Dual EQ filter with a single pot?

[fryingpan]

The circuit actually works very well. The only thing is that I wanted to try to implement this with as many discrete components as possible (BJTs and JFETs), in order to add some colouration to the signal (I'm looking at about 1-2% THD). Plus, the bridged t filter is a bit too wide for the scoop I'm looking for (the design usually yields Q=0.2-0.5, if not bootstrapped, but I don't know if you can bootstrap it with a BJT, for instance).

[marcelomd]

Hi,
I tried again with the right voltages. Works as expected =)
No prints because I managed to crash LTspice and lost the sketch =\

[iainpunk]

hey, try the boss/roland discrete opamp for this. it overdrives really nice, and is typically used at decreased headroom to clip earlier, i believe it works just as well at the full 9v


the BJT output is the opamp's output, the Jfet gates are the (+) and (-) inputs.

you can narrow the cut of the bridge t by changing the caps and resistors to these values. but i don't think you can get such a narrow Q without bootstrap.

i hope you find a way to make it all work,
cheers

[fryingpan]

Anyway, this is what I came up with, and it simulates quite well.



The LED should have an approx. 2.1V voltage drop (so an orange, yellow or green LED should work fine). Sorry, it's a white LED, the datasheet reports 3.6-4V forward voltage at 20mA, in general any InGaN LED should work.

This is the response:



and it does so with 0.1% THD for a 100mV input, growing to about 3% THD on transients (400mV, 12dB over). I also did a FET version, but that grows to over 10% THD with the same signal, so it would need attenuation on the input (and a subsequent boost) to work.

Now to design the treble section...

[iainpunk]

think about circuit order of operation.
if the bass circuit crates 1% thd, that harmonic content would get boosted by the treble control resulting in more apparent distortion, where if the treble control comes first, the problem is less prevalent.

cheers

[fryingpan]

I am in fact thinking of putting treble before bass, but anyway, 3% THD or even 6% THD on transients is healthy, it will provide some compression and some added harmonics without being very audible on bass (the Ampeg all-tube SVT is rated 300W RMS at 3% THD, for example).

[iainpunk]

ohw, yea, for bass, you can run around 7% THD before it starts to noticeably sound dirty.

cheers

[fryingpan]

In the end, I avoided using "discrete opamps" (which come with their caveats) and I came up with the following schematic, getting some inspiration by the Dan Armstrong Purple Peaker for the treble section:



This is the treble response (broad peak filter at 2kHz, with a treble shelf half the intensity of the former after that; many treble controls on bass preamps boost much higher treble, and I find it's mostly pointless unless you're looking at very modern sounds):



and this is the bass response:



Treble can be boosted by about 9dB, bass can be boosted by about 6dB (the very broad filter is centered at about 120Hz) with a corresponding 4dB cut in the mids @ approx. 300Hz (with the cut, the bass boost ends up being centered at 40-60Hz, getting to 0 boost at about 160Hz, aiding the lower octave a bit more but being significant all over the bass guitar's range). The range is deliberately shallow as larger boosts and cuts on bass and mids are largely pointless on bass (which is the problem with the "contour" switches on many amps, they're too heavy-handed).

LTSpice projects noise at -90dBV (which is more than fine, considering that the input signal will be in the whereabouts of -20 to -26dBV). At 400mV peak, distortion is in the whereabouts of 1% with the controls dimed (mostly 2nd and 3rd harmonic, it would appear), which is basically inaudible on bass, and could grow a bit (as it is now, the input signal could get to about 1V peak with little or no audible distortion). The distortion will be more prevalent in the bass octave, which would be bad if it were higher, but below 3% THD distortion in the bass octave is usually of no consequence (and it's mostly low order harmonics anyway, which on a harmonically poor signal such as bass guitar is fine). Power consumption is a bit high for an onboard preamp (about 10mA), but I have about 10dB margin for noise and I can get away with more distortion, and I could rearrange impedances to drop power consumption by half or more (there is also an extra buffer that could be removed, in the bass boost/mid cut section, but I left it because it makes matching the boost with the cut easier). Also, I could rework some of the coupling capacitors to avoid large values (if wanting to avoid using electrolytics) and the design could do with an input trim. The output trim is before the final FET amplifier, which adds noise, it can be easily moved later (and reduced to a much smaller value). I opted for mixing the filters with the clean signal as opposed to modulating the intensity of the boost themselves because there is no easy way, with discrete designs, to avoid frequency and Q shifts, and the distortion and noise is quite low anyway for the intended purpose.

Of course, all of this could be done with four or five opamps with less distortion and *maybe* less noise (low power opamps are not the quietest, eg. TL062, 4558) but I wanted the challenge of using only discrete components.