APM vs the big boys (heavy lift octo's)

[FletchyFletch]

Does anybody have any experience of using the Ardupilot Mega 2.5 on a heavy lift rig with 3-axis gimbal?

I can't seem to find any useable information on it in this sort of set-up and just wondering whether I'd be stupid to try and use it over say the SuperX or Hoverfly Pro.

 

[Bartman]

well, the APM 2.5 is already outdated and the newer board (Pix4 I think?) should be a better but but we don't have widespread use here or documented good performance across a wide variety of users. It was said here a while back that the APM 2.5 with eight motors and a high demand set-up like GPS and navigation isn't such a great idea, better to use 6 motors with the APM 2.5 but it's an old system at this point so you'd have to refer to the 3Drobotics.com website to see what is the most current system.

Bart

 

[sixshooterstang]

The APM 2.5 is qite useable with most machines IMO. You have near infinite adjust of your gain settings so heavy and sluggish is not an issue.

However I do suggest you go with their newest system the PixHawk. It has more space on board for added functions and other fun stuff. It has similar characteristics to the PX4 board being that it is a 32-bit system.
http://store.3drobotics.com/products/3dr-pixhawk

 

[FletchyFletch]

Thanks for the replies. I went for the SuperX in the end.

One more thing I'm confused about here. I've spent weeks researching and fiddling about with ecalc and have finally come up with a reasonable calculation based on the Tiger Motor MN3510-25. Everything seems fine on ecalc but when doing manual calculations the motor rating doesn't match very well with the battery selection.
Am I missing something here? (see the picture)

The MN350 is rated at 15A continuous, which seems a little odd.

View attachment 15172

 

[ovdt]

You need bigger props or a higher KV motor for safe hovering. I would recommend to go with MN4012-400KV for that AUW.

 

[R_Lefebvre]

IMO, the reason why you don't see APM discussed here much, is not because of performance or anything like that. It's just because it's too inexpensive. No offence intended, but this is very much a "you get what you pay for" forum. It's just the prevailing perception here. It's often true, but not always.

I think the system is now fantastic, and can't be matched by any other system on the market for performance and features, at any cost.

I did discover a problem with the processor load on octos, but that has been fixed. There is no longer a flight performance issue. However, you will lose telemetry and data logging when in GPS flight modes. I don't like that, so I wouldn't recommend the APM for any serious octocopter.

 

[Bartman]

IMO, the reason why you don't see APM discussed here much, is not because of performance or anything like that. It's just because it's too inexpensive. No offence intended, but this is very much a "you get what you pay for" forum. It's just the prevailing perception here. It's often true, but not always.

I think the system is now fantastic, and can't be matched by any other system on the market for performance and features, at any cost.

I did discover a problem with the processor load on octos, but that has been fixed. There is no longer a flight performance issue. However, you will lose telemetry and data logging when in GPS flight modes. I don't like that, so I wouldn't recommend the APM for any serious octocopter.

Bob,

How can you generalize like that and then go on to say you wouldn't recommend it for a serious Oktokopter when that's exactly what he's asking?

We're a put-your-money-where-your-mouth-is forum....show us what works and that's where the attention flows. We may not be a forum that is fixated on any one product so we don't have the usual flocks of fanboys running around chewing out everyone that disagrees with them but that's precisely why I love this place!

I'm not against a 3DR revolution.....what are we missing that is keeping it from happening?

Bart

 

[R_Lefebvre]

Well, if you're comparing the APM to another system that doesn't even offer telemetry and data logging... does it matter if it doesn't work on an APM Octo ? Regardless, Pixhawk is here now, so this will become a non-issue.

I'll show you some videos I shot. This is with my own Octocopter. 2 axis gimbal only, so the pan is Arducopter. The yaw isn't quite tuned, and the frame might be slightly bent causing yaw problems. I dunno. I'm not a professional, I don't have a lot of time to play with this stuff. This is like the 5th video I've ever shot. I was doing this shot for a producer who wanted to showcase a roadway (this was just a test).

Here's one with some general flying around starting at 2:00. This is like my second time shooting video. I'm flying LOS, aiming the camera blind. Just to show you stability while actually moving. This is fully manual, no GPS or inertial stabilization at all.

Here's somebody else who's better at shooting video than me. This is from a cheap quad.

And then here is the upcoming killer app. This is Spline Nav. It should be out with the next release. Might be Pixhawk only. Keep in mind this was shot from a Phantom with a brain transplant. So the minor shaking... not APM's fault. This is FULL auto, hands off. You could program this in, run it, tweak it, run it again. Wait for talent to be ready, and do the final shoot.

So, are there any technical faults? I don't know of any objective way to demonstrate flight controller performance.

 

[Bartman]

Bob,

maybe you can answer this for me

why do we still even talk about APM? If Pixhawk is out and ready for retail consumption then why do we still talk about APM? What was the PX4 thing that we heard about? Maybe there's a branding problem with too many things being named variations of the same system? Explain please!

thanks,
bart

 

[FletchyFletch]

The cost of the APM/Pixhawk was one off putting factor for me if I'm honest. I've gone with cheaper ESC's and Tiger Motors because, through much research, they seemed to be pretty much the same systems as some of the "higher end" gear that's going for three times as much. I just couldn't justify going for Ardu-anything based on the fact that there is a huge lack of information on the subject for professional rigs. I've played it safe and gone for the SuperX but would far prefer to be on an open source system if I could have my way.

 

[R_Lefebvre]

Ok, first, my name is not Bob.

APM is still the default system. Pixhawk only because commercially available like last week, and they're still filling pre-orders.

PX4 is basically the predecessor of the Pixhawk. It's mostly the same hardware, but is on a compact double-stacked board. The Pixhawk is the same components arranged on a single board, all wrapped up in a pretty case that the market likes. That's about it.

 

[Gary Mortimer]

Did you mention the Pixhawk has enough processing power to be getting on with other things ;-) Spline Nav will be what all the others try and catch up with this year. Its exactly what film makers need.

 

[Bartman]

Ok, first, my name is not Bob....................

for whatever reason I thought you were a Bob. Rich? Ricky? Ralph? Renton?

Thanks for the info.

 

[R_Lefebvre]

Rob, just not Bob.

 

[Bartman]

Rob, just not Bob.

ok, now that we've got that sorted out......

i saw the press release on the pixhawk that you posted over at diydrones. can you please post the same info here?

it looks as if the pixhawk has moved far enough away from its arduino roots that it should have its own forum? maybe we can kick it off with the pixhawk press release info posted to the new forum area?

 

[dazzab]

I'm using the APM 2.6 on a very expensive SkyJib coax octocopter carrying a full frame DSLR with great success as are several of the developers of the Arducopter code. I also have it on 2 other Droidworx copters, two 3DR and an AquaCoper. I'm also running the software on a PX4 in a Bixler 2 and in the near future on a TechPod. Can you imagine what it would have cost to put proprietary commercial FCs on all those? I'm broke as it is.

I've had issues of course, which I shared in the forums here but the outcome has been a very stable system that just gets better and better. I simply can't imagine using a component that is so important to successful flight where you couldn't talk to the developers or read about the development in an open environment. The combination of 3DR and the team of developers that have come together is moving ahead at an incredible speed. It's really exciting to watch.

The APM is not deprecated and will continue for quite some time. It's a solid little performer that will allow many people to enjoy flying at a reasonable price with all the bells and whistles. But the team is also moving ahead with even more wonderful work that the APM simply doesn't have the capacity for in CPU or memory. The Pixhawk will be running the same exact software as the APM for now. But that software will be expanded to greater functionality on the Pixhawk and will run faster. There is still great value in using an APM and the most recent release of Arducopter. As time goes on I think you will see a lot of professionals adopting it for it's superior capabilities and the ability to be part of a community rather than just a customer that is typically ignored when there are issues.

 

[Bartman]

I'm using the APM 2.6 on a very expensive SkyJib coax octocopter carrying a full frame DSLR with great success as are several of the developers of the Arducopter code. I also have it on 2 other Droidworx copters, two 3DR copters and an AquaCopter. Can you imagine what it would have cost to put proprietary commercial FCs on all those? I'm broke as it is.

I've had issues of course, which I shared in the forums here but the outcome has been a very stable system that just gets better and better. I simply can't imagine using a component that is so important to successful flight where you couldn't talk to the developers or read about the development in an open environment. The combination of 3DR and the team of developers that have come together is moving ahead at an incredible speed. It's really exciting to watch.

The APM is not deprecated and will continue for quite some time. It's a solid little performer that will allow many people to enjoy flying at a reasonable price with all the bells and whistles. But the team is also moving ahead with even more wonderful work that the APM simply doesn't have the capacity for in CPU or memory. The Pixhawk will be running the same exact software as the APM for now. But that software will be expanded to greater functionality on the Pixhawk and will run faster. There is still great value in using an APM and the most recent release of Arducopter. As time goes on I think you will see a lot of professionals adopting it for it's superior capabilities and the ability to be part of a community rather than just a customer that is typically ignored when there are issues.

what I love about the multi-rotor community in general is that we're all slugging away at it and, for the most part, remaining very civil with each other along the way.

thanks for the info, i'm posting a new 3DR forum area and it will be up in a moment.

 

[raybans]

Is this limitation true also with the new Pixhawk controller ?

IMO, the reason why you don't see APM discussed here much, is not because of performance or anything like that. It's just because it's too inexpensive. No offence intended, but this is very much a "you get what you pay for" forum. It's just the prevailing perception here. It's often true, but not always.

I think the system is now fantastic, and can't be matched by any other system on the market for performance and features, at any cost.

I did discover a problem with the processor load on octos, but that has been fixed. There is no longer a flight performance issue. However, you will lose telemetry and data logging when in GPS flight modes. I don't like that, so I wouldn't recommend the APM for any serious octocopter.

 

[raybans]

Non 3dr APM boards

Does anyone have experience with the CUAV APM 2.5 board available from Hobby-Wing ?
http://www.hobby-wing.com/cuav-apm2-5-apm-flight-controller-board-and-6h-gps.html

I am planning a Quad-Co-ax Octa build using the Pixhawk controller but INITIALLY I plan to go with a quad only build and a cheaper (hence CUAV) controller while I familiarise myself with the world of Arducopter. :tennis:

I am currently flying F550's and Gopro/H3-2d with Naza controllers as my 'classroom' -

 

[R_Lefebvre]

Is this limitation true also with the new Pixhawk controller ?

Absolutely not. Full load on the Pixhawk is using <10% of the processor resources.