I found the issue. I’ve noticed that when I move a Note Fader on the meloDICER, it’s sending Program Change to the MD (Notes are mapped to Patterns). So, I went to the Map Editor on the MD and turned everything off.
Two questions still remain :
Why Program Changes are passed from the MIDI IN of the OT to its MIDI out (meloDICER is going to the MIDI IN of the OT)?
Why reverting to X.04 fixed the issue? Was there no mapping at all in the Map Editor of the MD?
What are your settings for PRG CHANGE in the CONTROL menu of the MD?
Prg in, ON
Prg out, OFF
Channel, 16
Maybe try changing to another global and changing back. Could be the “cache” is wrong/corrupted.
Very good.
did this resolve in your snapshots?
thanks!
I tried setting up 2 new global settings and reverted back to the original with no luck.
Are you using the AutoChannel on the OT for the Send Program Change setting?
My octatrack is set to channel 16. I tried setting the octa/md to base/auto and that didn’t work either.
I think i’ve identified the issue. The X.05a patch will likely ignore the PRG change message if the channel is outside of Base Channel range.
I think we got it! Thanks again Justin 
i don’t think program changes trigger the map editor, like you say it’s notes that came relentlessly from the meloDICER maybe causing the freeze. this is a guess but the OT must be sending out the input from one of it’s tracks that’s being triggered.
this maybe be related to the X.05 patches regarding Base Channel range. depending on what OT channel the meloDICER was triggering and what channel that was sending notes out of.
Love that my oldest device is getting the most updates. Thank you so much.
The Machinedrum was already a mighty unit, and with each of these incredible releases its power continues to grow. Many thanks to the devs and testers for all your hard work!
Yeah, I reinstalled the firmware recently and haven’t seen the issue crop up.
Nice, maybe update your posts so it doesn’t confuse people scrolling through.
The fact no one else reported this it’s not helpful to leave them labelled as bugs, but useful to state how you resolved it if anyone else experiences. 
Just to update on this. I had a similar issue to the one @Roger one stated. The machines were silenced in some of my old patterns, with no apparent way to get the sound back. I haven’t had much time to look into it and I haven’t reinstalled. I think I went and made a new kit just to test out the NFX. So there maybe issue that only some of us have had?
I think what happens is the tracks you had neighbour machines on, after changing kit they remain that way until you trig them. The same in stock firmware and input machines.
@JustinValer
Thanks for the new FW, I would like to report just one odd thing that happened yesterday with X.05: I was playing with 4 machines enabled, stock FM BD (ch1), TRX-SD (ch2), on HT(ch3) a GND-PU and an FM-HH (ch10).
None of the new X.05 machines were loaded.
I twisted the BD (ch.1) FLTF + Function to globally alter the value temporarely, and once reached a certain level of turning, the complete pattern muted, I then checked the Synthesis page and for all channels the machines were unloaded.
Pattern trigs and locks remained in memory, but all the machines disappeared.
I don’t know if this thing was ever reported before, I still wasn’t able to reproduce it, but I’ll take a more precise note of everything in case it will happen again.
just doing some ground checks here; did you func+ext to reset after the modulation? or this happened before then? was the kit saved?
if it wasn’t; it would load an empty kit on func+ext. which sounds like what happened as your programming remained.