Rather than using a Power Distribution Board, is there anything wrong with the following setup?:
Join all ESC positive power wires together and to the positive of the battery.
Join all ESC negative power wires together and to the negative of the battery.
Connect the three wires (signal , positive, negative) from each (ALL four) ESC to the corresponding pins on the APM.
Helping myself to help myself "Each ESC has an on-board BEC, that makes four power sources on-board the quad. Pity we only connect one. If the PDB could be upgraded to include some Schottky diodes for isolation you would have quad redundancy on your power source".
So, if I followed the steps in my post, I would be OK so long as cut all but one red wire between the ESC and APM?
Whats wrong with just plugging in 4+ ubec's for redundancy... On my y6 thats what im doing.
If they aren't outputting EXACTLY the same voltage, you'll get a current loop between them that will drain your battery super fast.
And since nothing electronic is ever 100% exact, you'll always get some sort of current loop.
If you really want redundancy, then you'll have to come up with some sort of relay that will switch between the main BEC and the backup BEC when the main BEC breaks.
But then, one must ask, what would cause a BEC to break in the first place?
Having looked at the circuits of various Regulators (BEC) switched or unswitched - I can't find anything which suggests a problem with direct parallelisation (tm).
I would challenge this claim until demonstrated by experiment - however - there is nothing wrong with KISS and simply using one power supply.
That said - once you get into hexa- or octoquads which are redundant in order to prevent the obvious - I think this question deserves a better answer.
Most regulators are at-heart a switched power line - and do not have any reason to clamp or ground the output voltage if it happens to be higher than it's reference; they would simple turn off at that point.
In the case of a low battery (<out) , turning on could cause a small inverse flow - also called charging (in the case of MosFets - less likely in BJT).
The particulars would depend on the actual circuit - but that is the basic transistor theory IIRC.
I dont get how there would be a current loop, wouldnt battery drain be a function of amps? Its not like the arduino is going to draw more amps because its connected to more sources right? If I wire 2 5v batteries in paralell and connect them to my arduino it doesent do anything other than supply it with 5v at double the capacity.
Electrons flow wherever there is a gradient. If the ESCs aren't perfectly matched (and they often aren't), the high voltage one ends up trying to reverse the current flow in the lower voltage ones. Sometimes that's no problem, sometimes it's a minor problem, and sometimes it can be a big problem, depending on how mismatched they are.
You can google around and see the various discussions on this. Here's one.
This discussion (linked) has a few anecdotal (my copter fell out of the sky - which I expect is a story many many people could tell in one form or another), and proper electronic analysis which always reaches the same conclusion:
It's fine - but doesn't necessarily mean 4x power - it might however hold up the power in the event of /some/ failures.
If you want 4x power, you need a sub-ohm resister on each BEC; if you only want redundancy, the direct connection does something; but let's be realistic; it's just not redundant. For that one needs separate batteries, and current limited power supplies for each peripheral.
Ideally, this means the GPS/Autopilot system on one separate system, with the radio/manual control on another - with 3 batteries - each powering 2 propellers (of 6) on opposing sides. Then you'd need 2 gyro-stabilizers - one for the radio/manual side, one for the autopilot/gps side.
The tricky bit is the third system, it needs to stay up and receive commands from either stabilizer - so it needs two comm channels, and some logic to detect which stabilizer is in control.
I am just trying to make redundant all I can without spending any extra time or money. And hooking up all my BEC's to the arduino seemed like a good idea. From what I have seen, hardware failure in the esc's and motors seems to be a big issue with most crashes and being able to keep my Y6 in the air if any of the esc's go out is important to me and my 7D :)
Dean Kaman took out a patent on full redundancy (including double-wound motors) for the Segway. If you read that patent, you'll have a primer on Redundancy in Batt/ESC/BLDM systems.
The problem with ESC failure, is that too often Mosfets fail Shorted - this puts the full power of the battery through a single winding until something gives. in your case; with a Y6, I would add a fuse for each motor, or pair of motors - you might fly on a burnt ESC - and you'd need to bridge at least 2 BECs.