Use an additional VCA in ADA flanger?

[StephenGiles]

When my unit sweeps to heights hitherto never imagined, the volume drops off considerably - all part of the effect of course, but what if a VCA was placed on the wet signal, controlled by the LFO? I'm just thinking aloud here, well - with my 2 typing fingers, perhaps the 4007 could be further utilised for this so that the level is maintained as the flange sweeps up.

Downside of this could be noise, although that may be for the better as flanged noise sounds good to me. So leave the factory, leave the forge and dance to the New St George and let's have some thoughts on how to do it.
Stephen

[gez]

Don't have the schematic, but you might be able to send the clock to the input of a 4046 wired as a PLL then tap the control voltage (buffer the cap used in the loop) to send to the OTA.  The PLL would act as a (crude) frequency-voltage converter.

Downside is it's not that linear, but I've used it in (breadboarded) circuits and it's a cheap alternative to custom f-v chips.  Would have to consider the capacitance of the PLL input though.

[DiyFreaque]

Hi Stephen,

Well, if you want to find out if a VCA would be worth it, maybe you could try this simple tact first:

There are basically two outputs of the SAD1024 that are unused in Mike's rendition of the A/DA, and those two outputs should be identical to the outputs that are used (because of the parallel mux mode of the SAD1024).  If you were to tie those two outputs together with a clock null trimmer, then run those through an LDR/LED combo to the point the the other trimpot goes, I would think you could vary the current on the LED to determine how much of this 'second' signal is mixed in.  Or set it up as a sort of voltage divider, too.  Just a thought for experimentation if anything.  Guess you could do the same thing with a pot wired as a variable resistor or voltage divider, too just to see how it sounded (as long as the LFO sweep is slow, wouldn't want you to get CTS just testing if it would work with a fast LFO   ).  The LED/LDR optocoupler would make it so you could maybe automate it with the LFO signal....

I mean, heck, the original A/DA mixed all of the outputs together, anyway....

Cheers,
Scott

[DiyFreaque]

Forgot to mention you could control the LED current with a simple NPN tranny sink.  Run DC into the base with an offset pot/resistor to find the middle point of the LFO sweep and run the LFO through another pot/resistor to set the amplitude of effect the LFO would have on the LDR's resistance.  Then tie the anode of the LED to +V and the cathode to the collector of the tranny, with a safe, non-LED popping current limiting resistor from the emitter to ground (it would also serve the to limit the resistance range of the LDR).  If you wanted an expo curve, a PNP/NPN combo would work, too.

Cheers again,
Scott

[bwanasonic]

I always loved the heavy compression/post flanger tone of Frank Zappa's "Ship Arriving Too Late To Save A Drowning Witch". I always thought of this as squishing the peaks of the flanger peaks, but wouldn't the reverse apply as well? IOW - Aren't you describing post flange compression?

Kerry M

[Mark Hammer]

I forget the particulars, but this was something that Frank described as sounding "pillowy" in one of his Guitar Player interviews.  I seem to recall he used a Wasatch flanger, though I forget what exactly was different about it or what he did differently with it.  There may have been some external CV triks with envelope followers, or some highpass filtering or something, but again don't rely on my memory.  Quite probably, that interview is posted somewhere. If not, it is in my basement....somewhere...in one of the many boxes of GP back issues.

[Vsat]

Steve,
The volume reduction is probably due the  increased insertion loss at higher clock frequencies. Should be a graph  of this in the SAD1024 spec sheets. If you compensate for this, a benefit will be better notch depth at shorter delays. One method (going overboard for a foot pedal) could be to use a 2nd SAD1024 to monitor the amplitude of a 1000 Hz "pilot" signal. SAD #2 is clocked in parallel with SAD #1... the 1000 Hz tone is fed to the input of SAD #2 only... a level detector circuit watches the level of the 1000 Hz signal coming out of BBD #2, and sends a control voltage to a VCA connected to the output of BBD #1. If you want, could also put a VCA on the output of the pilot BBD, and  build a servo loop to maintain the  1000 Hz at a constant level. This technique is complex, and would really only be considered where it is essential that maximum notch depth be maintained for studio gear. For the purposes of most listeners, just a simple VCA tied to the sweep CV should be adequate.
Regards, Mike

[StephenGiles]

Thanks everyone, I'll mull this over for a while.
Stephen