Yesterday I had 2 more seperate and and independent MK failures which has really shaken my confidence in the MK power boards and for that matter the I2C data buss system. First flight I was taking aerial stills for a client and after about 15 minutes in the air I heard the MK lady say Error "major" so I quickly got the MK down to about 20 feet and I saw that one motor wasn't spinning. I got the copter down and it actually flew and responded very well on 7 motors at 25 pounds. If I had not heard the error or seen the prop not spinning I wouldn't have been able to tell I was running on 7 motors. So the configuration was doing what it should have been. Got the copter home and found the issue to be the power board. BL#6 has just stopped working, no signs of damage or physical signs of anything wrong. I then broke out my backup copter for more filming. This copter is identical to my first one except I'm using KDE 4012's instead of the TIger 4120's on my main rig. It was fine for the first flight. I then start my 2nd flight and the copter went up about 3 or 4 feet, and just completely flipped out uncontrollably. The copters motors were unresponsive and I couldn't even shut the power off to the motors. It just stopped once it couldn't move any more. I heard the MK lady say "error data buss". On inspection I found my molex cable hot glued and intact on both ends. The first thing I did was install a brand new spare cable I had, no difference. Copter now won't start up and has the I2C error on the telemetry. I then swap out the flight control board, no difference same problemm so I now have 2 power boards bad. The worst part is that there is no GPX file for this flight, it didn't record it. It recorded the first flight but it did not record this flight for what ever reason. When went to remove the SD card from the Navigation board it was still in the board. It did not get dislodged from the accident. Here's the video from the incident yesterday Password : mk So I refuse to run these boards anymore. I either have 2 options, the Hercules board or the MK Wookong. Can anyone give me info on either of these? For the DJI stuff who sells and supports the Wookong? I need to purchase it from someone that has a lot of experience with it. I might try the Hercules on one copter and the Wookong on the 2nd copter. What's the best forum for DJI?
Dave sorry to hear about your incident. The WKM is very reliable especially the A2, as for the Herkules is just a esc power board you still need a controller. Can you trust the MK controller. I built two units for a local aerial client and used the WKM with no issues at all. But you have to take extra care to connect all the wires in sequence. I hope this helps. Bill
There was a wrong channel mapping issue with Graupner once I tried, but it was long time ago. All guys seem using Futaba with A2.
Dave, if you're considering starting over from scratch (very sad, but might be unavoidable), I'd consider the Gemini flight control system from Zero UAV. They are redundant out of the box. And I'd mate that with something like the 75A ESCs from KDE, along with KDE motors. I think we are all going to have to step up one level (or two) to get true professional, rock-solid reliability. There's not a lot of them out there, but at least one guy whose opinion I respect ("SleepyC" on MRF) has good things to say about them. http://www.foxtechfpv.com/multirotor-autopilot-geminidual-redundancyfreeshipping-p-1515.html
Hi Yes it dose I have a MX 20 That I use with a Naza v2 and a GR16 and a WKM sometimes use a gr 24 but which I usally fly with a Graupner mx 22 and a dragron link tx But they all work and facilitate PPM.
Dave: I'm using my Graupner MX-20 with two Naza birds without any problems. I have my DJI s900/A2 bound with a Futaba right now, but am switching over tomorrow, for the telemetry. I miss seeing voltage, and just got this: http://www.openhobby.com/front/prod...de=001000000000000904&code=007000000000&sort= If you want to go Graupner let me know and I'll walk you through it. I really should make a video, as most of the ones out there are in German..lol You do have to remap the sticks, but it is easy. FS switch setup is very similar to the Futaba. I have two MK birds, a Quad and a CS6, two Naza Quads, and DJI S900 + A2. The CS is the most stable, but there are so many variables in MKTools, it drives me crazy sometimes, even in "easy" mode. The wiki is a morass of German and out dated material. The DJI's are much easier to setup and fly, but not quite as stable. (IMHO)
Sorry to hear Dave, I have always trusted my mk over my dji stuff but you just never know when bad things are going to happen to both.
Thanks for the input everyone. Steve thanks for the suggestion. As of right now for me at least, everything is open for discussion. What I can say is that unless MK does something immediately to address the board issues I will be switching systems. Holger is in Egypt till Nov 6th so I have no idea what he's going to do or say. I will be talking to as many people as I can and researching as thoroughly as I can. I talked to Lawrence from Aerial Technology last night on the recommendation of Patrick from KDE. Lawrence echoed what a lot of people say that DJI has no tech support and their quality control is hit and miss. To me quality control is a huge issue. He was telling me about 3DR's pix hawk and was strongly recommending it. It looks good on the surface. They have a phantom like copter called the Iris plus. It has all the same hardware and software you would use on a copter. I was thinking about trying it to see how it works. Right now everythings on the table.
All systems have their issues. I had an I2C that dropped a octo out of the sky on MK, Saw a DJI based system fly about 30 ft off a dock and drop into the water with a Dragon and numerous 3Dr systems that the Uni I work with fail for no reason (software related to bad code in the FC and lack of mathematical self checks). If you go pixhawk just know that fly-away is an issue with that system, as well as terrible sync times on larger slower motors (again, bad code). I have been experimenting with Minix (http://www.minix3.org/) on an Intel Edison with a custom FC and it works great. I have purposely injected bad code into the system and it refuses to fail due to how Minix is setup with driver redundancy. Very complex system but it works great and can fit into MK boards, ESC based systems or whatever floats your boat.
Dave: I've been in contact with Ingo Busker at MK. I suspect he's rather busy given Holger's on vacation, so he asked me to post this on his behalf: @Dave: Andy asked me to comment on the situation regarding the issue that you have been having with the MK boards. At the moment we do not have enough information to comment on what has happened. Please understand that we at MK take these kinds of reports very seriously. We have hundreds of pilots flying with our boards and we learn very quickly if there are issues -- and we try to respond very quickly as well.So please be assured that we take the issues that you have reported very seriously. We will investigate them and let you know what we find.Thanks Ingo
Everything fails. While I have only used DJI as far as multi rotor flight controllers go, I have used just about every system they have came out with. I have had a few instances where my wkm had issues but only one of them resulted in a crash. 3 of the issues it had where saved by flipping into manual mode and flying it a little bit in manual and then flip back into atti mode. The one time I have ever crashed a multi rotor my wkm pmu failed mid flight. I plugged everything in, took off and was about 25 feet up when it just shut off. Once I got home i plugged it in and it all booted up just fine only to power off about 1.5 minutes later and i felt the pmu and it would get hotter and hotter until finally shutting off. My DJI flight controllers have been working great since then (knock on wood) but I think the biggest difference between dji and mk is the customer support. With DJI you have support from the users but that really does not help you if something fails. With MK you can actually talk to someone who cares about you and the system. Pretty much when the Phantom came out dji switched their tone to just caring more about sales than repeat customers and keeping current customers happy.
HI Tristan I don't believe that "everything fails" or at least I believe in reliable flyers when you setup your equipment properly, and have good routine maintenance. The incident above in the video is completely unacceptable and I can't afford for that to happen again. I have worked too hard to build up a good reputation to have it destroyed in one quick instance. I was completely powerless to do anything to shut the copter off. It would not shut off remotely. It only stopped when it decided to stop. On any other day that parking lot would have been full of people, kids, and cars. I normally have a nice safe take off distance away from things but this incident would have stretched past any normal safe zone. I am a huge supporter of MK electronics and I love the support and attention that Holger provideds but I just don't think the Dual Quadro XL boards are reliable enough at this point to fly on a regular consistent basis. I double checked my flight logs, this year I had 315 flights and in that time I have had 5 power board failures. Thats a failure every 63 flights. It's just good enough for my needs. Could the Hercules board be better? The early indications are that they could be but I don't have time to be a beta flyer and I won't beta test my stuff on client jobs. Today I purchased a Wookong and will try that out over this winter. I will put a couple hundred test flights on it between now and March and see what happens. I plan to start out slow and take baby steps until I am confident and experienced with the system. What I think we should all be thinking about is robust datalogging. What do we do if the FAA comes out and demands better datalogging than these hobbyist systems can provide? There are robust auto pilot systems out there but you can easily spend over 10K. Maybe that's what it might come down to if you want to be serious about this application. After all I invested a lot of money in the cameras and camera gimbals. Maybe we are all investing in the wrong equipment? Things to think about.
Yea, my wording was poor. I just meant nothing is a for sure thing and everything can fail. We just have to do everything we can to make the risk as minimal as possible. I agree with the data logging. Someone needs to come up with a mini black box for these things that is more user friendly and easy to operate and make sure it records every flight without having to worry about reading and clearing the memory every few flights. Who knows, maybe one day the FAA will require us to have transponders on these things.
Not raining on you but Dave, even multi million dollar department of defense UAV's fall out of the sky. There are no aircraft that I am aware of that are completely failure proof. Doesn't mean that they are not reliable but what MK builds is not the same level as a DOD item, nor the cost as you point out.
Dave: I was looking at the video in the original posting. The motors, right after they start up, sound a bit raspy. Also, it looks like the copter got airborne almost immediately after motor start. Did it throttle up on its own? You mentioned that you could not power off the motors -- thus my question -- did you have control of throttles after the motors started? Andy.
I should have added that I was emailing with Ingo to try and determine the exact conditions where GPX file might not be written. I do know, from reading the code some time back, that the throttles have to be advanced above 40% -- the FC waits for that condition before writing to the flight log file. Hence my question above. Sorry for the weird questions, but I'm just trying to figure out what might have happened from what evidence we have -- which is basically what you and the camera saw and heard. Andy.
Nothing is failure proof. It boils down to what amount of risk are you willing to take on the parts involved. Everything in the predator can and has failed more then once, resulting in a crash. I quit flying my RED for the above reasons. Period. If someone wants it in the air I charge $7500 day minimum and they gladly go back to the A7s/BM4k (which grades right into the RED footage no problem). I feel your pain Dave, after this (I2c fail) I never wanted to fly again. Holger was very understanding and made sure we were ok. The new octo has 200+ flights on it since the crash and has had no issues.
Dave, I say this with the utmost respect, but what you wrote is probably written on more than one tombstone. If we were meant to fly, we would have grown wings. Those of us doing this stuff know it's hard, and we do it in spite of the risks (although some idiots do it because of the risks, but I digress). Anyway, minimizing damage and risk is the goal, not eliminating it. At least right now. So you have to go with a system that you have faith in, that you can rely on most of the time, and always have a backup plan (and backup hardware). I know I am not telling you anything you don't know, but I didn't like the sound of what you said, and I had to speak up. Steve
It's all good Gary. I know what you mean but there is a difference between never failing and failing one out of every 60 flights. At least that's my average. I am also not trying to build a completely safe auto pilot. Not sure what the stats for a military type auto drone are as I don't think we have any idea how many flights each one performs as they are top secret. I guess I am still frustrated with the flight that occurred above in the video and having 2 boards fail on the same day.