Pixhawk Hardware Reliability

Hi All,

I had an APM 2.5 on my plane for a couple of years and after the initial setup it performed reliably for those 2 years and was quite solid. Recently I decided to upgrade to the pixhawk and ended up buying two of them for two different planes. So far my experience with them has sucked. One of them died with the (No PX4IO board found) message which indicates hardware failure. 3DR agreed and replaced it (but it took a month). Now, just today the other one lost it's USB port and it's hardware failure as well. This is devastating to me and has me wondering if I should just give up on Pixhawk and just us the Eagle Tree Vector instead which so far at least has been easier to setup and with no failures ( I have two of them as well). 

Though I'm intrigued by some of Pixhawks advanced features and capabilities I really want something rock solid and usable. But I'm upset that I've wasted about $800 and countless hours on this experiment. Did I just have really bad luck with this hardware or is it a poor design and fragile?

Thanks for any insight. 


You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –


  • All,

    I think it's worth providing a bit more personal background to this discussion for context. As I mentioned in my original post I had an APM 2.5 installed on my main plane for a couple of years. Once I got through the learning curve on that and got it installed I played with pulling logs off and studying them in the beginning. But after that I just used the APM for plane stabilization and potentially RTL while I FPV'd and took my aerial video of many places. I didn't even use the OSD because it seemed like too much trouble to install. The APM worked fine for that, hands off, without ever hooking it to the computer again. As I got busy with other things I only flew once in a while. 

    Then I retired and decided to get more involved in the hobby again. I purchased one of the brushless gimbals but in order to hook it up I had to change my TX radio program and because of that I felt I needed to connect to the APM to make sure I didn't alter my flight modes. Well that didn't go well with the new versions of the software so I had to scrap the APM and replace it with something. Naturally, I thought the Pixhawk would be a nice upgrade so I bought one and then found I couldn't connect my Spektrum radio to it because it could only take PPM or sbus to control it (There were supposed downsides to the PPM adaptor). So I decided if I had to find a new RX I might as well get something out of it so I bought a DragonLink system (which is having their own growing pains (V3) that slowed me down a lot). Talk about a domino effect this was getting crazy both in expense and complexity !!!

    During my research I discovered the Eagle Tree Vector and bought one to try. I was very impressed as they seem to cater towards people like me that tend to get overwhelmed by too much complexity when the real goal for me is to just fly worry free and take nice smooth videos. So I want people to understand I'm trying to make the best of the money I've invested in Pixhawk and to push through if I can and learn and get it working for me. But if I get frustrated with some of the complexity and learning curve you understand it may be because of my own weaknesses because I can get very impatient and am a low energy guy like Jeb Bush. :-) It doesn't help that the weather sucks for flying so I can only build but can't test fly anything very well right now.

    • The Pixhawk world is not plug and play. The bottom line is the output rail of the Pixhawk was not designed to be a power rail. The Pixhawk designers enforce that by not powering it. If you have a few small servos that don't pull a lot of power you can probably get away with powering an empty spot on the output rail with a 5v ubec. Otherwise keep the power away from the pixhawk and use an external power bus for the servos. Computers are subject to noise and the last thing I want to do to the computer that keeps my aircraft flying is to risk interfering with it.

      Doing it the way you show in your diagram, you are using the Pixhawk as your power bus. If you are going to have a lot of servos and a gimbal powered that way it is a risk. Remember all the power is going through the Pixhawk power bus the way you have it. I would not do it that way.

      But you pays your money and makes your choices

      • The Pixhawk world is not plug and play.

        I knew that going in of course but I've been surprised at how difficult it can be at times, and some of it seems unnecessary.

        The bottom line is the output rail of the Pixhawk was not designed to be a power rail.

        I think it more accurate to say "Some very smart people with a lot of experience using the Pixhawk decided it would be better not to use the built in power rail"

        The Pixhawk designers enforce that by not powering it.

        I think they didn't power it so a user can provide the power of their choice and voltage level. Unfortunately, they also chose to tap that same rail internally for backup power to the Pixhawk but did that in a flawed way so that users have to stay away from it as a solution altogether. Sorry, but that's my interpretation of the situation.

        I also just discovered there is a Pixhawk 2 in the works. So it might be that some of these design flaws are addressed. But I find it interesting that 3DR refuses to manufacture it. And I saw a video not too long ago where, if I recall correctly, Chris Anderson stated that 3DR will be moving away from the retail hobbyist market and towards the commercial market. Just interesting new information that is relevant to this discussion if you think about it.

        Thanks for sticking with me on this discussion Mike, I find it fascinating.

        Here is the Pixhawk 2 thread I found...


        • I will concede that there are a lot of players in the mix. 3DR is just one of them. They did not design the Pixhawk. So each party may have a different attitiude towards use.

          The Pixhawk and Ardupilot have never been easy. Never will. They are the feature rich bleeding edge. The tip of the sword. There is nothing in the current flight controller market that can beat them. But that comes at a cost. Always does. Complexity. The Pixhawk 2 will not be any different. They are complex flying computers doing really amazing things considering the size and the cost of the components. Something you could not have bought for any price just a few years ago. These are exciting times for the Do-it-Yourselfer. On the other hand...

          If you don't need autonomous flight or scripting or follow me etc etc then they are not the right choice.

          But there are rewards for persistence if you do.

          • Thanks Mike,

            I agree with everything you just stated and think you have just summed it up nicely!

            I don't need any of the extras Pixhawk is providing right now vs. the Vector.

            That said, I do like the info Mission Planner provides on the ground and I can't help but think it would be nice to have that screen live when it's in the air. But I don't want to give up on the live video feed and have, so far at least, figured it had to be one or the other. I recently maiden'd a Penguin that can get a 30 minute flight time where as I used to be limited to 7 minutes. This might be just the plane to allow me to fly from inside a vehicle using autonomous flight and a data radio. This is something I've aspired too but the learning and implementation curve has kept me away. 

            So, bottom line, since I have the Pixhawk I'll keep working with it until I either give up, have success with some more advanced features, or just use it as a simple stabilizer as long as it stays healthy. 

  • If you don't want the pixhawk to control the servos you don't really need it at all...

    OP is controlling a plane not a gimbal with the servos. Pixhawk kinda needs to be involved.

    • I think the point is that the the Pixhawk circuitry and servo rail power are only isolated when the voltage is above 5.6V.  Pixhawk is intended to be flexible, meaning, if you really want to you could apply 10V to the servo rail and it'd power your 10V servos great (assuming the amperage is low enough for the busses to handle), but the FC circuitry will cutoff power to the Pixhawk, hence the need for powering the Pixhawk via the PM if you're dealing with servos on the servo rail.  

      In my plane I just use the PM + connect the ESC (which has a ubec) power, ground, and signal to the servo rail, which provides separate power for the servos.  I don't have the zener in there but it's not needed because the Pixhawk will automatically run off only the PM input in the event of a voltage spike that would shutoff power from the servo rail to the FC itself - there's no interruption in power.

      Bottom line, if you're using the PM and also have power applied to the servo rail for your servos, there's really no need to worry about the Zener - unless you're a zealot that wants triple or quadruple redundancy...in which case I wouldn't be writing this because it wouldn't have been a point of confusion in the first place.

      • Thanks for the clarification Matt. Your post makes sense and agrees with my new understanding of the way this system works. I'm using a separate BEC instead of the one included with the ESC but it's essentially the same wiring diagram as you just laid out.

  • Paul, I am not sure you stll understand #3. Servos pull a lot of power. So 3DR does not want you powering your servos off the output rail because it might brown out the Pixhawk when a bunch of servos were actuating at the same time.

    A servo cable has three wires. A ground, black or brown, a positive voltage, red, and a signal wire, white  or orange. You can send signals to control your servos from the Pixhawk output rail, just don't power the servos from the output rail. So what that means is you need the signal wires but you don't want to use the red power wires. So you power your servos by suppling power from a separate ubec say. With the ground and red power wires going to the servos from the bec. Then the signal wires come from the Pixhawk output rail controlling your servos. Make sure the ground for the bec is also common to the ground on the Pixhawk power module. So you might have a servo cable you make up, with red and black wires coming from a ubec and the white or orange wires coming from the signal lines you are using on the Pixhawk, this cable plugs into your servos.

    The RCin and the Sbus.out out should not be used to power anything but a receiver.

    Does that make sense?

    • Thanks for your input Mike however I can't believe your answer "3DR does not want you powering your servos off the output rail" I quoted the part of your answer that I can't believe for clarity. While I've been somewhat critical of the Pixhawk design and documentation in this area I'd be really critical if 3DR really wants you to cut all your servo leads and solder them together somewhere else to "avoid powering the servos off the servo output rail" That would make connecting and disconnecting servos quite inconvenient to say the least. 

      That said, your post, if implemented, does give the user a way to isolate the servos power completely from the Pixhawk, as it would be easier to understand if they had done it this way. To put it another way, if this subject can not be clarified to the satisfaction of a user and they want to be sure the 5v power rail is isolated they could install another separate "3 pin rail" and wire it back to the Pixhawk so that the power rail is completely isolated while the signal pins are connected to their respective pins on the Pixhawk. This would allow the servos to be plugged in and unplugged normally but insure the power was separate. I'm hoping that the outcome of this discussion thread will prove going to this extreme unnecessary however.

      It appears the RCin and Sbus out are not on the same "power rail" as the rest of the connector bus and the Pixhawk provides the power to these but at a limited current level so, yes, this part makes sense. I've worked around this by using the same BEC power to power both the Pixhawk servo rail and the RX power rail. Now as long as the zener diode proves unnecessary then I should be good.

This reply was deleted.


Neville Rodrigues liked Neville Rodrigues's profile
Jun 30
Santiago Perez liked Santiago Perez's profile
Jun 21