Pixhawk Feedback?

[Av8Chuck]

You hit one of my hot buttons. After 45+ years in the RC hobby I've watched companies dedicated to the furtherance of the hobby spend tens of thousands of $$ to deverlop new technology and superior products only to have it copied by overseas manufacturers that have contributed absolutely nothing to improving the hobby or the products used. Generally the copies have been inferior to the originals but they sold for less, which was all the justification a very lot of consumers needed to buy the copies. Much, if not most, of the software we use today was not developed in any way by those selling it for the least amount of money. The cloners don't give a hoot for the customer or product reliability, they carry no product liability, but people flock to them while ignoring the people that designed, developed, and initially marketed the product and still carry the product liability while providing customer support.

Ethics? Don't even mention ethics. Most, consumer or marketer, don't have any today and, I'm sorry to say, haven't for a long time. It's all about a buck and many get exactly what they paid for.

This is probably one of the biggest differences between hobbyists and professionals. Its one thing to purchase something that is poorly supported and fly it as a hobbyist, its a completely different thing to fly the same thing as a professional. As a hobbyist where hopefully the most you would lose would be the product itself, you might be able to justify purchasing the least expensive solution but if your flying in a situation where people could be hurt, property damaged and your reputation tarnished to the point where you can no longer be competitive and you go out of business, there's no way I'm trusting my career to a clone or knockoff.

Its difficult to even consider open source for the same reason, I'd like to be able to mitigate as much risk as possible and to do that effectively requires a whole other level of support and that goes in both directions, the consumer should expect to pay more for that support.

Most, if not all controllers developed in China rely almost exclusively on forums for their support so users end up supporting users. I don't want people to think I don't appreciate any help I've received on forums but that's not an effective way to grow a professional user base.

 

[R_Lefebvre]

Its difficult to even consider open source for the same reason, I'd like to be able to mitigate as much risk as possible and to do that effectively requires a whole other level of support and that goes in both directions, the consumer should expect to pay more for that support.

Yes, but actually that is a reason you should consider open source. So right now with Arducopter, you have the ability to pay somebody for any level of service you want. You can take free forum support from "the other place" where you tend to get people who have no idea what they're talking about. You can take free forum support from other websites where knowledgeable people participate. Or if you purchase from 3DR, you can get support on Ardupilot.com, where 3DR actually pays people who know what they're talking about to answer questions in a relative timely manner. With this, you're basically mixed in with all the other users. But the answers are usually correct and you can usually come to a conclusion to a problem within a week. If your problem actually turns out to be a bug, you usually get a fix for that on the next release.

But you have even more options. You can also pay somebody for a higher level of support. This could include "I don't have time to read the wiki, can you just tell me how to fix this?". To tuning support, from "here's my logs, what should I do?" to "if I send you this machine or fly you out here, can you get it flying for me?" You can even get custom code compiled for rapid bug fixes, or feature additions. All of this could be done by hiring an in-house person who is able to read the code because it is open source. Or you can hire a contractor <cough> such as myself </cough> who already has a demonstrated competence in the program. And, you can even pay one of the lead developers if you really really need something done.

You can't get that level of support from any other vendor that I'm aware of. Maybe some of the smaller American companies like Freefly? I'm not sure about that though.

For example, I have a client who wants to have a working demonstration of Lidar-based collision avoidance working within 2-3 weeks. Fairly basic, just "don't hit a wall despite bad pilot input or GPS error". Is that sort of thing an option with any other system?

 

[R_Lefebvre]

Oh, and in other news... a new era of Arducopter has begun. I've finished construction of a prototype gas-powered SRH, and successfully completed initial test hops this week. This is an 800-size machine with a Zenoah G29RC engine. My target for the project is to be able to fly with a 10kg payload for a significant period of time. The target is more industrial uses rather than commercial AP. So, spraying, or mapping with a heavy camera over remote and secured areas (wilderness, open pit mines, etc).

These machines could be built for about $6-8000 each, and you could fit several of them in the same transport space as an octo capable of similar payload. With blades folded it's only about 54" long, and 12" wide, ~12 lbs empty, and only 14lbs fueled and read to lift 10kg payload.

 

[Old Man]

If you were mass building what would your thoughts be about a 34cc engine that weighs less than the Zenoah, similar foot print, has better cooling, runs on gas but also possibly heavy fuel and/or injected? The injection stuff adds weight and electronic complexity though.

 

[R_Lefebvre]

I love injection! Carburettors are Byzantine technology!

Definitely sounds like an interesting engine. Even better if it was a 4-stroke.

 

[Old Man]

Too heavy. The engine is out there but not a hobby item.

 

[JoeBob]

Pixhawk Auto Flight: SuhWheeeet!

Got out to play with Spline Waypoints today. Starting simple.
Flight Plan:

69 Seconds of Air and Grass:
[video]http://vimeo.com/user25835851/review/106550094/929ad03cc5[/video]

Totally autonomous flight. Auto Take Off, etc. The chair wasn't exactly in the middle of the ROI; funny, it wasn't on the satellite imagery so I was guessing...
I liked how smooth the rotation was. Not jerky; no yaw jello.

I left the last 15 seconds on the video so that y'all could enjoy how that AlexMos Gimbal slyly corrected its lost horizon. "What? Me tilt with Yaw?"

I need help getting the altitude smoothed out. An earlier flight was set to 4 meters, and flew at .5 meters. I also had my Nav Speed set to 500 for that flight, so it was hauling *** in a circle just off the ground!

Baro Glitch Protection Distance is set to 200 (100 - 2000 range) (Pixhawk between the plates on a TBS Disco Pro- somewhat protected from prop wash)
AHRS GPS Gain was maxed at 1.000
Suggestions on how to balance GPS vs. Barometer for a steadier height?

I thought that 'Verify Height' was checked, but it's not checked now...

Would a Ublox NEA 8 GPS module be that much more accurate than the NEA 6 that I got from 3DR?

I've ordered a Sonar Sensor with the Pixhawk for my 2nd MR. Has anyone tried mounting the Sonar at 45 degrees down in the direction of travel to estimate the height the MR needs to be at before it gets there? (My goal is to cut nap of the earth circles 1 to 2 meters high at 500ms.)

I am so bloody happy to get this Pixhawk/Disco flying like I knew it could. I'm still scrambling to keep from falling off the learning curve.
Damn! What fun!

Now I have time to clean that smudge off my lens.

 

[R_Lefebvre]

Too heavy. The engine is out there but not a hobby item.

4 stroke doesn't have to have a worse power/weight ratio than 2-stroke.

 

[Old Man]

Ya know, I keep hearing that from people but for the last 10 years I have yet to see an affordable 4 stroke that didn't used mixed oil in an under 100cc size class that didn't weigh more for less power. BTW, this is what I do for a living now

JoeBob,

You really can't know how pleasing it is to hear success stories like yours!!!

 

[R_Lefebvre]

Most of my experience is from larger motors. You're probably right, that in the under-100cc class, maybe it's not possible. There might be economies of scale for the valve-train that just don't work out for small engines. But in larger sizes, a modern, fuel injected forced induction 4-stroke can blow away a 2-stroke on power/weight. But not power/weight/cost, that's for sure.

 

[R_Lefebvre]

Very chuffed with this:

 

[Av8Chuck]

Is "chuffed" a good or bad thing?

 

[R_Lefebvre]

Good.

Flew 2 tanks today, over 20 minutes each. I have it totally dialed in now. I accidentally hit 108 km/h just tooling around. :highly_amused: Have to recalibrate my brain to the scale of this thing to judge speed.

Vibrations are quite low:

That's actually very surprising. Better than both of my quads.

Will start working on auto missions now. And I'd like to start doing payload tests. But I need to come up with a solid mounting method.

 

[Old Man]

Is "chuffed" a good or bad thing?

Glad you asked, I was wondering the same thing

 

[jfro]

Glad you asked, I was wondering the same thing

according to 1st up on google...

• very pleased.
  "I'm dead chuffed to have won"
  
  
  
  
  

 

[jfro]

Any Pixhawk news on Multirotors?????

Mine has been sitting a while and maybe time to spend some more time with it......

 

[Old Man]

See how you are? Back door all of us with a Google search, lol. Good one and I wish I thought of it.

 

[Old Man]

Any Pixhawk news on Multirotors?????

Mine has been sitting a while and maybe time to spend some more time with it......

Whatcha mean by news? I've found flight modes available I didn't have with APM 2.6 and using Pixhawk things are working out pretty darn good overall. It's gettin used on larger and larger MR's with great success. Taking that to another level is that a lot of other controllers can't dial in the really large copters due to user input limitations with the software, where Pixhawk lets you walk right in an make adjustments as needed through Mission Planner without needing to be a software engineer. I wish I knew more about all of it than I do but having something new in front of you to explore everyday isn't a bad thing. The uBlox GPS is working out to be stellar.

 

[jfro]

See how you are? Back door all of us with a Google search, lol. Good one and I wish I thought of it.

That's the only way I know what some of the internet abbreviations and terminology means. Use it more than I'd like to admit.....

 

[Av8Chuck]

See how you are? Back door all of us with a Google search, lol. Good one and I wish I thought of it.

We don't think of doing that because we're OLD...