R&D Hex for Survey/Imaging/Conservation

[crayfellow]

Hi all,

We are a small product design firm embarking on our first major R&D project to explore aerial survey, imaging, and conservation opportunities. I'm also very interested in looking for new open source contributions we can make to reciprocate for all the great work that has already been done.

Here is the build so far, and I'm being as explicit as possible in the hopes of helping others:

• Frame: QuadFrame SIXcopter foldable with 25mm tubes in 400mm length (for up to 15" prop)
• Motor: KDE 3520 400kv
• ESC: KDE 55A
• Battery: 6S 2x8000mAh?
• Prop: APC 15x5.5 for testing/tuning, then Xoar 15x5 wood to T-motor 15x5 CF
• Radio: Taranis Plus X9D Tx / X8R Rx
• Flight Controller: Pixhawk (ended up with Fixhawk which has been working great and seems at least as well made as 3DR)
• GPS/Compass: u-blox M8N with HMC5883L mag
• Voltage/Current sensor: Attopilot 180A
• Camera: GH4 with Rokinon 12mm cine
• Gimbal: HD Air Studio InfinityMR
• FPV camera: Lumenier CS-600
• FPV Tx: Immersion 600mW
• FPV ant: IBCrazy bluebeam

Things I haven't yet figured out:

1. Motor/prop/batt; seems like a sensitive subject for some on other forums! With a limit of 15", and considering I'm hoping to power everything via BEC and filters where necessary (is that practical?), what motor/prop/batt would you suggest to balance stability, reliability, flight times, and reasonable performance?

Thoughts? I will try to keep this up to date as the build progresses and we learn more.

 

[crayfellow]

Here is the mounting for the InfinityMR gimbal:

 

[Mactadpole]

Hey crayfellow,

I do work along the same lines with my copters. I have recently completed a very similar build to what you propose.

http://www.multirotorforums.com/threads/quadframe-foldable-pro-sixcopter-medium-lift-need.22993/

I am now working on getting the gimbal working with the CP controller. I have run into a few problems in the process that I am still trying to work out. I seem to have a lot of camera shake/vibrations and I am trying to isolate where it is coming from. We've had bad weather for a while now and just can't seem to find flying time to get it worked out. I think my cheap CF props are one of the sources of problems. I am considering the KDE triple prop adapters with DJI 1552 props but I don't want to lose any flight time or at least not much. With my current build I am getting flight times of about 25 minutes which I am super happy with.

I have made a few mods since starting and will get those posted on the build pages soon. Main one is that I cut the battery shelf in half and mounted each half on either side of the FC for better COG.

I just measured on my center plates (don't know if its the same as the frame you propose) and the InfinityMR mount will work but holes would need to be drilled. I would recommend you take a look at the Quadframes folding hexa frame. It is super convenient to fold those arms in. I also asked Jakub if he was going to offer a smaller version of the InfinityMR for nex and A5100/6000 size cameras and he said yes. With the encoders that gimbal looks fantastic but a bit big for the cameras I would ever plan to carry. I'll post an update on my gimbal progress once I get to test it some more. I looked into the Arris when I was shopping and for multiple reasons I decided no.

Best,

Shawn

 

[crayfellow]

Hey crayfellow,

I do work along the same lines with my copters.

Hi Shawn! I would love to learn more sometime. Is there something we could study/design/develop to help you and similar pros?

I have recently completed a very similar build to what you propose.

http://www.multirotorforums.com/threads/quadframe-foldable-pro-sixcopter-medium-lift-need.22993/

that is so perfect! I am reading the whole thing and hope not to bother you too much with questions, but I may not be able to resist.

I am now working on getting the gimbal working with the CP controller. I have run into a few problems in the process that I am still trying to work out. I seem to have a lot of camera shake/vibrations and I am trying to isolate where it is coming from. We've had bad weather for a while now and just can't seem to find flying time to get it worked out. I think my cheap CF props are one of the sources of problems. I am considering the KDE triple prop adapters with DJI 1552 props but I don't want to lose any flight time or at least not much. With my current build I am getting flight times of about 25 minutes which I am super happy with.

Interesting. I know props are important, so I'm trying to balance that cost with motors (would like to try for KDE motors and ESC's). But yeah, others seem to be having luck with wood props, right?

I have made a few mods since starting and will get those posted on the build pages soon. Main one is that I cut the battery shelf in half and mounted each half on either side of the FC for better COG.

I just measured on my center plates (don't know if its the same as the frame you propose) and the InfinityMR mount will work but holes would need to be drilled. I would recommend you take a look at the Quadframes folding hexa frame. It is super convenient to fold those arms in. I also asked Jakub if he was going to offer a smaller version of the InfinityMR for nex and A5100/6000 size cameras and he said yes. With the encoders that gimbal looks fantastic but a bit big for the cameras I would ever plan to carry. I'll post an update on my gimbal progress once I get to test it some more. I looked into the Arris when I was shopping and for multiple reasons I decided no.

Yep I am thinking of the GH4 now, and I like the QuadFrame gimbal since it would require no frame modifications. I will be using an AlexMos 32-bit controller with dual IMU's, so I will hopefully avoid the integration issues others have had with the CP controller.

I did get the folding SIX, just arrived today in fact!

Still need to figure out motor/esc/prop/batt.

Cheers.

 

[Mactadpole]

Hey Crayfellow,

let me think about first question some. Mostly, I am looking for a grad student interested/motivated to seek more funding and start some projects I have in mind.

I really like the prop mounting style of just having the disc with two screws that go through it to hold prop on but may have to leave that style depending on what route I take. The highprop brand wood props I got are crap too. I think on this larger builds these big props can really have a much more pronounced effect on vibrations than the smaller prop'd copters. Never had this much issue with props before.

The GH4 seems big to me for what I am trying to do. The lighter the camera the more flight time. I think encoder based motors with controllers might be our current best solution so I would hold off just a bit on that gimbal decision. Have you seen the results Jakub has gotten with the InfinityMR?

I would highly recommend eCalc for your motor/esc/prop/batt combo. That is how I figured out the setup I have. I am certainly curious if there is a better one for this frame.

Will keep in touch.

Best, Shawn

 

[crayfellow]

I really like the prop mounting style of just having the disc with two screws that go through it to hold prop on but may have to leave that style depending on what route I take. The highprop brand wood props I got are crap too. I think on this larger builds these big props can really have a much more pronounced effect on vibrations than the smaller prop'd copters. Never had this much issue with props before.

Patrick at KDE Direct suggested T-Motor CF props. What do you think?

The GH4 seems big to me for what I am trying to do. The lighter the camera the more flight time. I think encoder based motors with controllers might be our current best solution so I would hold off just a bit on that gimbal decision. Have you seen the results Jakub has gotten with the InfinityMR?

Yep, I have. I also made the mistake of thinking Jakub ("Jakub J" on rcgroups, QuadFrame guy) and Jakub ("jakuban" on rcgroups, HD Air Studio guy), both in Poland, were one and the same!

Between the apparent adjustability/flexibility and option for encoders the InfinityMR gimbal looks like an excellent option. Since we may well be mounting "unique" or modified cameras, this is important to me.

I would highly recommend eCalc for your motor/esc/prop/batt combo. That is how I figured out the setup I have. I am certainly curious if there is a better one for this frame.

I looked at it, and despite a background in electrical engineering, I found it extremely intimidating. Much of my build is sort of up in the air, so it's a bit of a chicken/egg sort of thing. Since I'll use 14" props, I think a KDE 3510 475kv (or equiv), 14x4.8 CF props, and the KDE 35A ESC (or equiv) should be workable. My concerns with that setup is the substantial price difference vs. something like SunnySky/T-motor and ZTW Spider ESC's, a setup that plenty of pros are doing just fine with. I love the idea of supporting another US-based company, but ought not do so without solid justification as these funds are precious and limited. I also need maximum assurance that we're not putting priceless or expensive equipment in the air with a high probability of failure. It's a tricky balancing act. What do you think?

Also if we do go all out like this, what about KDE prop adapters as well?

Patrick at KDE also suggested 6S GensAce Tattu batteries. These appear to be nearly $300, do I have that right?? I want to be sure to add context to that recommendation with suggestions from others running highly reliable craft.

Cheers

 

[JoeBob]

Your flight controller, Pixhawk, supports up to 4S batteries natively. If you go with the 6S batteries/motors, you'll need to get a different power module to use a Pixhawk.

 

[crayfellow]

Your flight controller, Pixhawk, supports up to 4S batteries natively. If you go with the 6S batteries/motors, you'll need to get a different power module to use a Pixhawk.

Right, I was thinking the Attopilot 180A.

 

[crayfellow]

I also asked Jakub if he was going to offer a smaller version of the InfinityMR for nex and A5100/6000 size cameras and he said yes.

That was the camera I was thinking of originally. Do you use it for stills, or video as well?

 

[crayfellow]

let me think about first question some. Mostly, I am looking for a grad student interested/motivated to seek more funding and start some projects I have in mind.

Yeah.. that's a tricky part isn't it? Are you in touch with Conservation Drones? Maybe they can help.

We won't be good at helping to line up funding, but we are self-funding our own work hoping we will stumble into something beneficial for conservation, ag, forestry, etc.

 

[fltundra]

Patrick at KDE Direct suggested T-Motor CF props. What do you think?



Yep, I have. I also made the mistake of thinking Jakub ("Jakub J" on rcgroups, QuadFrame guy) and Jakub ("jakuban" on rcgroups, HD Air Studio guy), both in Poland, were one and the same!

Between the apparent adjustability/flexibility and option for encoders the InfinityMR gimbal looks like an excellent option. Since we may well be mounting "unique" or modified cameras, this is important to me.



I looked at it, and despite a background in electrical engineering, I found it extremely intimidating. Much of my build is sort of up in the air, so it's a bit of a chicken/egg sort of thing. Since I'll use 14" props, I think a KDE 3510 475kv (or equiv), 14x4.8 CF props, and the KDE 35A ESC (or equiv) should be workable. My concerns with that setup is the substantial price difference vs. something like SunnySky/T-motor and ZTW Spider ESC's, a setup that plenty of pros are doing just fine with. I love the idea of supporting another US-based company, but ought not do so without solid justification as these funds are precious and limited. I also need maximum assurance that we're not putting priceless or expensive equipment in the air with a high probability of failure. It's a tricky balancing act. What do you think?

Also if we do go all out like this, what about KDE prop adapters as well?

Patrick at KDE also suggested 6S GensAce Tattu batteries. These appear to be nearly $300, do I have that right?? I want to be sure to add context to that recommendation with suggestions from others running highly reliable craft.

Cheers

I've been using KDE motors since they were released and won't consider using anything else at this point. You made the perfect choice for motors (3510 475kv) in the 12 to 16 inch prop range, and Patrick and Chris have always gone out of there way when it comes to customer support.

 

[ary]

I've been using KDE motors since they were released and won't consider using anything else at this point. You made the perfect choice for motors (3510 475kv) in the 12 to 16 inch prop range, and Patrick and Chris have always gone out of there way when it comes to customer support.

for quad config using KDE 3510 475 kv using 2X 6000mah 35C. prop max 14x4.7.

 

[fltundra]

for quad config using KDE 3510 475 kv using 2X 6000mah 35C. prop max 14x4.7.

My current Config. 1st photo KDE 515kv.
Second is with my config with 475kv and Tattu 6S16000 15C spinning 14x4.8 tmotor cf's.
View attachment 24646 View attachment 24647

 

[ary]

My current Config. 1st photo KDE 515kv.
Second is with my config with 475kv and Tattu 6S16000 15C spinning 14x4.8 tmotor cf's.
View attachment 24646 View attachment 24647

Its a good idea to try single battery Tattu 6S16000 with low 15C. It is possible to use 15X5 prop. Just try to fliy it less than 5 min. check motor & ESC temp

 

[fltundra]

Its a good idea to try single battery Tattu 6S16000 with low 15C. It is possible to use 15X5 prop. Just try to fliy it less than 5 min. check motor & ESC temp

We are talking 240 amps continuous at an IR per cell of 1.7 or less, not going to be a problem.
Also, you should be strapping your bird down and testing temps and current draw way before you ever fly.

 

[Mactadpole]

If I could afford the KDE then I would get them. Still don't think I would go for Tattu batteries and definitely not a single battery setup. Part of all this is personal choice too.

I am now leaning towards the Tiger CF props myself as opposed to the KDE triple blade folders because of suppossed 5-10% efficiency loss. I am very happy with the flight times I get. I have a set of wood props that are 16 x5.4 (http://www.himodel.com/plane/HiPROP_16x5.4_inch_Beechwood_Propeller_Set_for_Multicopters_CW_CCW.html) and the hubs are so out of balance I pretty much can't put enough CA on them to get them balanced. But I am going to fly them to see what flight times I get with those and then decide whether I go with 15 or 16" CF props of high quality.

I currently use a Nex 5t camera but will be stepping up to the A5100/6000 with a Sigma 19mm lens. I too am hoping to move to some multispectral sensors too. The sony cameras are great for stills and video.

I have had some contact with Conservation Drones and they are a great resource. I have started one project in Ecuador: http://diydrones.com/profiles/blogs/drones-for-ecology-and-conservation-over-ecuador-s-pacific
What I am looking for now is to be an advisor for one or two self-motivated graduate students who really get this program off the ground in my lab.

Shawn

 

[Mactadpole]

Something is wrong with my eCalc subscription at the moment but once I get that straightened I can plug some numbers in for you and will post.

 

[fltundra]

If I could afford the KDE then I would get them. Still don't think I would go for Tattu batteries and definitely not a single battery setup. Part of all this is personal choice too.

I am now leaning towards the Tiger CF props myself as opposed to the KDE triple blade folders because of suppossed 5-10% efficiency loss. I am very happy with the flight times I get. I have a set of wood props that are 16 x5.4 (http://www.himodel.com/plane/HiPROP_16x5.4_inch_Beechwood_Propeller_Set_for_Multicopters_CW_CCW.html) and the hubs are so out of balance I pretty much can't put enough CA on them to get them balanced. But I am going to fly them to see what flight times I get with those and then decide whether I go with 15 or 16" CF props of high quality.

I currently use a Nex 5t camera but will be stepping up to the A5100/6000 with a Sigma 19mm lens. I too am hoping to move to some multispectral sensors too. The sony cameras are great for stills and video.

I have had some contact with Conservation Drones and they are a great resource. I have started one project in Ecuador: http://diydrones.com/profiles/blogs/drones-for-ecology-and-conservation-over-ecuador-s-pacific
What I am looking for now is to be an advisor for one or two self-motivated graduate students who really get this program off the ground in my lab.

Shawn

I had the same issues with a set of the tmotor beachwoods, they are junk compared to the cf's. Luckily, RMRC took them back.
As far as i am concerened, Lipo handling is main cause for failure. Your gaining nothing by running two batts. Your just creating more complexity.

 

[Mactadpole]

I had the same issues with a set of the tmotor beachwoods, they are junk compared to the cf's. Luckily, RMRC took them back.
As far as i am concerened, Lipo handling is main cause for failure. Your gaining nothing by running two batts. Your just creating more complexity.

Good to know about the tmotor beachwoods.

Like I said, a lot of this is personal preference. Definitely taking good care of batteries is essential. I agree with creating more complexity but at the same time to me I feel like I am adding some redundancy. Especially since I have had a power module meltdown on a single feed setup that let the copter just drop from the sky. I now set it up with two separate feeds, one for each battery to PDB. One of these goes through the power module connected to the pixhawk and the other goes straight to the PDB and I have redundant power to the Pixhawk in case of module failure.

 

[crayfellow]

If I could afford the KDE then I would get them. Still don't think I would go for Tattu batteries and definitely not a single battery setup. Part of all this is personal choice too.

Yeah. I don't know what I will do for batteries but probably will do the KDE motors/ESC's. As an engineer I'm not terribly comfortable accepting that there is some secret sauce in the KDE ESC's vs. anything else, though, and that is an important consideration since it significantly increases the overall cost. They may have some proprietary alg's in firmware, but I'd prefer to license their tricks so I can have the code and have something more sustainable for cost-effective replacements, etc.

As for my budget, I don't have a specific number although I'm trying to spend as little as possible for a solid, stable platform for R&D. Sometimes that may mean going nuts (KDE) other times it may mean DIY or other tricks to keep overall cost down without introducing undue risk.

I should say as an absolute beginner in multirotors it is pretty freaky to be putting $xxxx precious R&D budget in the sky so I have a mini quad practice rig, and a few simple rovers to test anything that doesn't require being in the air.