Thread Tools
Jan 26, 2011, 02:40 PM
Pants first, then shoes
Thread OP
Question

Quad with 8 motors or Hex?


I have seen some footage of the olsen Brothers quad which has 8 motors. They have the one on top one underneath configuration and in their video they mention the cover this provides if a motor dies.
Is their any negative impact on lifting ability with this configuration? I would imagine that either the lower or upper prop is trying to push "dirty" air?.

In terms of redundancy and lifting power is their any advantage to a quad with 8 motors (1 top 1 below configuration) over a hexacopter?
Thanks
Sign up now
to remove ads between posts
Jan 26, 2011, 07:47 PM
Pants first, then shoes
Thread OP
Sorry also wanted to add is there any power useage advantage of one over the other configurations?
Jan 27, 2011, 05:05 PM
Registered User
Yes, the efficiency is reduced (because power usage per pound of thrust goes up). A coaxial octocopter (what you call a quad with 8 motors) is about 30% less efficient (in terms of hover time) than a regular radial octocopter of the same weight.

Over a hexacopter it has the advantage of true one-motor-out redundancy (any single random motor can fail and it will still finish the mission). In special circumstances it can afford to lose as much as 4 motors (e.g. all the lower ones) and still finish the job (with obvious reductions in flight time and thrust). Compare this to a hexacopter, which can only fully handle a motor outage in 1/3 of random cases.
Jan 27, 2011, 05:21 PM
Registered User
Quote:
Originally Posted by tori3
Over a hexacopter it has the advantage of true one-motor-out redundancy (any single random motor can fail and it will still finish the mission). In special circumstances it can afford to lose as much as 4 motors (e.g. all the lower ones) and still finish the job (with obvious reductions in flight time and thrust). Compare this to a hexacopter, which can only fully handle a motor outage in 1/3 of random cases.
I don't follow this last statement. Where has the probability of a 1/3 come from? Surely a hexacopter could theoretically lose any two motors and still be able to fly? It would essentially be a quadcopter, albeit a very unstable one depending upon which motors go. It could possibly even lose three and still fly.

The worst cases for a hex losing two motors would be two adjacent ones, which would mean the centre of gravity is very off-centre. Or two CW or CCW motors go, meaning the remaining one which rotates in the same direction has to work extremely hard to counteract the torque of the other three.

?
Jan 27, 2011, 05:37 PM
Registered User
Let me clarify that. Imagine you have a hexacopter in the air, and you're performing experiments. Every time you cut power to one randomly selected motor. Just one. You do this maybe a hundred times. You will find that in about 1/3* of all the experiments the hexa switched to quad mode and kept flying without major problems. This is because 2 motors (out of 6, hence 1/3 probability) are on a zero-torque bar, so if one of them fails, the controller responds by killing the opposite motor.

Losing any other motor is still somewhat survivable - the controller shuts down another motor to cancel out the torque, but will lose yaw control in the process. The aircraft will tumble itself to the ground in a semi-controlled fashion, maybe break a prop on the landing.

* The probability of randomly picking one out of two specific motors out of a total of 6 is 1/3.

Edit: I'll draw a bunch of pics, stand by 30 minutes.
Jan 27, 2011, 06:32 PM
Registered User
I`m glad you did let me know that a 8x Quad is not better than a full 8x Octo.
I will start to build an octo soon how fast does an octo goes compared to a small quad ?

I will use the FY90Q on my octo
Jan 27, 2011, 06:45 PM
Registered User
So here's your run of the mill ISO Standard Hexacopter, floating about, minding its own business.



All of a sudden a motor #1 dies without as much as a warning. Fortunately it's a member of the zero-torque group (motors 1 and 4), so the aircraft's smart controller responds by cutting power to the other motor in that group, restoring balance:



The result is a somewhat heavier quadcopter, but still in perfect flight condition.

If a motor from nonzero-torque groups would die, the action taken by the controller would have to be more drastic. Motor #2 dies. The controller has to kill motor #0 to bring yaw torque back to neutral, and then throttle motor #1 up and motors #3-5 down in order to compensate for the missing pair:

Jan 27, 2011, 06:47 PM
Registered User
Quote:
Originally Posted by liteglow
how fast does an octo goes compared to a small quad ?
You mean the speed of flight? Well, that depends. They are in the same ballpark, no large differences between the two, so normal flight around 50 km/h, top speed around 90. It all depends on their thrust-to-weight ratios.
Jan 27, 2011, 07:02 PM
Registered User
Quote:
Originally Posted by tori3
So here's your run of the mill ISO Standard Hexacopter, floating about, minding its own business.



All of a sudden a motor #1 dies without as much as a warning. Fortunately it's a member of the zero-torque group (motors 1 and 4), so the aircraft's smart controller responds by cutting power to the other motor in that group, restoring balance:



The result is a somewhat heavier quadcopter, but still in perfect flight condition.

If a motor from nonzero-torque groups would die, the action taken by the controller would have to be more drastic. Motor #2 dies. The controller has to kill motor #0 to bring yaw torque back to neutral, and then throttle motor #1 up and motors #3-5 down in order to compensate for the missing pair:

That's brilliant, I see what you mean now. I didn't realise they would be arranged like that. Why not have all of the same direction props on one side though? So swap 2 and 3 in your diagram. That would make all of them zero torque pairs and lessen the impact of loss.
Jan 28, 2011, 01:42 AM
Registered User
You need two nonzero torque pairs in order to steer the aircraft. Two nonzero pairs can be commanded to steer without producing thrust (throttle the clockwise pair up producing more counter-clockwise turning moment and throttle the counter-clockwise pair down the same amount to reduce its clockwise moment, effectively ending up with positive clockwise moment and same thrust as before the maneuver).
Jan 28, 2011, 02:27 AM
Registered User
Tori3: Great help and great info about multicopter setup

I was afraid that the size of an octo would make it hard to (let`s say) follow an bike cycle down a track to get it on film, or someone skiing down a track.
the normal speed would be 20-50kmh i guess.
Jan 28, 2011, 03:39 AM
Registered User
Quote:
Originally Posted by tori3
You need two nonzero torque pairs in order to steer the aircraft. Two nonzero pairs can be commanded to steer without producing thrust (throttle the clockwise pair up producing more counter-clockwise turning moment and throttle the counter-clockwise pair down the same amount to reduce its clockwise moment, effectively ending up with positive clockwise moment and same thrust as before the maneuver).
Ah of course! I hadn't hadn't noticed that. Thank you for taking the time to explain it.
EDIT: See post 29. If you're running them in pairs then yes though.

Liteglow, sorry for mini-hijacking your thread. That's fast. A fun build
Last edited by websitedragon; Feb 11, 2011 at 09:59 AM.
Jan 28, 2011, 04:02 AM
Shoop da whoop!
Technotron's Avatar
Hey Tori,

Thanks a bunch for explaining; you really made it easy to follow :-)

So, do most controllers these day cut off motors automatically to compensate in case of a motor failure? My KK Hexacopter (with the latest controller board) should arrive this week, so I am curious to find out if it will behave similarly. Here's the link:

http://www.kkmulticopter.kr/index.ht...h_290&id_no=53

Edit: You know what I just realized? With this kit I could try both the Hexacopter and Y6 setup. Ace! How does a Y6 compare to a Hex? I've only only flown quadrocopters so far, hence my question.
Jan 28, 2011, 05:27 AM
1400 watt RW Zephyr
matwelli's Avatar
Interesting conversations on over/under motor combos vs top mount only.

What intrigues me is the loss of efficiency, so as soon as my motors turn up, i will do some testing with an eagletree logger and post the results.
Jan 28, 2011, 05:46 AM
Registered User
Quote:
Originally Posted by Chris Burton
So, do most controllers these day cut off motors automatically to compensate in case of a motor failure?
I'm working on an Atmega1280-based controller that will be able to do this, but I'm unaware of any released board that would. With most designs it's up to the pilot to compensate for failures with trims or stick movement.

Also, adaptive PID controllers should be able to tolerate some sort of failure automatically, without explicit failure handling code.


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion advantages with a hex vs a quad? LMAO Multirotor Drone Talk 4 Dec 30, 2010 12:05 PM
Discussion Interested in getting a Mikrokopter hex or a droidworx hex. How hard is set up? LMAO Multirotor Drone Talk 8 Dec 28, 2010 08:47 AM
Discussion Large Quad/Hex with GPS waypoints octoberfish Multirotor Drone Talk 3 Apr 23, 2010 11:04 AM
Question Quad versus Hex or Octo Hunter850 Multirotor Drone Talk 9 Mar 30, 2010 09:06 AM
Wanted 8 -32 plastic or nylon bolts with hex head or allen heads BrianSmith Aircraft - General - Miscellaneous (FS/W) 3 Mar 28, 2009 06:28 AM