This is in response to some feedback I got last fall on this forum:
I think the relevant parts to follow on are that the 250x250 scanning area is probably most suitable for us. Still looking at the same payloads I mentioned previously (the Cannon, the IC thermal, and ADC Tetracam). I think we'd like to go RTF right now since we're quite new to this approach, but it looks like the X8 isn't available anymore at the link you sent - this is the newer model? http://store.3drobotics.com/products/3dr-rtf-x8-2014
It looks like there's a bit price difference between the X8 and the quadrocopter models (or is the X8 just not spec'd out with the full programming and navigation options we'd need?). Perhaps a chat offline would be ok and more efficient?
I've also had a recommendation from the makers of a multi-band camera we want to fly: that they've had success with and recommended the agscout extreme from agriimage: http://www.agriimage.com/products/
Any experience or thoughts on that company?
The agscout's ability to handle 30 MPH winds was appealing as a collaborator is working at a flux tower at the top of one of our local mountains, gusts are an issue - not sure if the other options that are being discussed here would be appropriate.
Yes that is the latest X8. I've used it, I like it a lot. I'm not sure what you mean by price difference, 3DR's quad is $750 but has a very limited payload capacity.
Never flown that Agscout copter, we've only really used Mikrokopter and 3DR parts.
As for wind, I thinks that "maintain hoovering position in 30 mph winds" is a bit misleading. Yes, it can stay in one spot. But can it fly at cruising speed in high wind? That is what you need, since scans are taken on the move. So able to hover at 30 mph winds seems pretty standard for a heavy copter, actually. Those large landing gear, antennas, and camera set up look to be of just an average resistance to wind. I recently flew one of our light hexas in gusts up to 20, and it was considerably slower flying into the wind. So I at least wouldn't use that quote as selling point for this copter, it doesn't sound out of the ordinary at least in my experience/opinion.
Really it depends on how much you are willing to spend. For a basic RTF copter for flying 250x250 meters, 3DR is probably the best value I've seen. You can drive the price down by getting into the bits and pieces and designing/building your own copter, but that requires a higher skill level. If you have more funds to use, you can start looking at higher cost copters with more features, like the Agscout you linked. Off the top of my head, I can also think of the Draganflyer copter, another high end one.
I would love to build a custom copter system RTF for you, but ongoing maintenance would be a big issue because I would be the only one familiar with its specs, and I can't guarantee I would be available for support on an ongoing basis. So that's not really an option.
Any more questions?
Thanks again Stephen!
On cost - what I meant was that when I spec'd out the Quad vs 3DR (with some extra batteries, way-point capability, spares, etc.), the quad was maybe $1500 more than the 3DR. Agriimage might be pushing the budget, but we can certainly do the Quad or 3DR. The Agscout and Dragen get a little pricier because of the view as you go options and flight controls?
I agree on both sides of the custom model from you - that would be best, but we really aren't anticipating being able to handle the support side of it right now - perhaps later on?
I'm assuming that working with image processing wont matter too much what copter we work with?
Hmm, which quad are you looking at? I think we may be talking about two different products.
Dr. Ellis thinks I should do a start-up company for UAV work like mapping, building copters, and training. I am certainly considering it, but it's a big maybe at this point. Maybe in the future though.
And as far as image processing goes- you are correct. As long as the copter can go where you want it to, you should be fine. There's the possibility you could have issues with blurry pictures on some copters due to vibrations, but this is usually remedied by adding dampers to the camera mount, so I wouldn't worry much about it.
This is the quad: http://www.quadrocopter.com/HexaKopter-2-Ready-To-Fly_p_761.html
It sounds like there is no shortage of need for start-ups to do that UAV work (there are lots of ag opportunities and applications here in Southern Arizona even)
Haha, when you said quad I thought you meant a 4-rotored craft, not a copter from Quadrocopter. We use a modified version of the hexa 2 (ours fly with Arducopter brains and electronics from 3DR). It (the unmodified model) has some major restrictions such as not being able to fly further than 250 m from the take-off point. Plus it is more expensive for a smaller copter. The price point here is brand. Mikrokopter (being sold by Quadrocopter) is one of the oldest players in commercial multicopters, which is why it is much more expensive than 3DR's X8. I definitely would recommend 3DR over them, due to both price and ease of use.
I'll butt in here and offer some unsolicited advice. As someone else who has used both MK and APM extensively, I agree that you should go with APM. MK will work great at first but the more you learn and the more you want to do with it, the more restrictive you will find the system to be. APM's mission planning software is also much more robust than MK's, IMHO. APM still has its limitations and nuances, and MK beats it in some areas (onboard logging of flight data comes to mind) but, all in all APM is a more powerful and more flexible system. Then, if you go with the Pixhawk over APM (which I think comes standard in the new 3DR copters) you will have a highly capable controller that is no longer limited by hardware in any way.
No such thing as unsolicited advice on here, thanks!
Here's another high end high cost RTF solution, looks fancy:
Hi all, this discussion has been helpful as I consider possible new equipment. Stephen, it sounds like you have some experience with the new Pixhawk flight controller. Can you or anyone else elaborate on this? Basically I wonder if Pixhawk is better than the APM for stable autopilot flight.
I have been having difficulties lately with my 3DR quad copters and winds. For example, the other day I was trying to fly auto missions when the winds were forecast to be 9-11mph. I used to consider this to be an upper limit for autopilot flight, however the copters were having serious trouble dealing with gusts, and one of them crashed from its auto mission at 130m altitude during a strong gust. Granted the gusts did seem to me to be greater than 11 mph, although that was not forecast.
I am using 3DR quad C and D frames from last year, which are now discontinued. I recently replaced an APM 2.5 with a 2.6 because after several crashes, the old 2.5 was behaving pretty erratically in autopilot flight, even in light wind. I wonder though if a Pixhawk would be much better? The 3DR website seems to speak pretty highly of them.
A confounding factor here could be that the 4 propeller quads I am using may also be inherently less stable than a hexa or an X8 would be. I don't hear too much discussion here about 4 propeller quads. Are these still in much use among folks like Ecosynth, who are serious about getting high quality aerial photos?
Thanks for this and all the other helpful advice
Hey Steve. The short answer is: the Pixhawk will not be any more stable than the APM2.6 because at this time they are still running the same code.
Long answer: Although the Pixhawk has more processing power, currently it is running the same Arducopter code as the old system, so it won't be any better. The Pixhawk does have the potential to run much more software. For example: it can log more variables like the IMU output during the flight. HOWEVER: there is an update in the works which will only run on Pixhawk which takes advantage of the Pixhawk's advanced processing capabilities. At some point in the future, a properly configured Pixhawk will be more stable than an APM board.
In my experience with 3DR frames: the more times you crash and repair, the less stable your flight is. This is due to small damage building up. For example: motors become mounted askew due to the metal they are anchored to bending. Even if you straighten them out or pound them flat, they will never be quite the same. The other big thing is propeller shafts getting bent. The props still spin fine, but they are off balance. This introduces high amounts of vibration into the frame, which is terrible for stability due to the APM's accelerometers being susceptible to vibration. If your performance seems to be degrading, I would definitely blame the aging hardware. I have never found the APM itself to wear out unless it is damaged.
I've never actually used a quad for taking photos. I built one as practice early on, but we've never taken it up for a mission. I know as a rule of thumb: as the number of propellers goes up, so does the stability in wind. I recently flew one of our hexas in high wind, and there was no stability problem. Of course, there was a power problem. When flying into the wind the copter only got up to 2-3 m/s and it tilted at a crazy angle over 30 degrees since it was pushing so hard to fly into the wind. But it had no trouble staying stable.
Basically I would recommend making sure you have good hardware before you start upgrading autopilots, and a copter with more props will probably handle wind better.
Thanks for your reply Stephen. After trying a few things, eventually I realized the problem was most likely related to how powerful the motors were. When a new version of the arducopter software came out, I think earlier this year, there was a new parameter called THR_MID. This parameter represents how much throttle is necessary to maintain a stable hover. I believe is used as the basis of the ALT_HOLD and AUTO flight modes. It also seems like a straightforward way to tell if the aircraft is over- or underpowered.
I followed the procedure here to set THR_MID: http://copter.ardupilot.com/wiki/ac_throttlemid/
And discovered that my quadcopter, with 850 kV motors, was underpowered with my payload, which consisted of a 5000 mAh battery, a digital camera, and a dog collar GPS. So I got more powerful 880 kV motors, and then the THR_MID parameter setting was within the recommended limits. Since then the AUTO mode has been much more stable and consistent, based on how the flight logs look in google earth, and there have been no crashes, even in moderately windy conditions.
So it seems that quadcopters can be used for stable flight with a camera and a dog collar GPS, as long as the motors are powerful enough. 6 or more propellers would probably be more stable, since this would be an inherently more powerful and redundant setup. And damaged equipment of course wouldn't help anything. I did replace the damaged arms on the quadcopters, and have essentially brand new 880 kV motors that have never been crashed. With this the vibrations measured by the autopilot are well within the limits described here: http://copter.ardupilot.com/wiki/ac_measuringvibration/
Here are pictures of the same flight before and after I changed to the 880 kV motors, showing how they maintain altitude better.
Glad to hear it! I hadn't thought of being underpowered as a possible problem, but yeah that looks like it. The rule of thumb for hobbyists is this: with a fresh battery and flying in manual mode: the throttle should only have to be at about 50% on the radio controller to hover the craft. If you have to crank the throttle up higher to make it hover, then you are overweight/underpowered. Glad you're back in the air!