A word of CAUTION about the "water soluble flux" just mentioned (example: Kester 331). If you don't remove ALL of that flux, it is extremely corrosive! We do not use that flux for anything we do not intend to clean in an ultrasonic cleaner and we would never use it down inside a connector barrel. In my opinion you are MUCH better off using the far more active rosin core flux solder. You can clean or not clean that as you wish. It's just a cosmetic issue with rosin. There are also "no clean" fluxes like Kester 245 out there. They are cosmetically a little better without cleaning I suppose, but the flux is far less active. Best to stick with the "good old stuff"!
Now we are soldering wires directly to the motor's wires, which eliminates (12) M/F 3.5 mm bullet connectors in the whole system. Connectors are a point of failure - they are worth getting perfect, every time if you must have them.
Good call Ken! That is also the way I have done mine. Since you would have to solder connectors anyway, may as well just solder wires and use heatshrink. Currently mine are not buried inside the tube yet.
Just ensure the connections are the correct way around for the motor rotation (2 sets are normally reversed).
Hey Cluso99, we don't use those propeller saver O-rings. The force exerted on the propellers guarantees failure of the o-rings. Instead we are screwing the nut and washer down on the propeller.
Thanks ken. Mine had no force and failed just sitting there. Do your motors have threaded shafts or are you using the spinner style connectors?
I just finished a new frame for mine - It's quite a bit lighter and much neater than my previous one. I used pairs of square carbon fiber tube for the arms, sandwiched between custom body plates and motor pods I cut from 1/8" Delrin with my CNC machine. I added LED lighting strips to the front pair of arms to help with orientation (headlights!) and I'm very happy with how it turned out. I may machine another body piece for the top to serve as a camera mount, but it works just fine like this.
You can't tell from the pictures, but the body plates have counter-sunk slots cut into them to make sure the arms are properly aligned and angled, and all the holes, including the ones for mounting the ProtoBoard, were drilled at the same time.
I just finished a new frame for mine - It's quite a bit lighter and much neater than my previous one. Jason
Nice work - I'll even change the title of this thread since you're back in the quadcopter game. I'm truly inspired by the way you've assembled the electronics. Way to go!
Jason, that's a mighty fine setup. I agree with Ray, the dual booms look great.
I like the way you kept the ESC's in the center. Should be good for balance and faster response due to little weight on the extremes.
Thanks for the compliments, everyone. Keeping the ESCs centered and running the wiring through the booms is something I've seen on many other quad / hex copter builds, and I've always liked how tidy it looks, so I copied it for this one. Keeping the mass closer to center should help make it more agile, but I think my ability as a pilot is the limiting factor right now. I need to start practicing again.
I'm considering doing it again but with shorter arms - They're about 13" right now, but I have enough tube left to make a set of 9's. I didn't want them too small for this one because I want to mount my GoPro on it, and want the props out of frame as much as possible.
I also still have some oscillation so I'm going to try setting up the board with an LCD display and alter the programming so I can change PID values between flights. That should help me find some better settings and hopefully get it a little more stable.
In post #273, you say that the landing skids are cut from 1/4" delrin, but your BOM in post #305 only lists 1/8" delrin? Anyway, for cutting one quadcopter legs that is 6:00 minutes (1:30 per quadcopter).
From the pictures in #317, it appears that there are the following delrin parts:
4 - Skids
8 - Motor mount pieces
2- main body plates
1 -mini body plate
Assuming that to cut 1/8th delrin takes half the time as 1/4 delrin, I estimate that two of the motor mount plates would take as long as one skid, each one of the main body plates has as much cutting as four legs (or 2 leg times), and the small plate has one leg worth of cutting. So, that is
4 - Skids (6:00)
8 - Motor mount plates (6:00)
2 - Main body plates (12:00)
1 -Mini body plate(1:30)
So, ~26minutes of cutting time. Pololu will do cuts for $2.50 per minute, so to cut one quadcopter will cost about $65.00. The delrin will cost approximately $85 for a sheet, that is enough to make four (est.) quadcopters (http://www.mcmaster.com/#delrin-(made-with-acetal-resin)/=f6brff). So, the subtotal is $65.00*4 + $85 = $345, plus 15% for shipping costs so the grand total would be ~$400, or $100 for one quadcopter, Delrin parts only.
Cool! We shall encourage and support this any way possible. I think the costs will be lower than you estimate.
BTW, we have a wealth of information to post pretty soon. I've got all the drawings you'll need, professionally drawn and up to date. They're yours for whatever use, personal or commercial.
Cool! We shall encourage and support this any way possible. I think the costs will be lower than you estimate.
BTW, we have a wealth of information to post pretty soon. I've got all the drawings you'll need, professionally drawn and up to date. They're yours for whatever use, personal or commercial.
If anyone is interested, I'm still up for doing one-off parts for people with my CNC machine. A cnc will produce slightly more waste than a laser due to the thicker cutting tool, but I can also cut arbitrary depth recesses and even full 3d shapes as opposed to just straight profile cuts. For example, I've attached an image of the center plates for my quad to show the recesses that hold the square carbon tubes in place. This image is a simulation from my CNC software, which I can send as a preview if you want anything done.
Aww, come on. Our interest (and mine in creating this thread) is to see that the Propeller gets recognition in these devices. That's all we care about. Yeah, we'll sell a kit, but I'm sure others can do a better job than us anyway with the design and cost. Our interest is to see that the Propeller's unique design gets properly adopted in multi-rotor aircraft. To that end your content is appreciated on this thread, no matter how many posts you wish to make. In fact, several pages ago I retitled the thread to include JasonD in the title. That's you!
BTW, Jason - you're right down the road from us and I don't think you've visited Parallax yet. Come on over, maybe on a trip to Tahoe?
I just meant that if anyone has other bits they'd like to get made, have questions about cost, abilities, etc that aren't related to this topic that I'd rather not hijack the thread *completely*. You'll notice I didn't mind hijacking it a little. (but thanks for making me feel less guilty about it)
I have been meaning to pop in on you some day. I go to Tahoe a lot in winter, but I'm usually driving by you either at 7am or 5pm on Sunday. I know you're a devoted worker but that's pushing it. :-) I may be up that way the week before Christmas to have dinner with some friends. Will you be there the afternoon of Sat, Dec 17?
On a more topical note, I got some time in on my new quad today and have improved the stability. My PID loop values were WAY off where they needed to be. It still oscillates a little, but I'm closing in on it, and this is still using just an ITG-3200. I haven't incorporated the DCM code yet, but I hope to add that soon. I'm thinking of trying a hybrid approach where the main stability is just PIDs & gyros, but the DCM code is used to tweak the numbers when in "hands off" mode. It sounds good in my head - hopefully it works.
Will send you a PM now - I have a feeling that a meeting on the slopes could be the right idea. Since I've got FPV almost ready to fly and I have access to a unique property up here we should go out and collect some ELEV-8 video of cliffs and snow, don't you think? After all, my kind of business meeting would never occur on a golf course. We need a bit more snow to make a meeting on the slopes happen though - it's starting out as a dry winter!
Cool! We shall encourage and support this any way possible. I think the costs will be lower than you estimate.
BTW, we have a wealth of information to post pretty soon. I've got all the drawings you'll need, professionally drawn and up to date. They're yours for whatever use, personal or commercial.
Let us know if we can assist.
Ken Gracey
I wasn't able to use the Pololu quote service since I didn't have any files to submit, but I hope it will be cheaper that $100 per airframe kit. I wish Parallax would offer the hardware alone. Even just the Delrin cut would be a huge help: most people can source the rest of the parts easily enough.
My project partner (an Computer Engineer) and I (Computer Science) have decided to use a quadcopter as a base. We have 5 person quarters of work to get something running, so we'll see how it goes. Our professor wasn't very optimistic about what we could accomplish: he said that it needs to have a PID loop...
Our goal with this project is to create an "extreme sports" video platform. Since both of us are avid mountain bikers, our main use case is to do aerial videography. Most extreme sports video are from helmet cams or people standing beside the trail, such as this boring video that could be really nice if it had an aerial component. Professionals use a helicopter to get amazing footage such as this really well done video and another well done video.
Ideally, our in the stars goal is to have a platform that is able to autonomously follow a user down a trail. We live in Southern California, which is much like Utah from the video: devoid of trees, making object avoidance much easier. We hope that by using the predesigned/tested mechanical platform our project design goes more towards higher level functionality.
But that's only our goal, and we'll have to get the thing flying. I'm looking forward to being able to work on this project (especially since most of my scheduled class time is for it as well!).
Pololu is very expensive for laser cutting, but for many it is the only way.
If you need laser cutting, many sign shops or trophy shops have lasers. If you bring the material and files, especially if in Corel format, I'm sure you could get a fair deal.
If that doesn't work, I can laser cut for you for even cheaper.
@Cody, we can get you an airframe kit. We plan on offering three ways, but right out of the gate we'll just have the ELEV-8 Full Kit. When do you need this by?
I once asked my Epilog rep what most laser engraving shops charge for machine time, and he said $1.50 to $2.00. So, yes, Pololu is a bit on the high side.
I think many shops would like to charge that much. I have a feeling that those rates are on the decline, especially with the proliferation of Chinese machines.
There is a hackerspace in Seattle that charges $2/minute + materials for the laser. I've not seen it but I am told that they don't have a vent system for it - so they don't like to do a lot of cutting. They also charge $0.50/minute for the MakerBot. That would make my Space Shuttle worth $600!!!
StudentRND in Bellevue charges nothing for the MakerBot and the laser will be free as well, except for materials. Oh, and it is vented.
No vent system for the laser cutter?!! I don't know how you could even engrave without one. The fumes are not only stinky, but a health hazard and potentially combustible. Yikes!
Yeah, the MakerBot isn't quite ready for service-bureau prime-time yet. Such a device has got to be either a lot faster or a lot more precise to be any kind of profit center.
@Cody, we can get you an airframe kit. We plan on offering three ways, but right out of the gate we'll just have the ELEV-8 Full Kit. When do you need this by?
Ken Gracey
We plan on getting started developing at the beginning of January, just before CES. Just for completeness, I've found several quadcopter airframes (here, here, and here) but for the most part they are expensive and small. There seems to be a fascination with carbon fiber frames. Necessity, or simply marketing? I don't know, but buying a frame from one of those sources would cost us $200-$600. Doable, but we'd like the ability to modify as we see fit. Namely, scale the quadcopter to different sizes for different sized cameras.
As a side note, we are interested in just a frame because we want to do the electronics selection ourselves. We may also drive the few hours out to the nearest hackerspace and see if we can use that with a daypass. In any case, we'll probably make two quadcopters so that we can both develop simultaneously and have a spare if needed.
I'm looking forward to this project. I've spent four years in school just to get the background to get to this point. I'm ready!
Carbon gets used a lot because it's incredibly light for its strength, and it's very rigid. That means good flight performance, and less frame weight means more payload without requiring a bigger power setup.
Aluminum frames can be significantly heavier, and if they bend on a bad landing it can affect your subsequent flights - an off-axis motor will affect the stability quite a bit. Plastic and wood both work too, but they flex more, and can resonate with motor vibration, which can confuse gyros.
My first (decent) quad used a pair of aluminum tail booms for a model helicopter cut in half ($10 each), an X-shaped plumbing connector (to plug the four booms into), 4 T-shaped plumbing connectors to mount the motors to, and a round piece of hobby plywood which I screwed into the four arms to keep them stable. It flew just fine, so yes, a better frame will improve things, but as long as it's decently rigid and balanced it's not the most important part. If you're going to hang a $600 DSLR from it, then maybe you want to spend a little more to get something fancy.
I've been charging $60 to $75 an hour for machine time on my CNC, which is $1 - $1.25 a minute. That's cheap, but it's also gaining me experience with the machine and it's fun.
@Jason - flies as well as the Hoverfly setup. Your design appears to be working perfectly.
@Roy, I think you should be able to build an ELEV-8 on your trip to Parallax this week. At a minimum, you can get the motors and certain pieces prepared on Monday and the anodized booms come back by Tuesday/Wednesday. You should be able to go home with a complete ELEV-8. If you have an R/C transmitter please bring it.
Jason: Nice work with just the gyro. Thanks for posting.
Can you tell me what sort of results (data) do you get from the ITG3200 gyros? I have the Wii gyros (clone) which are earlier versions of dual axis gyros. I am getting approx 0,-29,-1 when stationary (so probably I am getting +/-512).
Would you mind posting your latest PID loop calcs?
BTW I have received my larger props in both CW & CCW now so I have: 8x4 and 11x4.7 and 3blade 8x4.
Here is what I am about to try regarding the calcs using your older code values (I need to recheck because I added this code a few weeks back and didn't get a chance to try it). QuadMain_004.spin
Comments
...Tiger
Good call Ken! That is also the way I have done mine. Since you would have to solder connectors anyway, may as well just solder wires and use heatshrink. Currently mine are not buried inside the tube yet.
Just ensure the connections are the correct way around for the motor rotation (2 sets are normally reversed).
Thanks ken. Mine had no force and failed just sitting there. Do your motors have threaded shafts or are you using the spinner style connectors?
http://vimeo.com/31514966
This guy makes his quad do fun things.
...Tiger
There's the type of stuff I like to see. Thanks for sharing.
Ron
You can't tell from the pictures, but the body plates have counter-sunk slots cut into them to make sure the arms are properly aligned and angled, and all the holes, including the ones for mounting the ProtoBoard, were drilled at the same time.
Jason
Nice work - I'll even change the title of this thread since you're back in the quadcopter game. I'm truly inspired by the way you've assembled the electronics. Way to go!
Ken Gacey
I like the way you kept the ESC's in the center. Should be good for balance and faster response due to little weight on the extremes.
Great work!
I'm considering doing it again but with shorter arms - They're about 13" right now, but I have enough tube left to make a set of 9's. I didn't want them too small for this one because I want to mount my GoPro on it, and want the props out of frame as much as possible.
I also still have some oscillation so I'm going to try setting up the board with an LCD display and alter the programming so I can change PID values between flights. That should help me find some better settings and hopefully get it a little more stable.
From the pictures in #317, it appears that there are the following delrin parts:
4 - Skids
8 - Motor mount pieces
2- main body plates
1 -mini body plate
Assuming that to cut 1/8th delrin takes half the time as 1/4 delrin, I estimate that two of the motor mount plates would take as long as one skid, each one of the main body plates has as much cutting as four legs (or 2 leg times), and the small plate has one leg worth of cutting. So, that is
4 - Skids (6:00)
8 - Motor mount plates (6:00)
2 - Main body plates (12:00)
1 -Mini body plate(1:30)
So, ~26minutes of cutting time. Pololu will do cuts for $2.50 per minute, so to cut one quadcopter will cost about $65.00. The delrin will cost approximately $85 for a sheet, that is enough to make four (est.) quadcopters (http://www.mcmaster.com/#delrin-(made-with-acetal-resin)/=f6brff). So, the subtotal is $65.00*4 + $85 = $345, plus 15% for shipping costs so the grand total would be ~$400, or $100 for one quadcopter, Delrin parts only.
Is anybody else interested in just the frame?
Cool! We shall encourage and support this any way possible. I think the costs will be lower than you estimate.
BTW, we have a wealth of information to post pretty soon. I've got all the drawings you'll need, professionally drawn and up to date. They're yours for whatever use, personal or commercial.
Let us know if we can assist.
Ken Gracey
Ken & Parallax... You're the best
I don't want to hijack this thread any more than this, so if you have questions you can PM me, or add them here: http://forums.parallax.com/showthread.php?132851-My-new-CNC-machine-and-why-I-ve-been-so-quiet-lately
Jason
Aww, come on. Our interest (and mine in creating this thread) is to see that the Propeller gets recognition in these devices. That's all we care about. Yeah, we'll sell a kit, but I'm sure others can do a better job than us anyway with the design and cost. Our interest is to see that the Propeller's unique design gets properly adopted in multi-rotor aircraft. To that end your content is appreciated on this thread, no matter how many posts you wish to make. In fact, several pages ago I retitled the thread to include JasonD in the title. That's you!
BTW, Jason - you're right down the road from us and I don't think you've visited Parallax yet. Come on over, maybe on a trip to Tahoe?
Ken Gracey
I have been meaning to pop in on you some day. I go to Tahoe a lot in winter, but I'm usually driving by you either at 7am or 5pm on Sunday. I know you're a devoted worker but that's pushing it. :-) I may be up that way the week before Christmas to have dinner with some friends. Will you be there the afternoon of Sat, Dec 17?
On a more topical note, I got some time in on my new quad today and have improved the stability. My PID loop values were WAY off where they needed to be. It still oscillates a little, but I'm closing in on it, and this is still using just an ITG-3200. I haven't incorporated the DCM code yet, but I hope to add that soon. I'm thinking of trying a hybrid approach where the main stability is just PIDs & gyros, but the DCM code is used to tweak the numbers when in "hands off" mode. It sounds good in my head - hopefully it works.
Ken Gracey
I wasn't able to use the Pololu quote service since I didn't have any files to submit, but I hope it will be cheaper that $100 per airframe kit. I wish Parallax would offer the hardware alone. Even just the Delrin cut would be a huge help: most people can source the rest of the parts easily enough.
My project partner (an Computer Engineer) and I (Computer Science) have decided to use a quadcopter as a base. We have 5 person quarters of work to get something running, so we'll see how it goes. Our professor wasn't very optimistic about what we could accomplish: he said that it needs to have a PID loop...
Our goal with this project is to create an "extreme sports" video platform. Since both of us are avid mountain bikers, our main use case is to do aerial videography. Most extreme sports video are from helmet cams or people standing beside the trail, such as this boring video that could be really nice if it had an aerial component. Professionals use a helicopter to get amazing footage such as this really well done video and another well done video.
Ideally, our in the stars goal is to have a platform that is able to autonomously follow a user down a trail. We live in Southern California, which is much like Utah from the video: devoid of trees, making object avoidance much easier. We hope that by using the predesigned/tested mechanical platform our project design goes more towards higher level functionality.
But that's only our goal, and we'll have to get the thing flying. I'm looking forward to being able to work on this project (especially since most of my scheduled class time is for it as well!).
If you need laser cutting, many sign shops or trophy shops have lasers. If you bring the material and files, especially if in Corel format, I'm sure you could get a fair deal.
If that doesn't work, I can laser cut for you for even cheaper.
Ken Gracey
I once asked my Epilog rep what most laser engraving shops charge for machine time, and he said $1.50 to $2.00. So, yes, Pololu is a bit on the high side.
-Phil
There is a hackerspace in Seattle that charges $2/minute + materials for the laser. I've not seen it but I am told that they don't have a vent system for it - so they don't like to do a lot of cutting. They also charge $0.50/minute for the MakerBot. That would make my Space Shuttle worth $600!!!
StudentRND in Bellevue charges nothing for the MakerBot and the laser will be free as well, except for materials. Oh, and it is vented.
Yeah, the MakerBot isn't quite ready for service-bureau prime-time yet. Such a device has got to be either a lot faster or a lot more precise to be any kind of profit center.
-Phil
We plan on getting started developing at the beginning of January, just before CES. Just for completeness, I've found several quadcopter airframes (here, here, and here) but for the most part they are expensive and small. There seems to be a fascination with carbon fiber frames. Necessity, or simply marketing? I don't know, but buying a frame from one of those sources would cost us $200-$600. Doable, but we'd like the ability to modify as we see fit. Namely, scale the quadcopter to different sizes for different sized cameras.
As a side note, we are interested in just a frame because we want to do the electronics selection ourselves. We may also drive the few hours out to the nearest hackerspace and see if we can use that with a daypass. In any case, we'll probably make two quadcopters so that we can both develop simultaneously and have a spare if needed.
I'm looking forward to this project. I've spent four years in school just to get the background to get to this point. I'm ready!
Aluminum frames can be significantly heavier, and if they bend on a bad landing it can affect your subsequent flights - an off-axis motor will affect the stability quite a bit. Plastic and wood both work too, but they flex more, and can resonate with motor vibration, which can confuse gyros.
My first (decent) quad used a pair of aluminum tail booms for a model helicopter cut in half ($10 each), an X-shaped plumbing connector (to plug the four booms into), 4 T-shaped plumbing connectors to mount the motors to, and a round piece of hobby plywood which I screwed into the four arms to keep them stable. It flew just fine, so yes, a better frame will improve things, but as long as it's decently rigid and balanced it's not the most important part. If you're going to hang a $600 DSLR from it, then maybe you want to spend a little more to get something fancy.
I've been charging $60 to $75 an hour for machine time on my CNC, which is $1 - $1.25 a minute. That's cheap, but it's also gaining me experience with the machine and it's fun.
[video=youtube_share;QNqwCrEYNwU]
@Roy, I think you should be able to build an ELEV-8 on your trip to Parallax this week. At a minimum, you can get the motors and certain pieces prepared on Monday and the anodized booms come back by Tuesday/Wednesday. You should be able to go home with a complete ELEV-8. If you have an R/C transmitter please bring it.
Ken
Can you tell me what sort of results (data) do you get from the ITG3200 gyros? I have the Wii gyros (clone) which are earlier versions of dual axis gyros. I am getting approx 0,-29,-1 when stationary (so probably I am getting +/-512).
Would you mind posting your latest PID loop calcs?
BTW I have received my larger props in both CW & CCW now so I have: 8x4 and 11x4.7 and 3blade 8x4.
Here is what I am about to try regarding the calcs using your older code values (I need to recheck because I added this code a few weeks back and didn't get a chance to try it).
QuadMain_004.spin