First off, i have concluded that Bluetooth (v2) is quite power hungry and am probably going to end up abandoning it and switching to using the nRF24L01 modules.
However, i encountered a weird situation re powering the circuit and that is the main reason for this question.
The issue is at the power supply end, i don't think there is anything amiss elsewhere but the whole setup looks like this;
FWIW: the setup works with power from the PC-USB through a 3v3 USB-to-TTL adaptor.
But when i try alternative power supplies, it fails - my conclusion is insufficient current for (or because of) the HC-05 - which is why i will switch to the nRF24 option.
The following are the various options that i had tried just for completeness sake, but the key point is the last one, #5 ; and of course #1 as reference point of what did work.
FYI, i break the circuit (for current-measuring) at the top-left part of the schematic - ie. where the power supply goes to the power rail.
Why is the circuit "stable" (ie. servo does not jitter) when i'm measuring the current ? (Is the DMM acting like some kind of buffering capacitor - i had already tried a 10uF and even 1000uF which didn't make a difference, well - slightly less of a jitter)
EDIT: just to clarify further - when i'm measuring - means the servo jitters when the circuit is left alone, but when it is broken and the DMM is inserted (to measure the current) the servo doesn't jitter anymore - what is happening here ?
EDIT2: forgot to add that Servo ground is also tied.