Twins: When one motor "fails" I want the other to quit at the same time.

[RCHenry]

The guys on the "Power Systems..." forum assured me that if I brought my question over here, one of you sharp electronic types would have an answer.
So, first a little background.
I have build and flown several twins with brushed motors and NiCds. I wired them all in series because when one motor quit I knew the other would quit at the same time--no roll-overs due to one-mortor flame-out.
Now I am "upgrading" brushless motors and LiPo cells. To my great disappointment, I find that I must wire brushless motor in parallel. If one motor quits the other just keeps on running. Hell, I could have that problem flying IC on less money and more power, but I have grown to hate the smell of nitro in the morning.
Here is what I want:
If one motor quits I want the other to quit dead at the same time. I do NOT want to have to day ANYHTHING manually that requires time to recognize the the problem nor anything that depends on my reaction time to solve it--like manually cutting the throttle. I want the system to do it for me, AUTOMATICALLY and INSTANTLY.
There was one suggestion over there that might make sense, and I would appreciate your thoughts on it. The suggestion was that I wire an external LVC between the RX and "Y" harness just before it splits going to the controllers. This person thought that with this setup, if one motor OR CONTROLLER dropped out so would the other motor.
What say ye??
Your help will be much appreciated.

[RCHenry]

One more point. When I posed this question over there, some got confused thinking I was talking about motor cut off when one motor reached the LVC point as set by the controllers. NO. That is not the question. I am talking about the event of malfunction of either a controller or a motor. In which case, I want them both to quit.

[Mr.RC-CAM]

A circuit solution would involve adding a magnet to the props, mounting a hall effect sensor to each motor's hub area, and then creating a custom circuit to monitor the RPM's. It would kill the Rx's throttle pulse when the motor speeds were substantially out of sync. This is best done with a microcontroller; it would not be an afternoon project.

Some comments.

(1) In your old brushed motor setup, you wired both motors in series, so that if one failed, the other would too. But, with a shorted motor, the failure mode results in exaggerated RPM's on the remaining motor. This would be much worse than if nothing was done.

(2) Some brushless controllers allow you to put two closely matched motors in parallel.

(3) My gut feeling is that a power plant failure, while in the air, with a quality brushless motor/controller, is much lower than a radio/servo related failure. And the added electronics to achieve your goal may (statistically speaking) result in something that is not as perfect as it may seem.

I'm not against the project at all. I just wonder if it is a serious enough threat that deserves a lot of attention. But, as a gadgety sort of fellow, I do think it sounds like an interesting project. So, hopefully someone will chime in with a solution that solves the problem and introduces no ill side affects.

RC-CAM

[RCHenry]

Thanks Mr.RC-CAM.
But what are your thoughts on the suggestion of putting an external LVC between the RX and Y harness just before it splits to the two controllers as described at the end of my first post?
Details:
My system consists of two 2826/12 AXI motors.
Two 40A opto controllers.
Two 4s packs wired in parallel making a 4S2P.
All wired parallel.
No BEC.

[RCHenry]

Mr. RC-CAM:
I am a complete electronic dummie, so please forgive me if I fail to understand. But when I was flying with the old brushed motor, NiCd, wired in series system I had motor failures, but the other motor ALWAYS just quit. Never did I experienced an increased RPM as you suggest. Or am I missing something?

[RCHenry]

Mr. RC-CAM:
Let me see if I understand you:
Wired in series each motor sees only half the pack's voltage.
If one motor is shorted, are you saying then that all of the voltage will then go to the secnd motor, thus increasing its RPM?
If that is the case, then I was just lucky because I never had it happen.

[RCHenry]

Again, Mr. RC-CAM.
This might inter the equation:
In the series system I used only one controller.
While with brushless, I must use two.
At least I have been advised by MANY not to use just one.
Really appreciate your thoughts

[Mr.RC-CAM]

There was one suggestion over there that might make sense, and I would appreciate your thoughts on it. The suggestion was that I wire an external LVC between the RX and "Y" harness just before it splits going to the controllers. This person thought that with this setup, if one motor OR CONTROLLER dropped out so would the other motor.I do not know of any external LVC's that would work as you described. None that I know of have any awareness of the health of the motor or controller.


If one motor is shorted, are you saying then that all of the voltage will then go to the second motor, thus increasing its RPM?That is what would have happened in the shorted brushed motor scenario.

In the series system I used only one controller.While with brushless, I must use two.That is generally correct. But some brushless controllers will work with two motors. There have been some discussions about this in the past.

[RCHenry]

Thanks so much for your help Mr. RC-CAM.
Please forgive my densness. I have no experience with any LVC.
But isn't it supposed to recognize an increase or decrease in voltage?
It was also suggested that I set the cutoff point on this external LVC higher than that set by the controllers. I am just wondering through the fog here.

[hul]

It was also suggested that I set the cutoff point on this external LVC higher than that set by the controllers.
2 ESCs cutting their motors on (a slightly different) low voltage would be the most frequent reason for one motor loosing power before the other. The fix for that is having a separate LVC set at a slightly higher voltage with a Y-harness as you described. You'll need a common battery for the 2 motors too, for this to work.
This will not stop one motor if the other motor stops for any other reason.
But a properly designed system with some reserve capacity in the motors, speed controls and battery should not fail in flight.

Hans

[Nightz]

Here's another idea, another project that would require allot of work I think.
Put together a micro that samples running current. Teach the micro what your average current would be for each throttle position. If the current was ever off at a given position, cut the power. A micro sampleing rate should be quick enough.

Here is another way. Put a coil around one wire going to each motor. The current through the motor lead will induce a voltage through the leads of the coil. Will require some calibration/correlation but each coild could be interfaced to a micro that looks for something out of "sinc".

Neither are quick and easy methods.

OH WAIT!!
This one may be the easiest. Connect one wire from eash ESC output to a comparator. Some micros have a built in comparator. When the signals are different you can cut both motors.

-Rocko

[RCHenry]

Hay thanks Mr. RC-CAM.
As you say, an external LVC will not cut both motors if one quits, but your information regarding how it will fix the problem of one motor losing power before the other due to the difference in ESC low voltage cutoff proves most helpful. This is something I most certainly will do. Thanks.
Only one question: Would it not be better to just disable the LVCs on the controllers all together and depend solely on the external LVC?
Otherwise, you electronic chaps are getting way over my head.
You would think that some enterprising manufacturer would do this. Maybe a speed controller just for twins?

[RCHenry]

Yes, Mr. RC-Cam, I use a common battery: two 4S packs wierd parallel making one 4S2P.

[Ironsides]

RCHenry:

Without dismissing any of the great electronic discussion, there is a mechanical semi-solution. In the early days, when engines were unreliable, the engines were splayed outboard. Extra power on one side pulled the aircraft to the outside lessening the need for extra rudder correction.

Note the thrust lines on the attached versus the fuselage datum line.

David

[Terry S]

Just my thoughts, but I think a brushless setup quitting will be very rare. I run a pair of out runners on a twinstar and when ever the aircraft starts to glitch or go out of control the first thing I do is kill the throttle and try to get a steady glide. When I have a steady glide I turn back for landing then try adding some more power to see if its ok. I dont think you need a gadget, just remember you are the pilot ! Interesting to see what ideas come up though...

Terry

[vintage1]

I have not responded on this before, as I could not see the wood from the trees.

Getting perfect synch is only possible by using a spcialised ESC. ONE ESC. those who have driven two identical motors off a single ESC say that once they are up and running - a flick does it - they stay up, and running.

That solves all problems really. Except for motor failure, but that is likrly to disrupt the sensing so much that both will stop anyway.


As far as running with two controllers goes, an extenal LVC solves the LVC problem, as does using LIPOS and never running tham anywhere near LVC..

Engine failure is likley to be a hairy bit of deadsticking anyway. No matter what you do.

But running an electric setup at conservative values is very reliable. Most failures occur because people will push the envelope.

Don't. Pick motors packs and XCs that can relaible accept 50% more than you need, and run them that way.

[willfly]

I agree with comments that B/L systems are far more reliable then the I/C engines and don't see it as a big enough reason to mitigate the risk of engine failure.

Having said this, if you do need something, it will involve custom development. Some circuit which reads shaft rotation; on stop, stops the other motor. Hall efffect sensors will do non invasive rpm readings, motor kill can be achieved by using MOS/HEXFETS in power line.

An electro-mechanical solution would involve using some sort of centrifugal switch on the motor shaft which switches on a relay in opposite motor. There is more to it then described here (like a manual push override button during initial start), but you should get the general idea.

[Acetronics]

Hi, all

I think to a gadget :rolleyes: ... when turning, a prop moves air ... Yess

So, if we use a tube ( like pitot ) behind each prop ... we know prop pushes on air.
a differential pressure sensor, a bit damped, with an input behind each prop could tell us the thrust or the flight is not symmétric.

we have the info something wrong there ...

Alain

no rewards claimed ... thanks !!! :p

[sajeev]

why not just sense if there is amps flowing into the motor ? differential sensors to measure current to both the motors. too much current , above the obvious would mean a short, no current or too less would mean the motor is out and the other motor too should be out....

~ Sajeev

[Acetronics]

:D
Hi, sajeev

Think global ... the glow and gas engines are mostly concerned by such a device !!!

Current draw ??? why not !!!

Alain

[GlowFly]

How about a different, lateral thinking approach such as is sometimes used on ic twins? Fit a rudder gyro. Then a single motor failure may render the model more controllable.
--
Steve

[sajeev]

ok then in the case of a gas plane why not sense fuel flow ?
or probably a sensor, kinda like a smoke sensor in the exhaust pipe ???

[sajeev]

hey how about two of those cheap RPM monitors used to tune the RPM. one behind each engine. no RPM on one engine, then cut the other engine too.
you could take the output from one of the display pins, feed it to a PIC that monitors for a 'high' on both engines. the moment it detects a 'low' the pic shuts down the throttle to both engines completely

[sajeev]

pressure sensors might not work since the glide of the plane will contribute to pressure in the tubes....

[Acetronics]

pressure sensors might not work since the glide of the plane will contribute to pressure in the tubes....

:eek: Did you notice, it's the DIFFERENTIAL Pressure that is to be used ???

The gliding gives the same pressure on the 2 sides ...DIFF Pressure = 0 !!!

pressure is a function of speed AND Thrust ... but I hope your two wings fly at the same speed !!!

So, Diff pressure is only a function of Thrust differences

LOL !!!

Alain

May be I'll accept a nomination to the Nobel Prize, after all ... :D

[Bthree]

Thanks Mr.RC-CAM.
But what are your thoughts on the suggestion of putting an external LVC between the RX and Y harness just before it splits to the two controllers as described at the end of my first post?
Details:
My system consists of two 2826/12 AXI motors.
Two 40A opto controllers.
Two 4s packs wired in parallel making a 4S2P.
All wired parallel.
No BEC.

If the above is the intended configuration then you would set the LVC on both ESC's to say three or even two cell LiPo cutout (or low cell cutout) so they would never activate, this will be your largest threat. Use an external LVC to control both speed controllers together (if you can find such a thing).

Don't try to run this setup on one speed cont. the power levels are too great

No idea how, but to sense a large out of balance current flow to the two spreed controllers using a hall effect device (clamp on ammeter) from say thr two red wires going to the ESC and a trip circuit, but really two brushless motors is going to be very reliable, many more times than a brushed motor setup and for that matter glow motors being the least reliable.

[skewbe]

RCHenry, your serial brushed motor setup was not foolproof, FYI. One motor could "lock up", or a prop could break/come off.

From what I understand, your best bang for the buck will be a single controller operating two brushless motors OR two controllers with their power leads in serial. It would ultimately depend on which is more reliable, the controllers or the motors and your wiring.

Any interlock that you come up with will have to be smart enough to allow the motors to "bootstrap" otherwise each motor will refuse to start turning because the other one isn't turning. Might I suggest a couple CDS cells (like the cheap tachs) pointed at the propellors that look for a threshold difference between the propellors rpms and cut the power if there is too much "slippage", assuming you are fixed on this gadget?