Arduino active rocket stabilization system update mid 2016
HTML-код
- Опубликовано: 20 авг 2024
- A long overdue video update, lost this file on my storage array for a while. Stabilization system is now a completed stack and fully bolted together but this design has been paused while I explore another method of fin motor driving, encodered motors and bevel gears to transfer the power to fin drive shaft.
More detailed project log with tons of pictures is on my blog at the following link
tminusarduino.b...
I am alive and working on stuff! New video is currently being shot and edited. Work on actual spacecraft and science support equipment at a 4 letter space agency has been eating my time and wearing me out.
Arsenio Dev this is dope
I know this comment might be a bit out of date for the video, but if you're still looking for cheap DC motors, you can get old printers for just a couple bucks a piece from places like scrap exchanges, goodwill, etc. Printers are full of tiny motors and some have rather large stepper motors too. Also, printers usually have dozens, if not hundreds, of little plastic gears. I've found a few plastic bevel gears in printers before. Good luck!
Are you not able to ram the motors in one direction, then detect when they stop moving using the encoders and use that to zero them on startup? This way you can remove the need for 4x potentiometers and 4x bevel gears...
That idea makes me shiver, though ...
Problem arises when I need matched high torque low voltage motors, not really available in most printers
Just a quick note to say how much I appreciate your taking the time to list what does NOT work! That's going to save a lot of people a lot of time.
No problem, this is exactly why I log my work!
Came upon this Video by mistake but I had to watch it till the end,I hope I get to see it to the finish and see it fly,Good Luck!
To save some weight, consider three fins instead of four. More complex software for controlling but less weight. In fact, you could even go down to just two fins if, via software, you can separate the normally composite movements into two sequential moves, one for rotating, one for deflecting.
I'm going for dead simplicity, four fins is easiest since
Servo write val
Servo write val-180
Had considered it but mothballed it due to being lost how to map XY onto 3 actuators
Or replace the metal screws on the arduino board with something lightweight. Antenna looks bulky too
this is just the testing mockup, not really concerned with weight or durability yet
I'm thinking you should definitely set up a wind tunnel to test out your designs. That way you can see whether the fins and controllers can effectively counter any forces pushing it "off-course". It will be very hard to test that on an actual launched rocket, they simply go too fast. Wind tunnel first to verify, then actual rockets later.
I wish I were close enough to you to actually help out. Looks like a fun project.
I've been watching some RUclips videos of model rocket launches. And what strikes me are the smoke trails that are left behind which are initially pretty straight but which very quickly end up all kinked. It shows how the wind speed and direction varies considerably on the way up.
I'm thinking you're going to have considerable difficulty getting the fins to respond fast enough in order to compensate for any deviations from the intended path. I'm thinking on the order of milliseconds. While you might well be able to do the computer processing quick enough, getting the motors that drive the fins to operate fast enough may well be the main challenge.
Used this at a manufacturing company I worked back in the day. At start up, Move the motor (in steps) until it stalls and then you know the position of the fin. Store value and return to neutral.
huh, interesting. Bet that works far better on steppers because I know Trinamic drivers can do "virtual" endstops to zero out based on stall out
really impressive
3:16 in my experience the servo splines don't really matter too much as long as you can press-fit the component onto the servo, then fasten it to the pinion with a screw(typically included with the servo). I recently had to do this for a 3d printed servo stabilized gimbal, and realized the friction fit holds pretty well. I've had more of an issue with the splines stripping from components once i figured out how to print them properly, and realized splines don't matter too much as long as you can press-fit then screw in. Screwing your component to the pinion also reduces the risk of stripping, and still keeps your component from sliding around even after it's stripped. Then again, screwing the fins to the servo horn like you did also works, with the caveat of slightly more weight.
5:51 brilliant
Man that's one seriously awesome project! Hope to do something similar one day. Beautifully done.
Why did google kill your adsense? Thats crazy
I honestly have no idea :( Hurts real bad though as I haven't made a single dime off of thse videos
There's a bunch of new development that hasn't been in videos due to bad internet, but check my twitter and see some of the latest designs
Man that's a bummer I'm sorry to hear that. They've just been going on a demonetization kick with all my videos as of late so I feel your pain. YT and Google are no longer the good guys they once were. I will check out your twitter; same name I assume?
Yeah, its super disheartening they're murdering educational content, especially chemistry aligned videos like yours. Yeah same name
600 is nothing ,DO IT! I spent 12 grand on a rocket shuttle Combo that went sub orbital ,the rocket crashed 10 miles away but I was finally able to glide the shuttle into my area ,but couldn't land it and smashed into someone's tent! I loved it! And now I'm putting the finishing touches on a Saturn V that is several thousand,but if you like it,you do it!
I'm going for L1 this summer, have a bit more disposable income now so I will be flying soon
I'm wondering about your choice of putting the control fins at the front of the rocket. Even very tiny angles of those fins will have a huge impact on the trajectory of the rocket. I suspect it's best to control the regular fins at the bottom of the rocket. Now, I know there's probably no place in the bottom of the rocket to put the servos, and you want them as far forward as possible, anyway, for center-of-gravity considerations. So how about putting the servos at the top and connect them to the fins at the bottom with control/pusher rods, much like is done in model airplanes?
lets hope spaceX gives you a job
Hahaha I'm not up to their standards yet
*north korea
If I were you I would apply. I am an engineer myself and I would consider you a good candidate for guidance and control position. This is currently a hot area due to popularity of drones around the world. Your passion and attitude is great. Not many working engineers would bother doing this as a hobby and this alone puts you well ahead of many other candidates. There maybe quite a difference between what you think about you and what a recruiter sees from a candidate. Let them know your RUclips channel. They will be impressed. You will learn A LOT once you join a company. Once you join a company like Space X, you should get paid well enough that spending a several hundred dollars for your hobby should be easy. Hope your qualification meets their requirement. If not, work for it!
@@grasuh
That may all be true but he has to at least have some academic qualifications to show, you can't just go with a project otherwise it would be simple to get a job ;D.
Why get a real job when you could just crank out MANPADSs out of your garage?
Tower Hobbies is a great source of RC stuff including servos, gears, electronics.
Have not read all the comments, and realize this is over a year old, but I would suggest doing a search on "associated Part #: 3914." It's basically the differential ring and pinion for an Associated TC4 touring car. Don't know if it would offer the precision needed but a stop at the local hobby shop may provide other options (the metal gearing from an RC8 comes to mind) that would be from the RC racing world.
Snagged some really nice purpose made steel bevel gears from the far east, video update coming soon
I would suggest that x function together similar to the rudder of an aircraft ...... y axis function inverted to each other to rotate the body ..... you will then have an top, bottom on your x and the aileron action/result on your y axis ..... That way you will only require two drivers/servos/motors ......
Put the servos in the fins and attach the output shaft to the body. servo in the fin, a metal tube attached to the motor and attached to rocket body run wires through the tube.
Unfortunately that would be even worse than the newer designs which are MUCH easier to integrate and have fewer failure modes
Rather than a potentiometer and another bevel gear, why not add an optical sensor that can detect the edge of the fin then home the fins on initialization.
This would:
1) Be cheaper
2) Be more accurate (since you would home to the actual physical fin edges)
3) Look really cool when it wags the fins to home.
Hmmm that might be a good idea but with a magnet and a hall effect sensor so it's flush with the body
Great stuff!!
This was something I thought about a great deal when I played with rockets... But we didn't have computers small enough back then and it's illegal in England anyway. :(
To solve the problem of space inside the rocket frame (around time frame 6) why don't you use a light plastic disk with its center connected with the rotating fin part and then use the servo along with a light plastic arm to rotate the fin! I think this will make your life easier.
Save money, go yo 3 fins. Also suggest you mount the fins off centre so the motor always carries a torque. This will reduce fin flutter caused by slack in the drive chain.
problem is that the math for 3 actuators gets painful real quick
Hey, I design opto mechanical stuff, I think your fins would be way more controllable with a flexure design, a fast servo and a reducer gear. You want your fin to be a flexure rather than a fin so that you reduce flutter. Also I imagine that you mostly want small ~5° rapid movement, not super large 90° movement. Flexure s can be as stiff or as flexible as your experiments decide.
Yes, I think flutter will be a major problem with direct-drive fins->servos.
Think about the sorts of cyclical loads you'd get with 20hz flutter. Many dead servos. Fins ripped clean off.
Props to you on your book sense. You have a gang of that. Some of the problems you encountered require common sense, you were lacking there. Awesome work, wish that had all that Arduino shit when we were kids.
Hahahah I know, much of this is learned through mucking about combined with what I learn from books and other research
this simply not going to work beacuse if its going about the speed of mach 1,the drag will strip off that little wings since it is directly connected to servos and dont have additional support may leads to losing the stability of rocket
i have no idea how i found this vid but to your design:
i think its a bad idea to mount the fins on top of the servos for 3 reasons
1st you loose resolution: you proably dont want to turn it 180 deg during flight maybe +-30° should be more than enough
but this 30° are restrickted to 600 the 2000 microsec of the arduino.writeMicroseconds() function
2nd force: you need a more beefy servo to mvoe the fins if you drive the fin directly
3rd mounting: as you explained
i would suggest to put the fins on a rod which get driven by a rod, so you can use any connection you like. please refer to this sketch
oi64.tinypic.com/2jcfvwj.jpg
Absolutly agree
I've got a software limiter on the servo range of motion to 30deg for demonstration, scaling back to +-15 for flight version. Currently exploring several drive options and testing them in a crap wind tunnel (multiple leaf blowers onto one airstream) to verify control authority
Sketch does not work, please use Imgur
imgur.com/a/r06je
Tacker Massacker Ok, that's interesting. Will run a prototype later and see how well it works under test
I realize that this would likely end up in needing completely new body and design, you might want to try using thrust gimbal as you could use 2 servos, although they might need to be heavier. I may try working on something similar but my printer is 150 x 150 x150😀
Hey! I'm actually very good friends with someone who is doing exactly that! Check out Barnard Propulsion Systems on here and twitter. I lent my voice to his channel trailer and we talk on the technical aspects all the time.
ruclips.net/video/fJji2f8LmnU/видео.html
Thanks
THANK GOD FOR 3D PRINTING AT HOME. THANK YOU FOR SHARING
ya you should definitely make it, if any other problems than quite.
vibration must be taken into account. vibration strongly affects the performance of the gyroscope.
Yep, on flight version the computer or IMU will be on vibration isolators
I actualy intrested to make the project but sadly i bought 2 mpu6050 they dint work, one mpu6050 dosent want to respond with the code, the second one it respond with the code but when i roll or spin it, it dosent wanna show
I wonder if rollerons like in AIM9's before the X version would work without too much guidance from the arduino.
no idea but making them has been added to my to-do
Arsenio Dev ok. tnx for the informative video, this will help in my robot arm project.
cheers
I don't see your problem . If you want to fix control fins on top of servos . Why don't you use self taping screws from the underside of the control horn into the base of the fin ?
Why not a spiral coil gear and small toothed gear to get your 90 degrees turn? With the gear reduction ratio you would get a much better control over your fins. I doubt you would get even close to needing to do a full revolution of the spiral coil, since after all, this is a rocket fin.
Why not use an rc flight control board for stabilization? That's what I did on my underwater rov and it works great. Regarding the fin stabilization, an exhaust deflector would seem to be an easier and more reliable solution.
4 separate servos is bad design. they operate in tandem, so use just one for each pair.
the hinge design is also bad. instead, put all 4 fins on a ball in a socket = one moving part.
Roll stabilization?
Psalm 133:1 How good and pleasant it is when brothers live together in unity!
I think it would be better to mount the fins centered on a metal axis so they can rotate. Then move the tip of the fin by a servo. Simpler, uses larger resolution, and leverage in your favour.
Sort of close, new video detailing design decisions coming soon
Try using gyros that are controlled by an RC helicopter auto hover radio system. The gyro unit would need two motors with counter rotating flywheels mounted in a internal frame that is attached to the servos.
just an idea, cool stuff.
Problem is reaction wheels don't really work too well in atmo due to aerodynamic loads
Would be less drag if you vectored the thrust instead of using the fins. Leaves the door open for spacex type landings too
My counterpart and friend Joe Barnard at BPS.Space is already doing that
Looks to me you might want to look at small hobby lathes, mills, and make some gear hobbing attachments for them. Since you can handle the electronics, you'll want to CNC them.
I really want a lathe or horizantal mill but they're PRICEY
If you mean a CNC manufacturing center, yes. You want one. :) But there are good deals out there on Bridgeport hand crank machines, and old SouthBend 9" lathes. They're good machines to start with, but are big. Looking at your shop the Tiag, or Sherline series might be useful for you, but big is always better. Those are very small machines.
and is it possible to program so that the missile hits the target?
I feel like this is the biggest conflict with 3D printing at the moment, resolution. You need too much to get it precise.
I cnc everything. You get it in any material and it has an ultra-precise finish.
O bollocs, such a shame you cant share the code. Well... I guess ill have to write my own then and find some smaller motors or a rocket license (no not a us one). Good luck with this project, its promising!
If I were you I wouldn't use resistor rather I will use encoder for pointing the fins to the exact location and for obvious reason.
why not mounting the fins with a bearing and coupling it flexibly to the servo so you don't put load on the servo?
Already implemented on new version
Can you make a rocket deploy wings and glide back down
It's possible but I wouldn't, gets the feds antsy
I have a question for you❗ that which metal is used to make a outer skin of rocket missile.
The idea of controlling the rocket's flight via movable fins is totally cool. The practicality? I'm not so sure. For a typical model rocket you really only have a very limited amount of time to perform any flight controls. In fact it's mostly limited to the powered flight time. Have you done any calculations relating controllable flight times to speed of moving the controllable surfaces?
Is this for normal powered rockets or high powered rockets? For the latter you MIGHT have enough control time. But then, what are you trying to control, keeping it on a straight-up flight or doing something more complex, like aiming for some target destination?
Have you done any actual flights?
You could solve this a lot easier with a bushing and mechanical linkage
hi everyone, i am trying to stabilize my rocket to make it fly as perpendicular as possible, i bought an mpu6050 and i used a kalman filter to combine the angular position obtained from the gyroscope with that obtained from the accelerometer, i have not yet had a way to test the rocket, however I noticed that as soon as I shake theboara little the roll and pitch values go up, someone has some advice.
any comments will be valuable to me, thanks in advance
Looks very nice! Inspiring others :D
what is the airfoil type of your control fins
Try u-joints with bearings. U-joint can be 3D printed.
Which hardware are used in rocket system?
Awesome video. Have you considered returning to the 5g servo motors and designing the fins with two points of contact; a post and hole on the top, and the servo horn mounted on the bottom. This should defer the majority of the stress on the fin during flight away from the servo and onto the post and hole joint, while greatly reducing the torque required by the servo. Another benefit, if a servo fails/looses power mid flight, the fin naturally return to an inline position. Servos generally have ~180 degrees of motion, and you probably only need +/- 20 degrees of fin deflectionfor this application. Should work!
Post & Hole --> __ |\
| \
| \
| \
| \
_ | \
Servo Horn --> | | | \
| | \
| | \
|_|___________\
I thought about it, but decided to keep on going to the micro gearmotors especially since Pololu donated a full set
So cool man! post an update, this stuff is too cool.
New one is coming soon, this time detailing the new motor drive system and going in-depth on the control code for that since at least THAT part will be open source and freely available
I am watching from nepal
If rocket speed is Mach 2 or above will Arduino manage? or you have to use CPLD or FPGA ?
Keep up the good work!!
Why four controlling fins? Why not just two? I haven't reviewed your goal but if it's just stabilization two will make the correction. Any sustained induced drag from fin angle difference between fixed and moving would be made up with performance from a lighter air frame. Three fins using a delta PIO configuration would be a challenge too. I'll sub ya to see where this goes. Looks like a fun project.
I know this is an older video but, why not make a reaction wheel configuration over moving fins? I know that the fins look cool but that is a lot of stress to attempt to handle and reaction wheels are used by full sized rockets for their reliability under the stresses exerted.
Reaction wheels will not have enough force to counteract the loads on the craft. Actually the full size rockets use gimballed rockets and differential thrust to do the rest
Arsenio Dev fair point.
What about gimbals?
Heeeey I like it, I am glad I saw this
Amazing ❤️ Thankyou 🍸🍸🍸 Could you share Arduino Sketch.
shorten your wiring, I know it doesn't sound like much, but a pin kit isn't more than $10 and it will save weight and wires going everywhere.... just a thought...
Yep, I'm getting custom harnesses made for the new version
could you incorporate GPS to make this a guided missile?
That is why he can't share the code
Perfect for boogaloo barn burner banquet.
use the bevel gears from 1-10 nitro car diffs or 1-16scale
I looked into those but they don't have the 3mm bore or a setscrew to lock it on the shaft. Found a solution though, details coming sooooon
I love this kind of hobby, where are you located? I am near GTA Toronto area and have lots of ideas to make things happen
Would love to see where you ended up on this design.
Coming SOON!
Great work. would me mind to share some cad dimensions?
look into a Teensy board to replace the arduino
what software application do you use to remote that
WORMGEARS? INSTEAD OF BEVEL
Where would one get their hands on servos like that?
Amazon, eBay. These are standard micros
Sick I wish I had a place or money to do it, ironically I got the brains for it and that's it
Total costs for this were about $45 in parts, most expensive was the TechShop membership but I sold custom laser etchings to offset. Look around for a local hackerspace, they'll help you out with 3D printing, code and electrics. For parts, eBay is your friend!
The major problem is finding an area that's not suburbia and basically there is no where that isn't ether forest or urban development
Google earth is your friend! Join the local NAR/Tripoli chapter?
Arsenio Dev I'm to young to do stuff alone and my parents are working all the time and I don't want to make them drive me at like 8PM
Ahh, get some friends to drive you around
Looks very interesting.
Hi,
I am currently working on a similar project, and I am looking to have the rocket's fins turn once hitting a certain altitude. I am using a BMP180, a barometric sensor, to calculate the altitude. However, the barometric sensor uses a lot of delays in order to take measurements, which interferes with the stabilization system, so the fins are not able to spin as fast. I have tried using schedulers, but I am relatively new to Arduino and could not get it to work. I also tried using timers rather than delays, but I think there are some delays built in the BMP180 library that also interferes. Do you have any ideas to fix this problem? If not, do you know any other barometric sensors that do not have as much delay, or have another solution to stopping the rocket at a certain altitude? Thanks a lot!
Oof, you're getting into ants in the fed pants territory there. Try switching those delay() calls in the library to millis() or just go baremetal I2C commands.
the question is , can this things fly ?
Technically yes, but I don't want to fly this one.
I think worm gear would be better than bevel gear. And there is a lot of them on aliexpress
working on a variant that uses those too
can you not just 3d print the servo gears that would be nice...
Would lego technics bevel gears/axles work instead of the 3d printed ones?
That would be a solid no, they're too bulky and not strong enough
Cab I use this mechanism to stabilize the rocket while falling down?
I need a 90 degrees angle.
I mean technically yes, but I don't want to make a JDAM
@@ArsenioDev Hah, me neither. I want this angle for a drop probe.
SpinOX, drop probe? I'm interested
@@ArsenioDev, I wanted to make a small model of the drop pod from HALO, but with a few sensors on board.
SpinOX hmmmm, why not static stabilization? Just add fins and/or strakes and call it a day? Then have an altimiter pop a chute at altitude
Can we read your arduino code?
Great project
Can it be automated?
Can arduino do tvc ?
What CAD software are u using?
Autodesk Inventor
Stl ?
did you print it in vase mode the nose cone?
It's called an ogive ( not sure of the spelling ) shaped nosecone. It's made for intermediate speeds rockets. Rounded nose for slow speed and spike for supersonic.
Great vid.
parts list pls
Can it fly to the designated location?
this looks more complex than just developing electronics and stabillization. the problem is the propellant.
Put on a APM ，let it fly to the destination route set automatically, fuel only need to adhere to 1 minutes, the purpose is to make it more like a missile!
apm???
No, I do not need the feds showing up.
just do not send them invite to your test flights..
Hi! Can you share Arduino sketch for this project?
As said before, no code or CAD will be shared due to it possibly being under ITAR arms restrictions and me playing it safe. To be fair, it's hilariously simple if you really think about it
Neat. Is this legal? Don't GPS chips shut down if they think they're in something like this?
This is 100% legal. No, the COCOM only kicks in at >1,000kts @ 18,000m. Further this variant does not pack a GPS unit
stl file?????
Nope
#JESUSCHRIST #love #AWESOMEGOD #unity #kindness #charity #children #JESUSCHRIST #GODOFLOVE #GOD #childbirth #child #son #daughter
do you have 3d file of fins ?
No
So I'm not the only one that, when he tries a lot of things that fails until it works, needs to tell to everyone how I failed until the final result !
Yep I've been massively helped by other people's work and their failures being told plainly so I'm doing the same
Why not use pro mini?
I'm going to be using a custom board and am ditching the AVR arch in favor of the ARM
no more updates?
Working on shooting a full update video, spoiler: it shrunk
Why use the complex gearing, driving four fins, and really messing up your balance point as well as with the center of thrust, when you can simply move to a thrust vectoring setup, by attaching the motor mount to a simple two servo gimbals using light weight push rods, or you can get even smaller, by building a gimbal for a rocket nozzle, positioned over a solid rocket motor (not V-2 style), but potentially a bell style, or better yet, the Toroidal aerospike nozzle tinyurl.com/p4w4ad6 with the spike gimbaled, which is likely a hybrid. I'm sure Barnard Propulsion Systems (BPS) channel is gimbaling the motor, and he is flying without fins at all as can be seen here tinyurl.com/nxxqto9 but, his goal is to land under thrust much like Xspace does. and he just got started with his program. but because at some point he will also need to throttle, a modified Toroidal aerospike nozzle is likely what he will need in the future, so he can gimbal his motors, as well as to add some thrust control, and vent excess thrust to the sides and slightly down when closing down the thrust to the main nozzles to land upright. only time will tell. try contacting BPS to see how he gimbaled his engines..
Hahahaha speaking of BPS, Joe and I are good friends! I wanted to do fins since it was a good challenge and also is the groundwork for an academic paper I'm considering writing for a blackbox (hush hush) project.
Since the Chinese Government is trying to grow their businesses there, they subsidize the shipping. Your seller can probably ship those items to you for less than I could if I had them from within the U.S. The result is likely that the seller is counting on the shipping costs as a part of their income for the parts.
Enjoy watching your video blog of your project and your explanations of what did not work as well as what did. Thanks for posting.
Whats use the materials
Please clarify the question