DIY feedback looper - interesting 'negative/self-cancelling' problem

[cactus friend]

Hi all, and apologies if this is the wrong forum, but it does have to do with a DIY pedal...

Some months back, I built a feedback looper similar to the one in this diagram.  I really loved the way it interacted with my friend's MXR Carbon Copy delay.  I could use it to cause the repeats to just wash over and over and start breaking up and sounding ridiculous until you couldn't stand it anymore and had to disable the feedback loop.  However, I had to return that pedal to him, and so I bought myself a Malekko Ekko 616.  For the most part, I like it better than the Carbon Copy...however, here's the problem: when I turn the feedback loop's pot up to where the repeats would start swelling and getting wild with the Carbon Copy, the signal actually attenuates to the point of disabling the delay!

My question is: does anyone know what's different about the Ekko that might cause this?  Is it because of the buffer?  That's selectable, and even with it disabled, this happens.  Does the pedal have some sort of 'protection' against this?  I can cause it to self-oscillate with no problems (I'd just prefer being able to control it with a momentary footswitch which I plan to build into my next looper).

Thanks for taking a peek at this - here's a video of the Carbon Copy and my looper creating the 'desired effect'.

[Quackzed]

sounds like an 'out of phase' issue... some pedals invert the signal output, so reintroducing an inverted signal back to the input(un-inverted) will result in cancellation of the initial signal-or some part of it, rather than 'addition' they 'subtract'.
you can either build a little circuit inside the looper that will allow you to switch the signal phase after the return jack from the 'offending' pedal... search 'splitter blend' -u2a in that schematic is an example of the fragment you'd need...
or you could add that little circuit into the delay itself- if theres room-...
-another option would be to use another pedal that also inverts the signal and run it after the delay to invert the signal twice, so that its back 'in-phase'.
i'd probably just build the u2a part of the splitter blend from 'green return' to the just before the 'blend control'... tack it in to your pedal so you can 'un-invert' the signal via the switch if you need to for certain effects...
or find another buffer type circuit that you prefer that inverts the signal and just use a dpdt 'true bypass' type wiring to 'insert it after the return...
http://www.runoffgroove.com/splitter-blend.html

[cactus friend]

I could hug you.  Thanks so much!  A phase issue, of course.  What I will do is add a phase toggle to the next version of my feedback looper!  Thanks again!!

[cactus friend]

Hm, one more dumb question.  An actual phase switch (accomplished by a DPDT toggle switch) wouldn't be enough, would it?  I actually have to invert the signal with a buffer?

[defaced]

Depends on how you handle grounding and power supplies and such.  It could get nasty, or work great.  An active solution (opamp, inverting stage, etc) would probably be safer.