Choosing ESCs for multimotor setup

This is for Tony
I recently got involved with a quadcopter and needed 4 ESCs to work the motors I already had. No problem, off to Hobby King and order 6 TowerPro ESCs. When I got them I discovered I couldn't get a proper startup sequence on 5 of them! I then discovered they're knock offs! They're marked TowardPro! After scratching around for a while, at Phils insistence I put the ESC signal coming from the RX on a scope, and discovered the signal was always the same! At or about this time I discovered the Throttle cut button on my Futaba T6EX! Using the trim slider I got the trim all the way down to +120 (I presume this is the number of bits required to get down to the bottom of 1024) Bingo! I now had 6 ESC that worked, at this stage, had I used my head, I would have saved myself an aweful lot of scratching around!
I now carried on building (assembling actualy) my quadcopter, with all the wiring done, it came time to switch on....................................You have never heard such a cacophany of sound in all your life! By dint of disconnecting the ESCs one by one I discovered that,in some cases the output signal from the quadcopter control board wasn't getting short enough to trigger the safe start conditions for the ESC, hmm, where to go from here?

I then hit on the idea of using the trim on the TX to start and then stop each individual ESC and motor. I had at this stage pulled everything out of the quadcopter so it was easy to do it on the bench. I had 7 ESC which I then linked in turn to the same motor and RX. I then used the throttle trim to just start the motor, then just stop the motor, noting the trim reading on each. This is what I found, to make things easy all readings have been referenced to zero, ie zero is full closed throttle trim.

ESC no, start at, stop at
1 52 44
2 56 48
3 40 32
4 44 36
5 148 140
6 52 44
7 48 40

As you can see, ESC no 5 needed a massive amount of trim to start, and this is the one that always worked, obviously to use it in a multi motor setup is an absolute no no, the revs from theESC will always be 10% low.
As for the rest, the total spread is 12 bits, which as a % of 1024 is 1.17%, enough to be easily trimmed out with rudder, although I would be tempted to use ESC 1 & 6 as the outers and 3 & 4 for the inners.
I think the moral of this story is, check your ESC start points! If they're too far out, chuck em back!

I would think that a slight mismatch of esc's on a fixed wing 'plane would be much less critical than on a multi-rotory-whirly-thing, where balance is everything, to achieve a stable hover.

I think what I'm trying to say here is this, we spend hours getting the alignements right, 500mm from wing tip to fin or more is not uncommon, we fiddle about for ages to get wing tips to fin within 1mm, that's .5%. I'll guarantee that Tony will have that Lancaster lined up, always to Sunday, and it'll be within .05%! Then we fit a set of ESCs that haven't been checked, and can potentially give you 2% thrust difference from Port to Starboard! Bottom line, give yourself a chance, it's not hard to do.

Eric wrote: I would think that a slight mismatch of esc's on a fixed wing 'plane would be much less critical than on a multi-rotory-whirly-thing, where balance is everything, to achieve a stable hover.

You're probably right Eric, I'm just tub thumping! I also have a roll gyro and a pitch gyro!