Replica Tap Tempo + Buffer = Plop, Plop, Plop

[karbomusic]

I built a couple of MPF102 buffers which are taken from the 2nd circuit on the AMZ site and built them into 1090NS inline boxes:



They work great, love em! However, last night I was using one on the same 9VDC rail and feeding T-Rex Replica delay while testing my new 9V/12V power supply I built. I noticed that only when these two devices are connected I can hear the tempo indicator plopping into the audio signal. As an FYI the Replica delay tempo is always on and flashing even when the delay is bypassed so you can set it before engaging the delay. At first I thought it was my power supply then I switched to another supply and same issue. Use two separate 12V supplies, no issue.  

I then did a quick test using a TMO18 transformer which I wired to the output of the buffer and that immediately fixed the issue but, I also immediately lost the bass response; had it not been for that it would have been easy to just drop it into the 1090 box just before the output and be done with it. Drats! I have pretty much reduced it down to the buffer box, if it isn't in the mix, the problem is non-existent; lots of other pedal combos, no problem. This brings me to the biq quesiton. What can I do to remedy this in the simplest way? I can't think of anything that is inherently wrong with the buffer but it does seem to be the straw that breaks the camels back and I don't want to start tearing it apart until I know more about who/why/what/where.

Points of interest:

1. Since the one power supply is 9/12V and not isolated the issue is much reduced (but still there) if I run the buffer at 9V. This isn't about the isolaton per se, just wanted to point out it is less at 9V.
2. The only modification to the schematic above is that I used a 10k resistor instead of the 3.3K. Unfortunately, I don't remember why.
3. The level of the plop is just above the noise floor so it isn't that loud, I could possibly deal with it at a live gig, not so much in the studio and it's just enough to be annoying.

Any ideas? I can solve this by isolating the power supply but since it only happens with this single combo and I am the one who built one of those, I want to make sure there isn't something I can solve or if there is something about the buffer circuit I can address. Additionally, it seems like a good excuse to post and learn something new I can apply in the future.

EDIT: Due to the transformer test working could this possibly be a weird incarnation of a ground loop?