Power Manager for FPV plane

[Kuba]

Hi,

I recently created a simple power manager for my FPV plane:







The reason I created it is simple: When I fly FPV I often cruise above buildings, cars etc. Potential failure of any servo or BEC could result in a crash for example into a nice and shiny sport car. What my Power Manager does is:

1. Protects against BEC failure
Power Manager has 2 power inputs for 2 BEC's. If one of them fails the other one continues to deliver power to the servos even if the broken one is shorted to the ground (I didn't add any over-voltage protection yet but I don't think it is necessary). In normal operation current is being drawn from both BEC's sharing the load proportionally to the voltage difference (if you use 2 identical BEC's they will share 50/50)

2. Protects against Servo failure
Basically I assume, that responsible FPV pilot not only performs regular tests of servos but also has 2 servos on important suraces (ailerons and elevator). Let's imagine a situation, where your plane is far away and a servo stops working because it is mechanically broken. Since we have 2 servos for each surface it should theoretically be possible to land (or at least choose a crash site far from buildings and cars). Unfortunately it isn't always that simple. When servo breaks down mechanically it's electronics try to move the servo anyway causing heat dissipation on the electronics which eventually leads to burning it. Now there are 2 possible ways things happen: either the electronics broke down and servo is not consuming any current (the good scenario) or it broke down causing short-circuit on the power line (the bad scenario). In the second one classic FPV setup will loose 5V for the servos (because one of them is creating a short circuit), probably burn BEC and we have a major crash. The first scenario is more optimistic but it may happen, that even when servo isn't consuming power (or creating short-circuit) it may be blocking the signal line which will prevent the other servo on the same channel from working and we still have a crash.

My power manager has 2 types of protection. The first one is a fuse that will burn out if the servo exceeds it's expected current draw. During tests I found, that typical 9g servo can pull 0,5A in it's highest load so I used 2A fuse. There is no way that fuse gets burned during normal operation, only if the servo's electronics break down and create a short circuit. Simple.

The second type is buffering the signals from reciever/autopilot to the servos and double outputs. If you look at my pictures you will notice, that almost every channel has 2 outputs. Even though they are the same channel their signal lines are not directly connected which means, that regardless of what happens to one of them the other one will always work and there is no way a broken servo could damage the reciever or influence other servos.

Let's consider Elevator:
In my Skywalker I have 2 servos for that surface (it is divided into 2 so if 1 serwo stops the other can still move it's surface). One of the servos is connected to ELEV_OUT1 and the other one to ELEV_OUT2. Even If i short all 3 wires that go to one of these servos the other servo will still be fully operational and I will most likely be able to land my plane safely.

As far as ailerons go most FPV pilots allready have seperate servos for left and right ailerons so the same situation applies.

3. Ability to power the reciever by different BEC
In some cases you may want to use different power source for RX than for servos. It is especially important when you fly on 35/72MHz radio that is very sensitive to power spikes. In my power manager you can either connect the external source or just use the same one as for servos.


Why am I publishing my project here? At first I wanted to sell a few of these power managers to other people on our Polish FPV Forum (I actually gave away a few to my friends for testing) but the Polish community found my project rather unnecessary and almost everyone claimed, that servo failure never happened to them and never will. I believe that my project will be more appreciated on this forum and I am curious what you think. If you think similar to Polish FPV pilots please feel free to criticize. Original thread (in Polish): http://rc-fpv.pl/viewtopic.php?t=4898

If anyone is interested in getting one of these I guess I could make that happen.

[Squawk]

I like the idea of redundancy!

But I have 0 experience with FPV flight yet. (Still designing an antenna tracker.)
So I am curious too, how some experienced folks here look at this idea..

[baka]

Some good ideas. Part of me always thinks the more kit you add, the more kit you have to go wrong, if your system controlling redundant systems goes wrong it could cause you to crash. I always check my equipment out before I fly but I can never say 100% that any of it wont fail randomly, same as a real plane. Will it stop me from going on holiday or flying fpv this year, no. You pay your money, you take your chances

[Kuba]

Some good ideas. Part of me always thinks the more kit you add, the more kit you have to go wrong, if your system controlling redundant systems goes wrong it could cause you to crash. I always check my equipment out before I fly but I can never say 100% that any of it wont fail randomly, same as a real plane. Will it stop me from going on holiday or flying fpv this year, no. You pay your money, you take your chances

The power manager is redundant itself and during development I kept in mind that it can not be a weak point. There is no CPU onboard that can go wrong, just pure basic elements. It is designed in a way, that even a disaster on the board itself (I can't think of a reason for such but still) will disable only "half" of the servos. It means, that you would loose 1 aileron, 1 elevator servo and for example 1 flap. This is still better than nothing.

[baka]

That really cool, if the pcb itself is just about full proof. You made it from scratch? If so great job, really nice. Any plans on selling it?

[Alex Villa]

I think this power manager is a good idea. Most people start building their airplanes with the very basics to fly. But as you get experienced building you start getting higher standards, this is where this kind of device will be interesting. After all real airplanes do have double or even triple redundancy, this is because designers found that if done properly redundancy adds security.
From my experience I can tell you that you can have very good ideas coming from the forum posts. But you should never let the naysayers to derail your projects. Many people are hostile to developers for various reasons.

[rcmike]

Great ideas. In giant scale RC planes they often use dual RXs and batteries in case something fails.

[CaptnKliegle]

Very impressive CB and quality!

All my servo failures so far have been stripping of the gears in crashes or tall grass catching on control horn. These failures are caught on site when I do a control check at the scene of the accident. Although I do remember an aileron servo on my Easy Glider Pro stripping inflight, AND with that particular aileron not at neutral. Flew for several minutes like that, until it was close enough to hear the gears grinding over head...plus it's flight characteristics were radically different.

My question is...

If a servo has an electrical short to ground, or an open circuit in flight (not due to crash impact), effectively the servo now is 'frozen' in it's present position at time of failure. With this type of failure, will a similar servo be strong enough to overpower it if two servos are operating the same flight control?

I supose one could separate the elevators from each other and have them work independantly (like on the MD-82,82,90, etc airliner) and split the rudder in two for each servo. Now you might have a CG problem, because you just doubled the amount of servos in the tail of the aircraft, hence doubling the weight at a significant distance from the CG.

Interesting product though!

Kriss KA1GJU

[Kuba]

If a servo has an electrical short to ground, or an open circuit in flight (not due to crash impact), effectively the servo now is 'frozen' in it's present position at time of failure. With this type of failure, will a similar servo be strong enough to overpower it if two servos are operating the same flight control?

I supose one could separate the elevators from each other and have them work independantly (like on the MD-82,82,90, etc airliner) and split the rudder in two for each servo. Now you might have a CG problem, because you just doubled the amount of servos in the tail of the aircraft, hence doubling the weight at a significant distance from the CG.

That is why I split my elevator in 2. With foam plane second servo should be able to deflect the surface by bending the material but I wouldn't count on that too much.

As far as additional weight go - 2 servos on the elevator is safety but 2 servos on rudder is a bit too much If you have ailerons rudder is not critical and thus it can be operated by 1 servo only (that's my opinion). I included double rudder output for bigger planes that use 2 servos by default.

In my skywalker I have double elevator servos in the tail together with RC RX, 1 rudder servo, big gps antenna and some heavy wires (8 lines in a shield for RC RX, 3x 3 2 wires in shield for 3 servos, GPS antenna wire). I also added carbon reinforcements (I know, carbon is bad for RF but I couldn't find anything better at the time). Such component placement makes me use 2x 3s5000mAh lipo packs in the nose to keep it balanced and I'm good with that since that's what I wanted to use from the beginning.

From the top: black GPS antenna wire, 2 elevator servos, RC RX "hatch", RC wire, rudder servo. No reinforcements on this picture yet.


EDIT:
I forgot, I also have a wire for rear light (1 line in shield) coming from the fuselage to the back of the plane

[Squawk]

I'm glad that most experienced people here think this is an asset.
I would like to incorporate something like this in my RC model..

(And yeah, it does look very neat and professional!)