How to eliminate heterodyning in my 301 Flanger? Grrrr...

[Morocotopo]

 Hi, I built a 301 flanger as per tonepad´s project, and used a 9V to 18V charge pump (as found in Geofex) inside the box to get the 18V needed by the pedal. But I get heterodyning noise, found thanks to markusw help. Not really sure what that is, I believe it´s the charge pump frequency interacting with the LFO clock frequency, and generating sum and differency frequencies that get in the audio path/frequency range.
Now, the question is, anybody can help me with how to get rid of it? Would shielding the charge pump PCB help? Something like enclosing it in aluminum foil connected to ground? Or....?
I REALLY want to solve this, I´ve been working on this flanger for months!
Maybe even an enlightened mind would offer a good explanation about what the hell heterodyning is/works.

Thanks in advance

Morocotopo

[MartyMart]

To "heterodyne" is to generate new frequencies by mixing two or more signals in a nonlinear device such as a vacuum tube, transistor, diode mixer, Josephson junction, or bolometer. The mixing of each two frequencies results in the creation of two new frequencies, one at the sum of the two frequencies mixed, and the other at their difference. A low frequency produced in this manner is sometimes referred to as a beat frequency. A beat frequency, or "beating," can be heard when multiple engines of an aircraft are running at close but not identical speeds, or two musical instruments are playing slightly out of tune.

I googled this, as I had not come across the word before, seems like you understand what it means in your post
I have 2x charge pumps, which give me 18v from 9v , using MAX1044 chips but put them in very small alu boxes.
Perhaps wrap the MAX section in Alu foil, ground it direct to box and separate from circuit ground, also put perhaps
another large cap to ground at "it's" 9v + connection ?
I think this would prevent the interaction.
MM

[markusw]

What also probably might help is sticking a LP filter between +18V and the main pcb (i.e. a 10 ohm resistor in series followed by a large electrolytic (few hundred µF) in parallel with a 100n to gnd). In addition decoupling the two gnds with a 10 ohm might help too.


Markus

[Mark Hammer]

One of the problems of the MAX1044 is its convenience.  Huh?   

Powwer upverter like the 1044 and others, provide a sort of "mini-AC source" by creating a high current clock pulse which is averaged out by the diode/cap ladder on its output.  The MAX104 absolves the user/builder of creating a HF clock circuit and does all that work itself.  Unfortunately, that also means the user has no choice in what that clock frequency IS, and at least in this one instance it would appear the fixed clock frequency is all wrong for the application.  Perhaps a better route to pursue is to use a CMOS invertor in a discrete clock like the way it is done in the upper right hand corner here - http://www.paia.com/tubehsch.pdf - and set the clock frequency to something that will generate less audible heterodyning byproducts.

[puretube]

dunno - but I`m afraid this kinda supply will also
rain into the rails
just like that well-known LFO-ticking does...

[Mark Hammer]

The power upverters are clever elegant solutions to certain problems.  Personally, though, I'm more the type that would prefer to either use two batteries or else start with a higher-voltage multi-purpose power supply and regulate downwards as necessary.  I might point out that the FL301's design is intended to provide a stable 12v out of 18v as the two batteries start to peter out.  The pivotal factor there is that whatever the BBD bias trimpot is set to, that should remain the valid bias voltage for a long time.  My sense is that, if an upverted 9v battery dies faster, that bias voltage will not remain valid quite as long.  Alternatively, maybe the smart thing is to start with a decent 12v wallwart and skip the power-conditioning stuff altogether.

[Morocotopo]

 Thanks for the suggestions/reflections, will try them as soon as I have time. I want it to work with a single 9V batt and the charge pump, I don´t want to have to use a separate power supply just for one pedal, or 2 batts and not have the possibility of PS power. Who knows, I might be able to achieve my goal, I think, around march 2008 and after about 325 threads/cries for help here... (I started work on this pedal several months ago...) Oh, the joys of electronics...

Morocotopo

[Morocotopo]

 Couldn´t wait, and stayed up late to try to finish this thing once and for all...
The LP filter suggested by you, markusw, worked, it reduced the heterodyning noise quite a bit (I put 5 100 uF caps in parallell to get 500 uF, quite funny looking...). Decoupling the ground(s) helped also, but made the bypass switch pop, so I left them out. Maybe because I have the switch setup to ground the input in bypass, and/or maybe because the resistors cause a different ground potential between different parts of the circuit?
So, It´s not perfect, still a little bit of noise, mostly when the rate is highest (a low, muted "bump" that turns into a pitched low noise when the rate goes into the audio range) but it has improved quite a bit. I´m gonna leave it at that, or I´ll never build anything else!

Thanks to all for the help. I hope to be of help too for someone, as I learn more and more in this great forum.

Morocotopo