Low Pass and High Pass filters in one stompbox

[PRR]

The "Gain" control seriously interacts with the low-pass tuning.

Wire "gain" as a 10K pot to-ground, usual Volume affair.

Not sure why it needs a gain knob.

[ElectricDruid]

+1 agree with PRR. I'd drop the gain control and there two caps associated with it. They're causing trouble and they're largely unnecessary. If you want it louder, turn the amp up. If you want it quieter, turn the guitar down. Simple!

[tubegeek]

If you want it louder, turn the amp up.

Fine advice in ANY situation.

If you want it quieter, turn the guitar down. Simple!

Um, are there any guitar players who can even UNDERSTAND that sentence?

[MrFrog]

Hi all,

It takes me a little while to find time to work on this, hence the delays.

I remove the "Gain Control" (both capacitors and the variable resistor) and then wired the "out" from the high pass section to the "in" of the low pass section. I did this by moving the OUT from row I to row B on the vero diagram.

There was no change in the response from the circuit. Using the audio probe I get audio everywhere except for the red marked area on the eariler picture.

I have replaced both the 1nF capacitor and the TL072 once already so I don't feel like it is either of them (but happy to try again if that makes the most sense).

I checked the legs of the TL072 have continuity with the veroboard so that should be ok.

Is the only thing left to do to change the capacitor and the TL072 again on the off chance the replacements were faulty?

Is the 10uF capacitor on the output somehow messing things up?

[noisette]

Didn´t read the rest of the thread, but that gain control is weird in that isn´t really gain, but it influences the first pole of
the lp. Unless it´s other end is connected to gnd (not what is shown), then it´s an attenuator, but still skewing the filter
response. Better to wire an output level potentiometer, like 10k to 100k log...
If you want a gain boost over full freq bandwith, for whatever reason, you would need to employ another opamp-

And don´t give up, those filters will be very useful!

EDIT:Should have read the thread. it´s been adressed already, nevermind

EDIT2: if you put a gain stage behind, you could configure it as a summing amp and have the filters switchable between
series and parallel, that way you´d get a bandpass and a bandstop(notch) filter! Just a thought

[ElectricDruid]

Hi all,

It takes me a little while to find time to work on this, hence the delays.

I remove the "Gain Control" (both capacitors and the variable resistor) and then wired the "out" from the high pass section to the "in" of the low pass section. I did this by moving the OUT from row I to row B on the vero diagram.

There was no change in the response from the circuit. Using the audio probe I get audio everywhere except for the red marked area on the eariler picture.

I have replaced both the 1nF capacitor and the TL072 once already so I don't feel like it is either of them (but happy to try again if that makes the most sense).

I checked the legs of the TL072 have continuity with the veroboard so that should be ok.

Is the only thing left to do to change the capacitor and the TL072 again on the off chance the replacements were faulty?

Is the 10uF capacitor on the output somehow messing things up?

Do you have clean highpassed audio at the output of the first stage?

Anyway, it doesn't matter - I think the only remaining thing we need to look at is the biasing. Your design has no midpoint voltage anywhere. The two "ground" points in the highpass filter should go to a "virtual ground" midpoint voltage. Alternatively, you could run the thing off two 9V batteries and make a bipolar supply, but that's not the usual way for pedals.
The "ground" point at the bottom of the 470p in the Lowpass filter could be taken to the midpoint voltage too, and in a bipolar supply circuit it would certainly be, but since it's a capacitor and no DC can cross it, it doesn't actually matter, and you can leave it as it is.

HTH,
Tom

[MrFrog]

Do you have clean highpassed audio at the output of the first stage?

Yes it seems that after the signal goes in to the TL072 on the lowpass side it kind of disappears (which is why I initally replaced the TL072). Although for testing I am using one of the little vox amplugs so its hard to hear if the highpass working through the earbuds. Usually I get the stompbox so the output is working then drag an actual amplifier over to where I work on the stompboxes and do a final test of the 'effect' itself.

* Midpoint voltage and biasing *
If there is an idiots guide someone can link to about this to this that would be great. I think I have pieced together what this is. From a laymans perspective what it looks like is normally a 9v supply gives you between 0 and +9v. The idea is that you use a 'magic box' to change this in to -4.5v (which is Vref) and +4.5v.

As far as 'magic boxes' go I think these are referred to (or are part of the family of) voltage dividers.
I thought if I could find some schematics or diagrams I might see how to fit it in to my current design.
I ended up on these two pages:
http://www.runoffgroove.com/splitter-blend2.png (the bit with two caps and two resistors.).
https://www.diystompboxes.com/smfforum/index.php?topic=116491.0

* What I think I need to do next *
If someone can let me know I am on the right track with this next bit that would be great.

I was going to use the ProcoRat voltage divider since despite being a bit more complex has a diode to help stop me from accidentally blowing up my stompbox, I can also understand more of items in the diagram.

Then in my stompbox I use the Vb and Vc instead of 9v and Gnd (I assume Vc is Virtual Current and Vb is virtual bias?)

* Final question *
I saw this comment on this page (http://www.....org/viewtopic.php?f=41&t=13878):

If they are all powered from the same power source, you should be able to use just the one 1/2 Vref for each circuit provided that it's fed to each op-amp via a 1M resistor, you could even use a spare op-amp to buffer the 1/2 Vref so that it doesn't get too heavily loaded.....

Should I have one of these 'Vref' magic boxes for each IC? If so I might just use the simpler one from the Runoffgroove schematic otherwise this stompbox will be the size of my house by the time I am done!

Thanks for all the help so far!

[ElectricDruid]

Nice work on the searching! You've got loads of good stuff there!

I would suspect that you can use that Proco RAT power input circuit that you've found exactly as is (below). If not, you'd probably only need to reduce R13/R14 (Try 47K or 10K, or even 4K7) to get the midpoint voltage more in the centre.



That Geofex link describes why this is necessary and provides the theory:

http://www.geofex.com/circuits/biasnet.htm

Ok, so if you take the Proco RAT input, take the three ground points in your schematic to the "VB" bias/reference/midpoint voltage.



The only things that go to the battery ground/0V are the op-amps pin 4 power supply, and the input and output jack's ground.

That should get you working. Good luck!

Tom


Edit: Sorry, didn't say - you can use the one "magic box" Vref supply for everything. There's not often a need to have separate ones, although occasionally it's a good idea.

[Babajoux]

Just wanted to know how it turned out?