Saturday 3 May 2014

2, 4 & 8 Band Gyrator EQ

Moving on from the HM2 which obviously has quite a powerful EQ, I thought I could use the circuit to make a handy standalone EQ.  I was going to just use a quad opamp and make this 2 band with 2 of the channels being used as the input buffer and output gain stage, but thought it would still be pretty compact using a dual opamp for the input and output and so settled on 4 bands.

I have included a table showing the example frequencies you can achieve using common component values for R1, C1 and C2, and just altering the value of the R2 resistor.  Another option would be using 4 x 10K multi-turn trimmers instead of the R2 resistors so you could vary the frequency for every band and really fine tune it to the gear you're using.

You can use the Excel calculator I posted here to select your own components and get the frequency and Q you want if you prefer to fine tune the bandwidth.  I also included a trimmer so you can set to unity precisely, or give a slight boost if you like.




And now everything is modular of course, there's no problem in adding another opamp to give you a further 4 bands.




And the baby of the family



90 comments:

  1. Really interesting Mark! Gonna do some testing with this circuit once i have the time for it, need to get into building again, fallen behind the last month. Drivers License has taken all my time :p
    Thanks!

    ReplyDelete
    Replies
    1. Oh yes I saw on FB :o) Congratulations mate
      I think this may be quite useful in another distortion pedal to give it the HM2 kind of EQ. I quite like chopping things up at the moment and seeing what you can make with the little bits :o)

      Delete
    2. How much boost does the trimpot offer you think?

      Delete
    3. The original had a 3K3 resistor there which would give you 2x gain. With the trimpot set to maximum you'd get 4x. So it's not going to give you a massive boost but should get you well above unity. Increasing the size of the trimpot will increase the gain if you want more.

      Delete
  2. What a great idea. I really love your chopped almost utility veros like this one. Thanks again.

    ReplyDelete
  3. I originally did this as per the HM2 but on reflection I've altered the 68K resistor from pin 3 of IC1 to vref. This sets the input impedance and so increasing it makes sense when used at the beginning of the circuit like this.

    ReplyDelete
  4. I've just had a look at the Boss GE-7 schematic and this is basically how they have done it with a few extra buffers and stages here and there. Although I know the formatting for this will come out wrong, these are the values they have used for the various sliders:

    --------------------R1------R2------C1------C2
    100hz-----------100K----330R----1.5u----56n
    200hz-----------82K-----330R----680n----33n
    400hz-----------100K----330R----330n----15n
    800hz-----------100K----330R----150n----8n2
    1.6khz----------82K-----330R----100n----3n9
    3.2khz----------82K-----330R----39n-----2n2
    6.4khz----------Not a gyrator. Lug 2 > 47n > 820R > vref

    Oh and another thing I noticed is that in the GE-7 schematic the pots are all 10K, so maybe try both on a breadboard before committing yourself and soldering, to see which works best for you.

    ReplyDelete
    Replies
    1. mark is there any math type i can use to find "my" values? cause i couldn't download the calculator link...

      Delete
    2. This is what I based the gyrator calculator on:
      http://www.muzique.com/lab/gyrator.htm


      Delete
  5. Very interesting, thanks.
    The 10K trimmer determines the general level ?

    ReplyDelete
    Replies
    1. Yes, it just sets the gain of the final opamp and so you can tweak it to give you unity or a slight boost if you want. You could even omit it and have it as an external pot if you prefer.

      Delete
  6. Hi Mark!, excellent idea!

    I was wondering if is possible to use this vero as a tonestack replacement of any other pedal, something from:

    In -> Pedal Circuit -> Pedal Tonestack -> Pedal Circuit -> Out

    to:
    In -> Pedal Circuit -> Gyrator EQ -> Pedal Circuit -> Out


    is this possible?
    a

    Thanks master!

    ReplyDelete
    Replies
    1. Yes you could if you can make the existing tone stack inoperable and tap off the effect board, or you could just have the effect output going straight into the EQ input and so running them both in series which would probably be the way I would go.

      Delete
  7. Does anybody know what is the Eric Clapton mid boost center freq and its Q ?
    Observing the EC mid boost schematic, it strongly remind me of a discrete op amp in a gyrator configuration, but I'm not able to identify the R1, R2, C1, C2 value....
    It could be easier building it with a a normal opamp (741, tl0x1, ecc.)

    ReplyDelete
  8. Is this like cut/boost EQ or just it just cuts frequencies?

    ReplyDelete
  9. Can I use this EQ circuit in my amps effect loop? I mean, can it handle the bigger signal levels coming out from the preamp without overload/distortion?

    ReplyDelete
    Replies
    1. I don't see why not, it's exactly the same type of tone control used in the Boss GE7 and similar EQ pedals, and I've used them successfully in an amps effect loop in the past.

      Delete
    2. Ok, thank you for the answer! I haven't used any EQ pedal in my effect loop yet, so I have no experience with it.

      Delete
  10. Hi Mark.

    The pic says "note IC2 rotated 108*" if my eyes are ok it looks like BOTH IC2 and 3 are turned 180. Correct?

    jeff

    ReplyDelete
    Replies
    1. Yes it's just because the notes were made for the 4 channel version. All of the quad opamps will need rotating.

      Delete
    2. Thanks man....happy wednesday!

      Delete
  11. these are te values i'm going to use.
    and i want to know if i don't get something right.

    r1: 180k
    c1:0.1μf
    c2:0.022μf

    and now for the r2
    1) 12k---73.05hz---q 1.82
    2)4.3k---122.03--q 13.03
    3)1.2k---230.99---5.75
    4)220r---539.49---13.42
    5)56r---1069.3---26.59
    6)20r---1789.28---44.5
    7)4.7r---3691---91.78
    8)2.2r---5394.88---134.16

    what do you think guys?
    i really need to be sure on that.

    ReplyDelete
    Replies
    1. The spread of frequencies looks good to me, but I'd still be very tempted to socket a couple of the components simply because the Q may be a little high with some of them. That may suit your rig and sound perfect but it also may take some trial and error using different component values so you can get the same frequency response with a lower Q, and you're not going to know until you try it.

      Have you considered breadboarding it first just to see what you think about the response? You could just breadboard the input and output opamp and a single EQ channel, then just swap R2 to see what you think of the response for each of the 8 listed above.

      Delete
    2. tag it mate!
      now verified.
      it's working really nice. the trimmer has a really small effect (for my ears) on the general gain, so it's not something that i couldn't live without.

      i've used other values from what i've posted yesterday. i'll make another post tomorrow with those i've tried and with voltages too just in case someone what to have them as a guide.
      to be honest, i still don't fully get what Q stands for...so my resistors will stay in sockets for sure...!!!

      thanks for everything

      Delete
    3. Awesome, thanks for verifying matey. If you increase the value of the trimmer then you could get a bit more of a boost by increasing the gain of the output stage, but I only really included that so you could give a slight boost if you wanted match unity. But you can increase it if you think it could be useful.

      The Q tells you the width of frequencies you're affecting. As the Q gets higher, the frequencies affected get narrower. This diagram should help make it clear, and you can see it's a case of finding the happy medium between the frequencies you're choosing and the Q of the filter.

      I've done a 2 channel version of this using just a quad opamp which I'll add to the post. I reckon that one maybe useful to include in the same box with builds which may not have a particularly good EQ, so at least you can tweak a couple of important frequencies.

      Delete
  12. Excellent job, thanks for the pic

    ReplyDelete
  13. What's the formula for the frequencies and Qs? Calculator is gone. Just getting the Filedropper front page.

    ReplyDelete
    Replies
    1. You can get it from here. The calcs are at the bottom of the Filter tab

      Delete
    2. Actually try this version. I noticed some of the calcs on that one weren't working properly, I must have deleted something I shouldn't from the original spreadsheet

      Delete
    3. Thanks! But it don't works with the OpenOfice Calc / MS ExcelViewer and I don't want to buy the whole office pack just for calculating some frequencies ;)
      But a search brought me this which looks pretty similar to the excel calculator.

      Delete
    4. Damn, just saw that you posted that link here :o

      Delete
  14. I want to do the 2 band, basically with a bass and treble sweep, that I will put in series with another effect in the same box. So, as I understand it, I need to find a good center frequency with the lowest Q possible to get the widest sweep? I've been playing with the numbers on the calculator and I can't seem to find the best combo. Anyone have any suggestions for a bass and treble?

    ReplyDelete
    Replies
    1. I haven't experimented with it yet, but I'd be tempted to start with the bass and treble values used in the HM2 and moving on from there. Whether you're a metal head or not, the HM2 EQ is powerful and well regarded and it may not need any further tweaking no matter what the music style.

      Delete
    2. I added the 2 band in series, but I'm having trouble. First off, I overlooked the note about the IC orientation, and after turning it the right direction, and then having to replace it, I got it sort of working. I get no response from my bass control (for with I used the example components for 159.24 Hz), and I only get responce from my treble control (example components for 3394.92 Hz) at the last 10 percent of pot, at which point it is quite useless. Did I fry something else when I inserted the IC wrong?

      Delete
    3. i've build the 8band one as you may already read above, and some of my pots have the same "last 10 percent" problem. as i build it i left it that way, but sometimes i feel i shouldn't...

      also i guess that we shouldn't always go for that lowest Q possible, cause that mixes the pots range so that you could easily get the same hz from 2 or even 3 pots...at least that is what i get at some sets of that eq with the values i've used...so i strongly believe we should do some serious maths on that, and choose the right values.

      Delete
    4. Yes I agree, lower Q by itself doesn't mean quite as much if it means frequencies overlap. I would suggest using the values used in the Boss GE-7 and going from there.

      Delete
    5. If you enter the values used in the Boss GE-7 in the calculator you will see they tried to maintain the Q in 3.5. So if you want a freq that is not in the GE-7 I would suggest tweaking the values trying to maintain the 3.5 Q.
      I´ll try the 8 band one with 6.4khz freq as the highest freq. In the schematics I don't understand why this only freq has a different configuration. So tweaking the values in the calculator my best attempt to maintain the 3.5 Q was this:

      R1: 82k
      R2: 720R
      C1: 0.01uf
      C2: 0.001uf
      Freq: 6.5k
      Q: 3.37

      Thanks!

      Delete
  15. I am having an issue where if I move any of the 4 pots from the middle the sounds almost cuts out. With the trimmer all the way down it works, but there isn't much boost in the frequency, only cut. I checked for bridges and everything looks good. It was working for the past week just fine up until now.

    ReplyDelete
  16. I'm trying to achieve the Mesa Boogie Mark V Graphic Eq with the bands: 80/240/750/2.2k/6.6k. Searching here at googl, I've found some guy that calculated the Q value with the Mesa specs and got it:

    80Hz /Q=1.17
    240Hz /Q = 1.937
    750Hz /Q = 2.12
    2200Hz /Q=0.673
    6600Hz /Q = 1.0

    I'll try to calculate with these values...

    ReplyDelete
  17. Stupid question: Is C1 an electrolytic cap??? oO

    ReplyDelete
    Replies
    1. Nah, just a 100n greeny or poly or whatever you have.

      Delete
  18. Well... Theorically, the Mesa Mark IV GEQ has different frequencies compared to the labeled ones... There are some discussions about this on the web. They use a RLC circuit, with a inductor, so you need to use this calculator to find the Fs and Qs:

    http://www.cvel.clemson.edu/emc/calculators/RLC_Calculator/index.html

    Using it, you'll find the right frequencies and the Q values for each band... But again, there are some VERY different bands comparing to the labeled on the chassis, take a look:

    80Hz -> R=470 / L=1H / C=3.3uF ---> Real F=87.61Hz / Q=1.171
    240Hz -> R=470 / L=0.39H / C=0.47uF ---> Real F=371.7Hz / Q=1.938
    750Hz -> R=470 / L=0.22H / C=0.22uF ---> Real F=723.4Hz / Q=2.128
    2200Hz -> R=1K / L=0.068H / C=0.15uF ---> Real F=1576Hz / Q=0.6733
    6600Hz -> R=1K / L=0.033H / C=0.033uF ---> Real F=4823Hz / Q=1.000

    I really don't know why!! The 6.6k is so far from the 4.8k, so... Maybe, following the maths is the right thing to do... The Boss GE-7 has some little differences, but they are kinda insignificant.

    Another detail> Mark IV uses 50k as the pots instead of 100k (boss uses 10k), but I don't think this would affect the sound at all.. What do you think, @IvIark?

    ReplyDelete
  19. Can this be run at 18V? The TL074 datasheet says it can take +/- 18V, which would be 36V rail-to-rail...since we're not using a bipolar power supply here, and VCC- is connected to 0V, you could give it upwards of 30V (with appropriately rated caps), correct?

    Also, this site has a nice calculator for figuring out the bandwidth by inputting the Q factor and center frequency (which your excel spreadsheet calculates).

    http://www.sengpielaudio.com/calculator-cutoffFrequencies.htm

    -Ryan

    ReplyDelete
  20. This comment has been removed by the author.

    ReplyDelete
  21. Hi , Can i use this for bass guitar ? what is the bandwidth and Q Factor value for bass ?

    ReplyDelete
  22. Replies
    1. hi mate! i just saw your comment.
      i think that for this build, the best way is to perfboard it and the choose your values for each pot.
      when i first build it, i just did maths for the values and the results weren't as good as i was expected as you can read from my comments above.
      so...your ears are the best advisors for this one.

      hope i helped.
      savvas

      Delete
  23. Just finished a little compact 2-band one the other day. 9mm pots, no footswitch (toggle switch instead), in a 1590A. Made a PCB for it based on this vero layout. It's tight.

    http://i.imgur.com/2YqbjJR.jpg
    http://i.imgur.com/vaFgk63.jpg

    -Ryan

    ReplyDelete
  24. IvIark, what do you call Vref ?
    Is it the link where all the R1 are going, so where there's a link on the left of the TL074's ?

    I'm gonna make the best bass eq to use with distorsion ever ;-) I will tell the components when the pedal will be done but I can already give the band freqs which I think are the most appropriate : 50, 120, 400, 600, 830, 1k2, 2k5 and 5k (peaking for the seven first ones and shelving for the last one).

    ReplyDelete
    Replies
    1. Vref are 4,5V which you get from the voltage divider (Row 7 on the 8 Band) using two 10k resistors from 9V to GND.

      Delete
  25. please give me all BOM man ,i want build this eq. thank for a great layout

    ReplyDelete
  26. This comment has been removed by the author.

    ReplyDelete
  27. This comment has been removed by the author.

    ReplyDelete
  28. Hey guys. has anyone experienced that the low frequency band doesn't seem to work right? my 80Hz Q0,9 (R1: 180k, R2: 22k, C1: 0,1uF, C2: 0,01uF) Band doesn't cut or boost anything audible really... while the other ones work fine. Do I need to use different values? Or a higher Q? My other three Bands have a slightly higher Q (approx. 1,5)

    ReplyDelete
  29. Hey everyone! :) my build works alright but I used 100k lin slide pots and they seem to respond really bad... Most of the action is happening in the very last upper bit of the pot. Any ideas??

    ReplyDelete
  30. oh I got it!! Use 1k lin pots and it works like a charm!! But, any Frequency below about 160Hz wont respond properly, so forget about that low bass...

    ReplyDelete
  31. Hey all.

    For some reason I cannot get my head around how to wire the pots from the vero? What exactly is going to the all band 1 and 3? I'm assuming band a and b 2 is going to log 2 of each pot? Thanks.

    ReplyDelete
  32. The 1 and 3 wires are connected to 1 & 3 lugs on all of the band pots. Then Band A, B, C and D are the frequencies you choose to use yourself from the table (or your own calculations) and you connected lug 2 to the points shown.

    ReplyDelete
    Replies
    1. Thanks heaps. Makes total sense now. Looking forward to putting it together. :)

      Delete
  33. Any ideas what to do about hiss? My circuit is quite noisy, especially as it put it in one box with an OD...

    ReplyDelete
  34. Hi, the trimpot also cuts volume or just boosts? I'm planning to use it as a cut with EQ to clean up my amp overdrive. Thanks.

    ReplyDelete
    Replies
    1. it also cuts. use 1k lin pots though, way better response.

      Delete
    2. Thank you! Any suggestions on the freq of a 3 band EQ to use this way?

      Delete
  35. Hi, I make EQ 4 bands, The pedal works, but it saturates / distortion. An idea ?

    ReplyDelete
  36. Hi Mark, can you reupload your calculation file?
    you're current link is dead.
    Thanks

    ReplyDelete
  37. I need some help with this one. I have it all built but all I am getting out of it is a high pitched whine/whistle. I thought that I saw somewhere, where that noise was indicative of something. I've already switches out the opamps pianos and no change.
    Thank you

    ReplyDelete
  38. Hi there! I begin the 2 bands EQ and noticed a little mistake. Le the board is a 18x11 not a 18x12 ;) I will make this with the first band around 83Hz Q around 5 and 803Hz and Q=1.5 to add this to the multicab sim :)

    ReplyDelete
  39. Thanks for all your layouts, it's great to see so many builders doing this.. my question is could this be used as outboard eq in mixing ? If so any ideas on what needs to be altered for this to work correctly as an outboard eq?
    Thanks again

    ReplyDelete
  40. Hey Mark.

    Question. How critical is the potentiometer value to the gyrator circuit? Will it affect the Q too much if I change it to a 20k-W from the 100k? I notice the BOSS GE-7 uses 10k pots on the same circuit. Does this pot function like a blend or is it part of the gyrator?

    ReplyDelete
  41. Hi there ! So I did the 2 bands EQ with Band 1 R1=430K, R2=1.6K C1=0.22uF and C2 =0.022uF Fq= 87.26 Q=5.18.
    For band 2 I have, R1=33K, R2= 1.2K, C1=0.1uF and C2= 0.01uF, Fq=800.19 and Q=1.66.

    But I've got a problem, I put it in the multicab sim (before it, to avoid the clipping of the Mutlicab IC) but I don't have the unity gain even when the trimpot is all the way up.To hear something, I need to put all the way uo the volume of my distorsion and the multicab sim.
    Really need it for rehearsal so any help please ?? thx !

    ReplyDelete
    Replies
    1. No body can help me ??

      Delete
    2. If you try a bigger trimpot it will increase the gain. Try 20K or even 50K

      Delete
  42. This comment has been removed by the author.

    ReplyDelete
  43. Hello! What pot values did you use? And also, does the whole pot taper function normally? Thanks!

    ReplyDelete
  44. Hey everyone. Ive been working on an 8 band and have been plugging in some values trying to find useful bands to use. As said about the ge7, q values of approx 3.5 seem to be good as starting points. A lot of posts here had some strange capacitor values as well, so my design has some more common values used.

    Note that my hz values are very slightly rounded.
    Hz. R1. R2. C1. C2. Q
    90hz -150k-330- 1uF -47nF- 3.4
    150hz- 68k-330- 1uF -47nF- 3.1
    270hz- 47k-330-470nF-47nF- 3.7
    520hz- 39k-330-330nF-22nF- 2.8
    870hz- 12k-270-220nF-47nF- 3
    1900hz-5k6-120-220nF-47nF- 3.2
    3800hz-10k- 82-100nF-22nF- 5.2
    6000hz-6k8- 47-220nF-10nF- 2.5

    I chose to make the 520hz mid band and 6k bands a bit wider for more of a mid scoop control and air-y boost or cut on the 6k. I made 3.8k more narrow because 3.8-4k is a really shitty and harsh frequency in most cases, especially in guitars so i made it narrow to do a more surgical cut.

    Going to try it out and see how it goes.

    Whats the consensus on pot values? Im going to try 100K for now but whats best? 10k? 50k?

    ReplyDelete
  45. Ive been told that r1 and r2 set the gain of the controls? The capacitor values are difficult to find if I use set values for all r1s and r2s

    ReplyDelete
  46. Can you repost the calculator? The link's gone down.

    ReplyDelete
  47. I second that motion to repost the Excel calculator. Thanks!

    ReplyDelete
    Replies
    1. I used this and it worked out nicely. http://www.muzique.com/lab/gyrator.htm

      Delete
  48. Is there a fixed resistor value that will give me unity gain or will it differ slightly from circuit to circuit?

    ReplyDelete
  49. Built this and whenever the 9v gets plugged in for the first time I get an annoying high pitched squeel. I found that shortcircuiting some resistors and capacitors with the IC helps the squeek disappear. Any ideas for a permanent fix skipping the need to open up the pedal everytime?

    ReplyDelete
    Replies
    1. I have same problem. I have tried this circuit several times and never worked for me.

      Delete
  50. Okay, help me out. I just built the 8-band EQ, but I didn't look much into how the EQ parameters were built prior to (I don't have a formal education in electronics). The "Example" values seemed like they would work fine for what I was looking for—which is just an 8-band version of the GE-7. I used B100k pots, 100k resistors for R1, .1uf caps for C1 and .01uf caps for C2. Here's what I did for R2 values, along with the frequencies I was going for:

    RA = 10k, Freq 159
    RB = 4k7, Freq 232
    RC = 1k5, Freq 411
    RD = 390ohms, Freq 806
    RE = 220ohms, Freq 1073
    RF = 100 ohms, Freq 1592
    RG = 22 ohms, Freq 3394
    RH = 7R5, Freq 5814

    However, I noticed that the pots don't really seem to do much. Pot H seems to make a mid-range shift, and pot G makes the volume cut out entirely if I turn the pot all the way up OR all the way down. The other pots don't impact the output much at all. Before I go into a deep dive to see if I soldered something wrong, do the example values actually produce a noticeable EQ change? Or was this project meant to produce more subtle parameter shifts?

    ReplyDelete
    Replies
    1. I also built a second EQ—the 4-band, and I'm still running into issues. I used B100k slider pots instead of the regular pots that I used in the 8-band above. I took values from someone else's post above:

      Hz. R1. R2. C1. C2.
      A 270hz 47k 330 470nF 47nF
      B 520hz 39k 330 330nF 22nF
      C 870hz 12k 270 220nF 47nF
      D 1900hz 5k6 120 220nF 47nF

      The pots do seem to impact the EQ, but they only increase the treble, and only when I have them in the center: meaning that if I slide them up or down, they get more bass-ey, and all of the pots have this issue. There isn't much in the way of altering the EQ in any useful way, even if I use the center as a baseline. It's just "Do you want more treble, or no?" As I move from A to D, the level of treble seems to go higher, so that's a start.

      It seems like terminal 1 on each of the pots should go to ground? What frequency range should I shoot for to adjust the bass?

      Delete
  51. Hey there, is it possible for you to reup the mentioned excel? And what's the db range ?

    ReplyDelete