would fit in a 1590A or G.

[ElectricDruid]

You should have a read of the ESP pages about filters:

http://sound.westhost.com/dwopa2.htm#filters

There's an example of a 2nd order (12dB/octave) highpass filter on that page, with equations to adjust the values. It's currently got a -3dB point at 112Hz, so doubling the two resistors will lower that by an octave (to 56Hz) which is in the right ballpark. You could then tweak by ear, using different resistor values (but trying to keep the 2:1 ratio) or tweaking the two cap values (and keeping them both the same).

The worst part would be getting the dual supply for the op-amp, but that's just two resistors and a cap for a "virtual ground". There's loads of stuff on here about using a virtual ground with op-amps.

HTH,
Tom

[GibsonGM]

ESP has some great info, for sure! I've hacked a bunch of their stuff for pedals.  If the voltage swing required isn't large, it should be pretty easy to run many of those circuits on single supply.  But like Granny is saying, looks like he doesn't need to go active.  Likely doesn't need 2nd order, either, if it's just to cut rumble....some playing around should be enough to make this work.

[Granny Gremlin]

ESP has some great info, for sure! I've hacked a bunch of their stuff for pedals.  If the voltage swing required isn't large, it should be pretty easy to run many of those circuits on single supply.  But like Granny is saying, looks like he doesn't need to go active.  Likely doesn't need 2nd order, either, if it's just to cut rumble....some playing around should be enough to make this work.

2nd order would be benefitial (you can lower the F3 point and still make sure enough sub bass is being eliminated) BUT the problem is the phase reverse; a full 180 degree flip with 2nd order.  In a balanced line situation this is easy to deal with, but with single ended guitar signal this adds complexity that is not desired - either go active or you stick a transformer in there.

[GibsonGM]

ESP has some great info, for sure! I've hacked a bunch of their stuff for pedals.  If the voltage swing required isn't large, it should be pretty easy to run many of those circuits on single supply.  But like Granny is saying, looks like he doesn't need to go active.  Likely doesn't need 2nd order, either, if it's just to cut rumble....some playing around should be enough to make this work.

2nd order would be benefitial (you can lower the F3 point and still make sure enough sub bass is being eliminated) BUT the problem is the phase reverse; a full 180 degree flip with 2nd order.  In a balanced line situation this is easy to deal with, but with single ended guitar signal this adds complexity that is not desired - either go active or you stick a transformer in there.

Totally agree.   We gotta keep it simple stupid and all that...if it works with the R-C, so much the better!