Jump to content

Passive Mid Cut Filter Design


SwedishLuthier
 Share

Recommended Posts

I have a set of fantastic sounding rock flavor single coils. Actually they are a split humbuckers designed like a P-bass, but the sound is exactly like over wound traditional single coil pickup without the noise. Anyway I would like to have a bit more flexibility from them. One alternative is to unwind a few hundreds of turns, put a lead wire there and wind them back to original specs. But I have been playing around with a parametric EQ and noticed that if I cut about 6 dB at 550Hz with a pretty normal Q with I get exactly the more traditional strat sound I'm after. I would like to have an on board passive filter to do exactly that. I have tried to read up on band pass filters, notch filters and similar. And even thou there are calculators available on line I'm really completely lost when it comes to analog filters and need a helping hand. So what I need is a passive 6 dB, or adjustable cut, filter that I can switch in and out of the circuitry. Anyone know of such filer and possible have a schematic drawn up? Any help at all is really appreciated here.

Link to comment
Share on other sites

Passive notch filters are a little tricky because most topologies require an inductor (or multiple inductors, depending on how many poles you want the filter to have) and inductors are not too accurate, can generate noise and have very strong magnetic fields inside and near them in addition to non-linear resistances (with respect to frequency). If you wanted to go active, that would be much easier and take less space because the 9v battery would probably take up less space than the inductors needed to realize a passive notch filter. You might be able to chain up simple resistor/capacitor low pass and high pass filters, but those aren't terribly accurate. I would say if you want to make the circuit yourself is to use capacitors and an inductor-simulating opamp arrangement. You get much more control that way. Of course, if you go active, you might be able to find an off-the-shelf mid cut circuit.

Link to comment
Share on other sites

I was hoping for a RC circuit. I can easily make an inductor myself as I make my own pickups. It is essentially a coil with an optional steel core. So if an inductor is needed that shouldn't be a problem. And yeah, I know I can get an off the shelf active EQ type of on board unit, but I would like to keep this all passive for two reasons:

- It will not be dependent on a battery

- I have a prejudiced against "active" electronics on board.

I know that they are perfect for some things and so on, but call me conservative in that aspect...

I have found on-line calculators for high and low pass filters, like this for one pole filters:

http://www.muzique.com/schem/filter.htm

but as I have too little theoretical knowledge in that field I cannot really make any thing out of them.

Anyway I appreciate your input Ripthorn

Link to comment
Share on other sites

Here you go: http://www.guitar-mod.com/rg_passive.html

That uses caps/resistors and an inductor to cut out the mids.

Edit: Oops, that operates at 850Hz though.

Edit 2: They have some information on making your own http://www.guitar-mod.com/rg_mods_faq.html including an equation to give you the value of inductor and capacitor you need to cut at 550Hz

Edited by Keegan
Link to comment
Share on other sites

If you wanted to leave out the inductor, it's possible by cascading an rc lowpass and an rc bandpass. Your crossover frequency would be 550 Hz and for a 6 dB drop, 550 Hz has to be the 9dB down point for each of your filters. You can tinker some with the Orcad Pspice demo and get all your frequency response plots with that to help tweak the values. Of course, something else you will need to take into account is the input and output impedances, as those can shift the notch frequencies of your filter. If you are fine with an inductor, that will facilitate things and you can achieve a higher Q if you want. If you still want more info, I could maybe get a schematic with some values up and going in the next couple days or so. I would just need to know what kind of Q you are looking for. Then again, the above link might be all you need.

Link to comment
Share on other sites

That site with the pre-made notch filters is cool, but if you want to do your own:

http://www.mouser.com/catalog/636/998.pdf

Seems cheaper and easier than winding your own, and they're tiny.

Best,

Todd

OK, somebody should check my math, but I think the 42TL018 using both sides is about 5-6 H

which gives a notch a little below 500 hz with a .022 uf?

Sorry this isn't my field, we need an Electrical engineer, not a physician doing this math :D

Edited by ToddW
Link to comment
Share on other sites

I have already tried the Torres mid cut circuit that looks very similar (possible the exact same Moser inductor that Todd linked to) to the Rothstein, and sorry to say, but it sucked big time. It cut mid all right but also attenuated treble pretty much like a trad tone pot, so I am loosing both mid and treble.

Anyhow I might try to use the existing inductor from the Torres thing and calculate a different frequency. That formula in the second link was really good to have. Thanks Keegan. The inductor from the Torres unit has a 1.8H inductor so if I put in F=550 HZ I get 4.65X10-8, 46.5 pF if I’m not mistaken. Can anyone confirm that?

The best site for LC filters I have found is this:

http://www.allaboutcircuits.com/vol_2/chpt_6/2.html

But I would also really appreciate if someone knew how to make a passive RC filter, just to be able to compare them.

Ripthorn:

I really have no idea of the Q I need. Typical me to call it “a pretty normal Q” earlier on. But I can see that the attenuation is roughly 3dB at about250-300Hz and in the range of 1.3 Hz and the peak attenuation at 550Hz is about 10-12 dB. If you would like to make something out of that tiny bit of info I would really appreciate it.

Thanks for your input everyone

Edit:

When having a closer look at the Rothstein Mid Scoop / tone switch I can tell that they are exactly the same unit as Torres sells. Stay away from that if you need a pure mid cut filter

Link to comment
Share on other sites

Your math is a little off. 4.65e-8 is actually 46.5 nF. That is a fairly typical value (47 nF). Looks like you need a Q of about 2, which is really quite low, but that is because I usually think of this stuff applied to things like wireless communication, where you need really high Q's. In order to get the proper design, I would need to know a little more about the inductor (what is it's self resistance at 550 Hz, in particular), but other than that, you can use your same inductor and capacitor. Once you know the inductors resistance, finding the attenuation at 550 Hz won't be hard, and the schematic is just a resistor, inductor and capacitor in series where you take your voltage out after the resistor but before the inductor and capacitor to ground. I will see if I can whoop up a schematic in paint real quick and post that.

Link to comment
Share on other sites

notch.jpg

The calculate for your resonant frequency is still the same, but to determine the attenuation, you want Rs/(Rs+R1)=0.316 to have a 10 dB drop at 550 Hz where Rs is the self resistance of the inductor at 550 Hz. Hope this helps, let me know if you need more info.

Edit: I know the inductor drawing sucks, it's paint, what can I say?

Edited by Ripthorn
Link to comment
Share on other sites

Your math is a little off. 4.65e-8 is actually 46.5 nF.

He he. I'm not surprices Ikeep mixing things ut all the time, nano, pico, micro... Yeah I think I have a few 47nF caps sitting around somewere already. The measurements on the inductor, 1.8 H and 168 Ohms were taken with a pretty cheep RCL meter that uses 250Hz. So I'll use a 115 ohms resistor for R1, right?

Anyway Im very happy for your help

Link to comment
Share on other sites

If you have a signal generator or something that can put 550 Hz into your inductor, you can just measure the resistance across it while the signal is going through and use that value as Rs.

Link to comment
Share on other sites

Whell no, I dont have that. But I'll use 115 ohms as a base and fiddle around in that area and se what happens and how it sound.

I'l try it out later today and see how I like it

Thanks

Link to comment
Share on other sites

If you can get the number off the xicon inductor, you should be able to look up the resistance at 1khz, I think that's what's on the data sheet. Also, that schematic seems to have left out the pot that's in series with the cap and inductor. This is getting wired like a tone control, right, so one side of the pot to the hot, and the wiper to the cap, then to the inductor (+/- a resistor in parallel like the Bill Lawrence Q filter) to ground. The other end of the pot isn't grounded.

Anyway, to me, you're going about this wrong. You know the values from the formula, and what it sounds like now, so why not use your ear at this point. If the current torres values cut too much treble, try a smaller cap. If that doesn't cut enough mids, you can go back to the same cap and use a bigger inductor . . . or both sides of the transformer xicon mini transformer.

I don't think you'll find your sweet spot with the formula, but I am very curious to hear how it works out.

Todd

PS) I have a 6 way rotary switch wired up with 3 different cap values and two different inductors to give 6 different tone control choices on the guitar I'm working on now. Two are standard tone controls with just a cap, the others are different notch filters. Almost at the point I can wire it in. :D

PPS :D ) http://guitarnuts2.proboards45.com/index.c...3605&page=1

Edited by ToddW
Link to comment
Share on other sites

I guess I need to state that I have never built one of these for a guitar. What I gave above was from a filter design book from my filter design class. If there is something easier that someone else has already developed, then go for that, I am just thinking from a purely electrical engineering standpoint. Though you wouldn't have to wire it up for a pot at all, you could just hook it up to a push pull pot (or toggle switch) where the switch simply engages the notch filter (if you get it to sound how you want). Though I guess a pot would let you choose your cut amount. Lots of options.

Link to comment
Share on other sites

you can get quite a good filter utilizing the rc network lpf 1.9k .15uf cap. 558.7hz freq range. for the hpf 2M 150pf or .00015uf if my conversion is solid gives us 530.8hz. hope that helps. also you can add a single buffer in there such as a mosfet or a jfet and do a dual design. the current rc single filter section gives you -6db per octave. with two in series you can double that.

Link to comment
Share on other sites

I guess I need to state that I have never built one of these for a guitar. What I gave above was from a filter design book from my filter design class. If there is something easier that someone else has already developed, then go for that, I am just thinking from a purely electrical engineering standpoint. Though you wouldn't have to wire it up for a pot at all, you could just hook it up to a push pull pot (or toggle switch) where the switch simply engages the notch filter (if you get it to sound how you want). Though I guess a pot would let you choose your cut amount. Lots of options.

Hey Ripthorn,

I looked back and he did say he wanted to simply switch it in and out. Missed that, figured it'd be on one of the tone controls. My bad.

Best,

Todd

Link to comment
Share on other sites

I built a simple test board to be able to switch caps in and out. The best result I got was with a 5nF cap (or really 10 50nF in series…) but that still didn’t shape the treble the way I would like it to do. It looks like the inductor is right for me as the low mid cut off frequency sounds right. But I cannot get enough treble. Possible I need a steeper treble cut of part of this band pass design. Isn’t it called a two pole filter or something like that? You can really tell I have no real clue of what’s going on…

For now I’ll have to either find an alternative schematic or try what ansil suggested. I just have to get those parts.

Link to comment
Share on other sites

The only issue with what ansil suggested is that it is active. If you want to go active, you can do the rc design as it is very simple, or if you need a higher Q, you can actually use active parts to simulate an ideal inductor and use it with capacitors for a higher precision filter. If you do go active, it might be easier just to buy something off the shelf. I can look to see if I have a schematic for a higher pole filter around here, or some other way to boost the Q. The only thing is that the Q for a passive filter (and most active filters) is symmetric about the resonant frequency. I think you would have to be clever to get an asymmetric rolloff for a notch filter. Let us know if you want to go active or stay all passive.

Link to comment
Share on other sites

The only issue with what ansil suggested is that it is active. If you want to go active, you can do the rc design as it is very simple, or if you need a higher Q, you can actually use active parts to simulate an ideal inductor and use it with capacitors for a higher precision filter. If you do go active, it might be easier just to buy something off the shelf. I can look to see if I have a schematic for a higher pole filter around here, or some other way to boost the Q. The only thing is that the Q for a passive filter (and most active filters) is symmetric about the resonant frequency. I think you would have to be clever to get an asymmetric rolloff for a notch filter. Let us know if you want to go active or stay all passive.

actually there is no real problem with what i said. if you read above it says that you can add a single mosfet or jfet to help drive it. you will lose a little signal with the way i showed itabove but it will work all the same which is why i used the resistors i chose above. for the hpf that has reistor to ground i chose a 2M so that it wouldn't affect the impedance as a lower resistance would monkey with the imp of the vol pot.

and i used a small resistor for the lpf so you wouldn't lose that much signal. seeing as its really only attenuating the mids everthing else is cool you shouldn't notice that much of a difference.

ciao ed

Link to comment
Share on other sites

I built a simple test board to be able to switch caps in and out. The best result I got was with a 5nF cap (or really 10 50nF in series…) but that still didn’t shape the treble the way I would like it to do. It looks like the inductor is right for me as the low mid cut off frequency sounds right. But I cannot get enough treble. Possible I need a steeper treble cut of part of this band pass design. Isn’t it called a two pole filter or something like that? You can really tell I have no real clue of what’s going on…

For now I’ll have to either find an alternative schematic or try what ansil suggested. I just have to get those parts.

Peter,

Did you use the transformer from the torres circuit? IIRC, you have one of those, right?

I have two of those mouser/xicon transformers, I think the TL018 and TL218. Using on 10k winding should give me about 1.4H, using two in series should give me about 5.6H

I plugged those into a spreadsheet formula and the 4 filters I'm starting with are a .042uf and 1.4H for 650, a .022uf and a 5.6 for around 450hz., a .042 at 5.6H at around 320hz, and a .1uf and 5.6H for around 200hz. Position 2 will actually be a "Q-filter", and position 1 will be a standard .022uf. Easy to wire with a 6 way switch and a 500k/DPDT switch to cut the circuit out completely. That's just math, and obviously I'll tweak by ear.

I think .068uf and 5.6H would give you a freq around 500hz. If you use the TL018 you'd get more like 5.3H so it'd be a little higher frequency. If you decide to play around with this, the transformers are only a couple of bucks from Mouser, but S&H is around $7 I think. Course, that's in the US . . .

Best,

Todd

Link to comment
Share on other sites

Todd, you are right. I played around with that Tottes transformer. One problem is that Torres clips all of the legs on the secondary side and one on the primary side. I do not have the moser parts number for the transformer but it is stanped with 0616S. I measured it with my LCR meter and with 250Hz (only frequenzy availible with that meter) it gave me 1.8H. At what frequenzies are your numbers?

You're probable right, I need to get a few more inductors to play around with . You haver giben me a few good starting points. I'll contact Moser and asks if they ship to Europe.

And if I understood things correct ansils circuit also would work passivly (is that the right word, passivly...).

Thanks guys

Link to comment
Share on other sites

Henries is henries I think. The resistance you measure may change at different frequencies, that's basically the point of the inductor, but the inductance of the coil shouldn't change. Any EE's who can help confirm that? Inductance won't change with frequency? Maybe impedance will? :D

Either way, you have a 1.8H inductor.

To put the notch of your filter at 454hz, put the inductor in series with a .068uf cap. That's close to A440 with a standard value cap.

To put the noch at 578hz, use a .042uf cap.

That math doesn't account for the resistance of the inductor, but I'm not sure how significant a 500 ohm resistance is.

Let me know how it sounds if you try it, I'm curious since mine isn't wired in yet. Heck, I just re-sprayed the guitar today after my first try checked, so I'm still a few weeks away from hearing it.

Best,

Todd

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...