3 band parametriq Eq troubleshooting (AMZ/Sabrotone)

[bvdl]

Hi,

I've built this pedal:
http://www.sabrotone.com/?p=578

This is the work in progress (I followed the Eq with a clean MOSfet boost from AMZ):


While testing I found out I might rethink the condesor for the lowend frequencies, but what bothers me most are following issues:

1: A slight, (soft) high frequency noise. (Can I solve this with a parallel condenser according to the frequency of the noise? Would 9v be sufficient, or may the circuit suffer from clipping by not having enough headroom?)

2: Althoug it is capable of crisp sounds, it robs the signal of some of the very high end of my passive SSS Stratocaster pickups. (Tested with only this pedal in the chain, is it possible that impedance has something to do with it? I have a buffer I still need to use to test it yet)

Some other factors:

I built the circuit on veroboard and am quite certain the joints are soldered well.

Cuts are neat, and I scraped between traces with a sharp cutter + freed the board from flux, checked for improper wiring and shorts...

I use a wall wart powersupply from Boss, providing 9V-DC (introducing high frequency noise maybe? my other 10 homebuilt pedals don't suffer from noises.)


There's some other work left to do, but so far this are my only concerns about this biuld.

Thanks for your advice,

Benny.

[Rob Strand]

1: A slight, (soft) high frequency noise. (Can I solve this with a parallel condenser according to the frequency of the noise? Would 9v be sufficient, or may the circuit suffer from clipping by not having enough headroom?)

This can be reduced using pre-emphasis on the input opamp and de-emphasis on the output opamp.  This method helps reduce noise without introducing clipping problems.  Check out the Boss GE-7 Equalizer it uses this technique.

2: Althoug it is capable of crisp sounds, it robs the signal of some of the very high end of my passive SSS Stratocaster pickups. (Tested with only this pedal in the chain, is it possible that impedance has something to do with it? I have a buffer I still need to use to test it yet)

I think the cause of the high frequency roll-off is the 100k resistor (R10) at the input.  It is loading the pickup.
Increase this to 1MEG.   In addition you could increase the 1MEG resistor (R11) to 2.2MEG.

[MrStab]

+1, that 100k resistor is insanely low. check this out: http://www.muzique.com/news/pulldown-resistor-vs-input-impedance/

[MrStab]

upon some thought, i would look for other potential issues before adding pre/de-emphasis (which you should). i say this because me and at least 50 people i've made similar circuits for haven't reported this issue at stage volumes, and because even some rack PEQs have their emphasis well above the guitar range yet it's not a widely-reported problem (eg. http://www225.pair.com/audio/waltzingbear/Schematics/Urei/545.JPG 10.6KHz).

switch the MAX1044 to an ICL7660S* and short pin 1 to pin 8. and never buy a MAX1044 again! it really confuses me that people still use this 10V-and-it's-dead, no-oscillator-boost chip! (though i think some 1044s have boost pins...)

[Rob Strand]

IIRC the 1MEG resistors in the gyrators don't help with noise.

[ElectricDruid]

IIRC the 1MEG resistors in the gyrators don't help with noise.

Noise voltage is proportional to resistance in ohms, so 1 Meg's certainly aren't helping. Other factors are temperature (not something we can alter significantly) and bandwidth (don't build a circuit that amplifies radio signals). High gains (like in highly resonant/peaky EQs) obviously boost noise at those frequencies.

http://www.resistorguide.com/resistor-noise/

Tom

Edit: It's worth pointing out there's a trade-off here between low currents (e.g. high resistor values) and low noise (low resistor values). You can't have both. The typical values of 1K-100K are a compromise between these two extremes.

[MrStab]

so i honestly, honestly do not know how i thought this was an SVF, i need to seriously wait before coffee kicks in before posting!
or did something change in the Matrix?!!! luckily my only contributions were regarding charge pump noise and the bias resistor, both still relevant regardless of the circuit.

all i can say re. PEQ gyrators is i found them a finnicky pain-in-the-behind to both make and to use, as the parameters don't work independently.

actually, i think i remember Sabrotone or Tagboard had a similar circuit that was an SVF with a charge pump, so maybe i'm confusing it with that.

[PRR]

> high frequency noise.
>  robs ...very high end of my ....pickups.

The obvious reasons are 2Vpp signals in a path that could pass far higher levels, and that low 90K input impedance.

As said, make the input resistors a Meg or two each. Using TL071 there's no reasonable limit.

The first stage can be rigged for gain of 2 or 4, with opposite cut at the end. This gets hiss 1/2 to 1/4 of before, relative to signal.

http://s22.postimg.org/4qyv53j9t/bvdl_1.gif

[PRR]

> Noise voltage is proportional to resistance in ohms

No; to *root* of resistance.



Any other result would violate conservation of energy.

> 1MEG resistors in the gyrators don't help with noise.

Not IMHO substantially, for several reasons.

When boost/cut is centered the gyrator hiss cancels at the output.

They are only 1Meg when that section is set to its lowest frequency.

(dubious) C1 and "Q1" are the C and R in a LCR filter. That other junk makes a fake "coil". At first blush, hiss in the "coil" is rolled-off by rising impedance. (I admit this may be wrong, but I'm too sore to wrestle it in SPICE.)

[blackieNYC]

http://www.diystompboxes.com/smfforum/index.php?topic=113018.msg1044916#msg1044916

A good short read.  Shorter - try 5532s. I had a somewhat similar problem. And also took PRRs advice.  Not enough gain if you do a midrange cut. Mine is quiet and reliable.
PS: I think you add (space)width=600 right after the first img  "[img width=600]" something like that