question on removing the envelope detector on an envelope filter

[Marcos - Munky]

I built a DOD FX25 a few days ago, and it's a fun effect to play with. What I liked most of it it's the filter sweep. Just by curiosity, how could be the envelope part removed and the filter be controlled as a regular wah?

Schematic for reference:


I believe the envelope part is the circuit at the lower part of the schematic. Would a pot from r12 and r13 to ground do the job? If so, any guesses on the pot value?

[Josh?]

Before I Start:

DO NOT TRUST ME BECAUSE I'M NEW TO BUILDING PEDALS

Now that that's out of the way, the bottom section looks a lot like the part of the DOD 440 (another envelope filter) which provides a control voltage for its LED to do things to an LDR.  Here, however, it looks like the control voltage is going straight to those op amp-looking things.  To make a cv, you Might be able to use a pot as a voltage divider to determine how much voltage the op amp-looking things get. It would probably look like this, but I have no idea of what the values should be, or even if I'm on the right track:




Hopefully someone smarter than me can help you out, it sounds like a really fun project!

[Keppy]

Josh's suggestion would work, but the pot would probably get scratchy due to a lack of smoothing. Since there's already a smoothing capacitor in the circuit, we might as well use it.

The key here is that the bias inputs of the LM13600 are current inputs, and the voltage at those inputs is tied at 2 diode drops above ground (about 1.2v). Since the 9v supply is a fixed voltage, and those inputs are a fixed voltage, there is a 1M/5k voltage divider created by R9 and R12/13 (these are kinda-sorta in parallel, since they connect to the same voltage).

Why the 10k resistors? The 13600 is spec'd for up to 1mA current on those inputs (with definite smoke above 2mA). Those are the safety resistors, to make sure that only 1mA can get to those inputs. Don't mess with those.

So, in order to do what you want, replace R9 with a pot wired as a variable resistor. The value's right there in the schem, 1M. Since the opamp can't drive C7 all the way up to 9v, you probably want the pot in series with a fixed resistor to keep the range similar to what the envelope filter gives you. Maybe a 500k resistor 500k pot combo. C7 will help smooth any scratchiness you get from the pot.

This is basically what you said in the first place...

Would a pot from r12 and r13 to ground do the job? If so, any guesses on the pot value?

... only connecting the pot to 9v instead of ground.

You also need to disconnect IC1a somehow. I'd suggest switching the cathode of D3 so that instead of connecting to the R9/C7 junction, it connects to ground. This is assuming you want the envelope follower to be switchable. If you just want to remove it, you can get rid of everything from C4 to D3.

[Josh?]

Those are smart ideas Keppy!

What's the sciencey reason a pot by itself would be scratchy, if you happen to know, and how does a capacitor solve this problem?

[antonis]

>What's the sciencey reason a pot by itself would be scratchy?<
DC..

>how does a capacitor solve this problem?<
Blocks DC..

[ElectricDruid]

A pot is scratchy because the wiper doesn't keep perfectly in contact with the track as it moves. It bumps up and down a little bit, and the resistance wobbles about as a result. If you've got DC across the pot, that wobbling shows up as noise. New pots are much better than old worn ones generally, but no pot should be expected to be noise-free.

The capacitor helps because it acts in tandem with the resistance to make an RC lowpass filter. Basically, the cap takes a while to charge up or discharge, so little spikes of noise are ignored because they're already over before the voltage on the cap has had time to change.

HTH,
Tom

[ElectricDruid]

You could have a look at the frequency control on this project too:

http://files.effectsdatabase.com/docs/diy/penfold_autowaa_article.pdf

It's a very similar thing, done slightly differently, so it's interesting to compare and contrast.

Edit: It's also worth pointing out that these state variable filters have a bandpass output after the first OTA (used in the DOD), and a lowpass output after the second OTA (used by Penfold). So you can experiment with the other output to get a different sound.

[Ice-9]

Check out the BYOC wah schematic in the link below, they have an added switch which chooses between auto and manual modes. You could make a similar addition to your circuit with just a couple of extra components.

http://byocelectronics.com/filterinstructions.pdf

Schematic is near the bottom of this build document.

[antonis]

My bad..

Didn't see "by itself"..

[noisette]

Hey, you could simply do it like the wasp filter clone by rené schmitz, which point is to do exactly what you want
https://www.schmitzbits.de/wasp.html
The one important thing is not to exceed the max. current spec. for pin 1/16 on the 13700, pin 5 on the 3080 (I think it is 2mA).
You could try if you get enough sweep by using the 10k resistors, disconnect c7/r9 from them and connect to your pot/pedal
via 1k resistor...

[Marcos - Munky]

Thanks a lot for all your answers! I got the idea on how to do it.

Here's what I have in mind. This one I built will end in a 1590A as a envelope filter, probably with a bandpass/lowpass switch. Some time ago I have built a 1590A wah using a circuit based on Colorsound Inductorless Wah, which I didn't liked how it sounds to tell the truth - yeah, I built it just for the challenge and because I had an spare 1590A. Since I liked the FX25 sweep, I was thinking of building another one to put in that 1590A wah. The wah is controlled by an led + ldr with a light screen because of the lack of space in a 1590A for a pot, so I won't need C7. Just need to play with a resistor in parallel and other in series with the ldr to get a similar sweep range.

Just a final question before the layout designing part: since I'm getting rid of IC1a, how important is to keep the IC1b buffer? Can I skip it and go right to C8 or it's a good idea to keep it?

[ElectricDruid]

Just a final question before the layout designing part: since I'm getting rid of IC1a, how important is to keep the IC1b buffer? Can I skip it and go right to C8 or it's a good idea to keep it?

I'd say it's important to keep *some* sort of buffer. The input impedance of the OTA is pretty low otherwise, so I'd say its sensible. If you don't want to dedicate an entire single op-amp chip to it, use a transistor or FET buffer, but I wouldn't leave it entirely unbuffered, personally. Some other people may not care as much and may feel differently. YMMV, etc.

T.

Edit:typo

[Marcos - Munky]

Yeah, I was guessing it may be worth to use a buffer. I was thinking of a transistor buffer, but honestly it'll probably take as much space as a IC buffer, or even more. So I'll probably go with an IC buffer. Thanks again!

[antonis]

Yeah, I was guessing it may be worth to use a buffer. I was thinking of a transistor buffer, but honestly it'll probably take as much space as a IC buffer, or even more. So I'll probably go with an IC buffer.

IC buffer has absolutely no fun..!! 

JFET buffer w single Gate bias resistor has also no fun (and it can easily ruin your day(s))..!! 

MosFet buffer w resistive voltage divider bias starts to be amusing..!! 

BJT buffer w bootstrapped Base bias resistor could be consider fun..!!   

Dual BJT buffer w multiple bootstrap is the life and soul of the party..!! 

[Marcos - Munky]

Dual BJT buffer w multiple bootstrap is the life and soul of the party..!! 

This is probably too much for so little space available in a 1590A wah

Thinking about an IC buffer or mosfet buffer, I have lots of 2N7000 here.