This Week on the Breadboard: The 'lectric Mama Flanger

[Chuck D. Bones]

The 3-day course I attended was a subset of that material. Part-way into the first day, I turned to the engineer sitting next to me and said "You know, I was already doing some of this stuff and I just chalked it up to laziness." He said "Yeah, me too." It felt good to have our "lazy" techniques validated.

I'm going to download all of this stuff, thanks!

 

[fig]

Premise

"Not much of what you learned in school turned out to be much use"

 

[Chuck D. Bones]

So true, as least as far as electrical engineering was concerned.

At one point in the lecture he made a disparaging reference to "rotton old tubes" and then in the next breath allowed that they were a hulluva lot more linear than transistors.

 

[jubal81]

ED even has an LFO and BBD clock VCO in one chip. Begs the question why I'm messing around with this analog shit.

('cuz I wanna, that's why)

OMG, you're right. I forgot all about his flanger project.

 

[Chuck D. Bones]

I finally got the diode sine shaper circuit working the way I like.



Any further developments will employ the Flangelocious LFO/clock chip from Electric Druid. That chip will replace the MN3102 and everything in the bottom half of the schematic. The only extra part it requires is a +5V reg. Credit where credit is due, it was some design notes on the Flangelicious page that set me on the right path for tuning the sine shaper above. Turns out that a sinewave is not exactly what we want. What we want is a "bent triangle" wave. Notice that the waveform below does not really flatten out at the top and bottom like a sinewave does. We don't want the LFO sweep to stall at the extremes, which is what happens with a sinewave LFO. Backing off on the shape trim makes the waveform a little more pointed at the top and bottom.



The shape trim lets us fine tune the "pointiness" of the waveform and compensate for variations in diode Vf. The range trim lets us maximize the sweep range. The waveform above was measured at pin 3 of the RANGE pot. This is the largest control signal that the clock VCO can accommodate. As I mentioned in an earlier post, U4 can be a TL072. The only downside of using a TL072 is the amplitude of the waveform above will about be 2V to 3V smaller because the TL072 output cannot swing to the rails. There is still plenty of sweep range available. The only trimmers that are critical are the blend trim and color trim. The others could be hardwired with only slightly diminished performance.

I should add that this thing has a HUGE sweep range, much more than commercial flangers. At the slowest speed, approx. 0.011Hz, and a high range setting we get a nice chorusing effect. For flanger effects, I turn the RATE up and usually set the RANGE below 10:00. I might try a log-taper pot there. The waveform above is max LFO speed. I had to use a stopwatch to measure the slowest speed because my 'scope doesn't go that slow.

 

[Big Monk]

I finally got the diode sine shaper circuit working the way I like.

View attachment 16227

Any further developments will employ the Flangelocious LFO/clock chip from Electric Druid. That chip will replace the MN3102 and everything in the bottom half of the schematic. The only extra part it requires is a +5V reg. Credit where credit is due, it was some design notes on the Flangelicious page that set me on the right path for tuning the sine shaper above. Turns out that a sinewave is not exactly what we want. What we want is a "bent triangle" wave. Notice that the waveform below does not really flatten out at the top and bottom like a sinewave does. We don't want the LFO sweep to stall at the extremes, which is what happens with a sinewave LFO. Backing off on the shape trim makes the waveform a little more pointed at the top and bottom.

View attachment 16228

The shape trim lets us fine tune the "pointiness" of the waveform and compensate for variations in diode Vf. The range trim lets us maximize the sweep range. The waveform above was measured at pin 3 of the RANGE pot. This is the largest control signal that the clock VCO can accommodate. As I mentioned in an earlier post, U4 can be a TL072. The only downside of using a TL072 is the amplitude of the waveform above will about be 2V to 3V smaller because the TL072 output cannot swing to the rails. There is still plenty of sweep range available. The only trimmers that are critical are the blend trim and color trim. The others could be hardwired with only slightly diminished performance.

I should add that this thing has a HUGE sweep range, much more than commercial flangers. At the slowest speed, approx. 0.011Hz, and a high range setting we get a nice chorusing effect. For flanger effects, I turn the RATE up and usually set the RANGE below 10:00. I might try a log-taper pot there. The waveform above is max LFO speed. I had to use a stopwatch to measure the slowest speed because my 'scope doesn't go that slow.

I’ve been looking for something that gives me the classic Electric Mistress sound but with unity volume considerations and a blend control. This might be it!

Either that, or put a blender pedal on my Mooer E-lady.

 

[JamieJ]

Any chance of this becoming a boneyard build @Chuck D. Bones ?

 

[Barry]

Any chance of this becoming a boneyard build @Chuck D. Bones ?

I added it to the wish list

 

[Chuck D. Bones]

I have one more thing to try on this design and that is using the Flangelicious LFO/Clock chip. Time to pull the trigger on that Mouser order so I can have some microcontrollers.

 

[fig]

Oh goodie! I love shopping. Can I push the cart? Can I? Can I?

 

[Chuck D. Bones]

You can ride in the cart if you promise not to stand up.

 

[jubal81]

I have one more thing to try on this design and that is using the Flangelicious LFO/Clock chip. Time to pull the trigger on that Mouser order so I can have some microcontrollers.

Sweetness! Exciting stuff.
Any chance you could explain how you're doing the programming for those of us less digitally inclined?

 

[Chuck D. Bones]

I'll be doing it for the first time, so I should probably make a youtube video, including all of the fuck-ups.

The initial setup includes buying a USB programmer on eBay, installing programming SW from MicroChip on my PC and downloading the hex file from Electric Druid. So far, I have only accomplished the 3rd task.

After that, install the chip in the programmer, plug in the programmer, run the programming SW, load the hex file and press the PROGRAM button.

 

[cooder]

Such a cool project, can't wait for the version with Electric Druid chip!

 

[Chuck D. Bones]

The deal with ED is their code is free for personal use. If you want to go commercial, then you either buy programmed chips from ED or pay a license fee and program your own.

 

[danfrank]

I think they use Microchip PIC ICs. I'm surprised ED is giving out the code, it's very nice of him! These PIC ICs have an option where the programmer can "burn" the inputs to the chip after programming so the IC can't be copied but is still able to execute the code that was programmed into it.
Back in the '90s these type ICs were used to trick cable boxes into "test" mode so the box would decode all cable channels. What a hoot!!

One more thing... When using the ED (no! not that ED!) chip which has to be run at +5 volts, the 3207 also has to be run at +5 volts unless the clock signal is boosted in amplitude to the voltage that the 3207 is using.
Thanks for this thread Chuck, a very interesting read!

 

[Chuck D. Bones]

That is correct, sir!

 

[Chuck D. Bones]

I've got this flanger running with the ED 4KNOBFLANGE chip. Works well except I'm fighting a noise issue. There's some slight ticking that's tied to the RATE and DEPTH control settings. Goes away if I disconnect the output of the BBD. Standby for further developments.

 

[cooder]

All systems on stand by ready to rip...

 

[Chuck D. Bones]

[edit]
Thought I fixed it. I rearranged the power distribution and grounding and increase the filtering on the BBD output a little. Here's how it looks now. The 5V reg is hiding behind a capacitor to the left of the LFO/CLK chip. Compare the parts count to the all-analog LFO & clock gen. The RATE, DEPTH & MANUAL controls have a huge range. Only two trimmers req'd. The RATE, DEPTH & MANUAL are voltage-controlled, which means they could be hooked up to an expression pedal or other control source. I'm calling this close but no cigar. I want to spend some more time playing it and reviewing the circuit design. I'm considering the merit of adding an output buffer. This circuit is layout-sensitive since the 4KNOBFLANGE chip puts digital noise on the power and ground rails.