New Project at GGG - 6 Band EQ

[Ben N]

Looks like a string of Anderton tone boosts tuned to different center freqs, no?  Anyway, it looks like a very nice & useful project.  Props.

Ben

[Transmogrifox]

just replace the capacitors with jumpers

are you talking about the electrolytic in the feedback loop of the first stage, or are there others? This is the only one that looks 'goofy' to me.

Just the two 10uF caps on the input stage, so it's DC coupled.

Maybe if we're lucky somebody like R.G. will pop into this thread and comment about the buffer.

Thanks for posting. It looks like a nicely laid out board and I think I'll end up building it here pretty soon.

[smallbearelec]

The next one is a gene splice between my original 800mS delay (DHEcho) and Scott Swartz's vastly superior PT-80 (300mS) design. So, 800mS, with Scott's companding and filtering. And yes, Scott gave his blessing and input. 

Please, don't bug JD with any questions about the SHEcho (Swartz-Hazelwanter Echo) just yet. It will just delay getting the project posted, and many of the questions you have might be answered by what is posted. After it's posted, fire away! 

Looking forward to this.  I still have one more HT8955 left in my parts bin.

No worries...supply of those is pretty good.

NB I also now have NOS TO-18 can 2N4861 FETs for Hornby-Skewes treble boosts, also 2N2646 UJTs if somebody wants to try doing a Repeat Percussion (though I don't know if that's the one that the Vox circuit used.)

SD

[StickMan]

I've been goofing around with a design like that that uses a TL074 where one opamp is used to provide an active Bandaxall tone stack, and then the other three are used to provide the gyrators for a 3 band midrange eq.  The yield would be a 5 knob tone control. 

No input buffer, since I'm going to incorporate it in a distortion box.


dave.

[StickMan]

Now that I've had a closer look at it, there are a few things that concern/puzzle me.  Just like a previous poster, I'm concerned about looking like an idiot, but I have to ask:

First, what's with the 100K resister on the input?  What's that going to do to the input impedance, and how much of a load is that going to put on the guitar pickups?  It doesn't look like the input phase of any other stompbox circuits I've seen.

Secondly, in a single battery configuration, it looks like ground is relocated to Vref (which is 1/2 V+).  But then there's no "true" ground for the circuit.  My main concern is how that will interact with the other stompboxes in the the chain.  I'm mean, if the signal ground for that stompbox is sitting at +4.5v compared to the ground coming out of the boxes in front of and behind it (not to mention the amp and the guitar), what will happen?

Apart from that, I think the idea of using the gyrators in the negative feedback loop of the opamp is very cool.  Once again, though, most of feedback loops I've seen all ground back to a true ground, even if the signal is biased to Vref.  What's the effect here, with no ground other than Vref?

dave.

[Transmogrifox]

First, what's with the 100K resister on the input?  What's that going to do to the input impedance, and how much of a load is that going to put on the guitar pickups?  It doesn't look like the input phase of any other stompbox circuits I've seen.

100k is typically fine.  Some tube amps go as low as 47k.  Most stompboxes have about 500k input impedance.   The Fuzz Face has about 10k input impedance, and the loading on the pickups is part of the effect.

Once again, though, most of feedback loops I've seen all ground back to a true ground, even if the signal is biased to Vref.  What's the effect here, with no ground other than Vref?.

With stompboxes, you use a battery or a transformer isolated power supply, so this isn't an issue.  By defining circuit ground at V/2 and connecting signal ground here, then the 4.5V looks like 0V to every other stompbox, thus you have converted your power supply to +/- 4.5V.  If you want to be conventional about it, then connect ground to the negative terminal of the battery, and don't connect "Vref" to ground anywhere, and you're good.  In reality, no stompboxes ground to a "true ground".  "Ground" on a circuit is arbitrary.  You could take a power supply designed to supply +/-15V and make a funny resistor divider and define the center of the divider as "ground".  For example, you may measure -5V from this arbitrary ground to the power supply -, and +25V to the power supply +.  Now you have a +25/-5 Volt supply. 

It's usually better to define ground on the positive or negative terminal of a voltage source (low impedance of some kind).  That way current flowing to ground doesn't screw up the other voltages in the circuit.  In the case of Vref, it is adequately decoupled with capacitors to a low impedance voltage source, and the only currents we expect to see in a stompbox circuit ground are AC, so the current on the ground will flow right into the voltage source, and little anywhere else.

[Paul Perry (Frostwave)]

Maybe it's relying on the leakage thru the electro.

[Transmogrifox]

Maybe it's relying on the leakage thru the electro.

I had that thought...but then did the transfer equation, and it appears that there is no purpose for having the capacitors since it seems it would not block DC--and that would be the purpose of the caps.  I think the fact that Dean posted it means it works...it's just a matter of answering the big question of why it was done this way in the original design.

Here's the equation for the unity gain state (where the sum of the feedback resistor and pot equal 100k):
Vo/Vin = [(1/jwC) + R2]/[(1/jwC) + R1]   
for R2=R1,
Vo/Vin = 1

The capacitors don't do anything.
In this form of the equation, we directly see that as frequency gets very large (w>>1), then the capacitors do essentially nothing no matter what the values of R1 and R2 (and this would be expected intuitively).

Now I have the question about what happens as frequency gets very low (w--->0).  We can see by manipulating the equation to this form:

Vo/Vin = [jwCR2 + 1]/[(jwCR1) + 1]   

As frequency approaches 0, the "jwCR2" and "jwCR1" terms go to zero, leaving the transfer equation:
Vo/Vin = 1

The only effect left is for the moderate values of w, where there is some effect from the capacitor. For these frequencies, there would be a slight increase in input impedance--and likely negligible, and mostly below audio range.

My conclusion is that the capacitors do absolutely nothing and you can save yourself some money by putting a jumper in place of the 10 uF caps on the input buffer.  It may save some distortion and noise as well.

[ethrbunny]

nice. i was just about to start questing for an eq.

At the bottom there is a note about not sharing power with other fx - why?

[Transmogrifox]

nice. i was just about to start questing for an eq.

At the bottom there is a note about not sharing power with other fx - why?

Virtual ground used on this pedal is connected to ground.  Most other pedals connect ground to the (-) terminal on the power supply or battery.  Hooking this up to another pedal sharing the same power supply would be tantamount to putting a jumper between Vref and power supply (-).  With minimal hacking, you can designate ground to the power supply (-).  Per my math (above) you would need to short the 10uF cap in the output-inverting terminal part of the feedback loop, yet retain the 10uF cap on the input for DC blocking.  (-) terminal of cap would be oriented toward input signal (reverse of what's shown).

[KaptenSpark]

Just built this thing and got into trouble. The eq pots does nothing and the humming is extreme, what could I've done wrong? Everything looks right to me.

Built it on perfboard, the one with the solder islands, not lines or whatever it's called.
Used one battery as suggested and connected "c" to ground.

help me pleeeeease

[jrc4558]

Ahem, check the voltages on the chips at least! Also the orientation of electrolytics for power filtering. You know, the usual... If it hums, its very often DC in the signal chain, or a broken ground continuity. Check that too.

[TELEFUNKON]

do I see problems coming up with people hooking up the battery negative wire to a stereo input jack ring for power switching on by plugging in the cable?

[NoFi]

So if i want to use this circuit with a single 9V supply, are the following connections correct ?

- A would go to +9, B to battery negative, so that C becomes VRef at 4,5 Volts.
- input and output jacks go to ground
- 10k (pulldown?) resistor goes to ground
- input 10uF cap needs to be reverted
- Pin5 of U1b goes to Vref
- 150uF to Vref
- 220k resistors to Vref

And by doing this will it work fine when hooked up to other circuits ?
Thanks for this project !

[Transmogrifox]

So if i want to use this circuit with a single 9V supply, are the following connections correct ?

- A would go to +9, B to battery negative, so that C becomes VRef at 4,5 Volts.
- input and output jacks go to ground
- 10k (pulldown?) resistor goes to ground
- input 10uF cap needs to be reverted
- Pin5 of U1b goes to Vref
- 150uF to Vref
- 220k resistors to Vref

And by doing this will it work fine when hooked up to other circuits ?
Thanks for this project !

"...150uF..."  You mean 150pF, right?  If not, that's the reason it isn't working.  And yes, the 150pF could be connected to to Vref (though it's just as well to connect it to ground).  The important connections to Vref are the 220k resistors and U1b Pin 5.

The 10uF in the U1 feedback loop must be shorted (just place a jumper instead of a capacitor in your perf layout).  Otherwise, the op amp will be saturated to the positive rail.  Measure your voltages and I'll bet that the output of U1b is at 8-9 Volts, and/or moving around a lot.  This was one of my concerns about the input buffer being configured the way it is.  Those caps the way they're configured don't make it any different than if it was DC coupled.

The rest of what you stated is correct.

EDIT:
And "B" (battery negative) is connected to ground on your jack like any other stompbox.

[NoFi]

Thanks a lot Trans !
Yes i meant 150pF and took good note about the 10uF cap that needs to be replaced by a jumper.
I haven't built it yet but will surely do, it would be an nice addition to a distortion project already running with a 4.5Vref.
++
Greg

[Dean Hazelwanter]

Hi...

Sorry for not being more involved with this thread. It's been hellishly busy at work, and every spare (and not so spare) moment has been taken up with my most recent project. I just finished coding for a new project last night. It is a PIC based controller that controls 32 loops/relays, with 64 configurable presets that get saved to internal FLASH memory. There are 2 parts: the rack mounted head which can select which preset (either directly using 0-9 keypad, or via UP/DOWN keys, and displayed on a dual 7 segment display) and which of the 32 relays are enabled for each preset. The head (which can operate independently without the pedal) talks to the floor-mounted pedal array via a CAT5 cable. The pedal has 10 preset switches and BANKUP/BANKDOWN keys to cycle through the 6 banks, as well as 3 x 7 segment LEDs to indicate the pedal's current bank, as well as the current preset. Changes initiated by either the head or pedal are displayed on both displays. A total of 3756 lines of assembler code between the head and pedal.  I'll be the first to admit that I'm not a programmer, and I think this project was penance for my previous job. For almost 10 years I tested software, and our motto was 'anyone can break the program, but it takes a real man to break the programmer!'  The prototype/testbed works fine, so it's time to package it up. A big thanks to Mark Hammer for his input regarding the realities of marketing something like this. If someone wants to market/manufacture this, let me know! 

Anyway, back to the EQ. Thanks to Trans and others for stepping in. While I didn't run this through a circuit sim or do the transfer equations or anything like that, it does work the way it was originally posted - I didn't see any rail to rail instability or anything like that. If I would have seen problems like that I sure wouldn't have posted it. I'm not saying it can't or won't happen, but I sure didn't see it. The beauty of it is, you can still use the layout posted at GGG. Try it either or both ways. As for the non standard Vref arrangement, yes, that's why I put the warnings in the notes included with the project.