Actively Stabilised TVC Model Rocket

แชร์
ฝัง
  • เผยแพร่เมื่อ 3 ม.ค. 2025

ความคิดเห็น • 206

  • @Assassinboy-oq7vf
    @Assassinboy-oq7vf 3 ปีที่แล้ว +76

    honestly this is impressive

  • @Bob_too
    @Bob_too 3 ปีที่แล้ว +43

    Amazing achievement Jacob and perfectly narrated explanation of the design and control systems. Great critique of the control system performance, sw simulations and next stage improvements.
    I’m looking forward to the next episode !!!

  • @Hermis14
    @Hermis14 3 ปีที่แล้ว +3

    As an aerospace engineer, I am satisfied with this video. Keep up the good work!

    • @Tom06
      @Tom06 3 ปีที่แล้ว +1

      Same

  • @STRIKEcorperation
    @STRIKEcorperation 3 ปีที่แล้ว +24

    In addition to the delay between the servo recieving a position command and actually moving there, there is an additional delay between when your PID loop calculates an output and the Servo library produces the corresponding output. The Servo library works at ~50Hz, meaning that roughly half of your calculated output values are discarded. You can actually take advantage of the fact that the servo library runs off of the timer interrupts. Every time the output compare interrupt for the timer-counter occurs, it runs an interrupt service routine (ISR) that does some stuff. By adding a single line to that code, you can indicate in a volatile global variable that the PID loop should run once. This ensures that the PID loop will always have the exact same delay every time instead of a random amount between 0-20ms.
    If you measure this delay, you can also add it to your HIL simulation so that you are taking account of the majority of temporal disturbances.
    Hope this helps you out in the future!

  • @MrFriskyWhiskey
    @MrFriskyWhiskey 3 ปีที่แล้ว +3

    Your first and only video so far. Great content. Articulate, simple and effective. Subbed.

  • @nicholasn.2883
    @nicholasn.2883 3 ปีที่แล้ว +9

    This was cool
    And I like how this was done with really basic components. I might try something similar one day

  • @AnthonyHandcock
    @AnthonyHandcock 3 ปีที่แล้ว +1

    I know near to nothing about model rocketry and watch these vids mostly because "Ooh! Sparkly thing!" but that's very impressive. Clearly your Arduino programming skills are a bit better than my level of screaming at it and calling it a bastard because it keeps insisting on doing what I tell it to do rather than what I want it to do.
    Well done.

  • @jackriminton7959
    @jackriminton7959 3 ปีที่แล้ว +1

    This is marvelous, please post more!

  • @phoenixmatrix5688
    @phoenixmatrix5688 3 ปีที่แล้ว +1

    That's the best rocket project i have ever seen in youtube

  • @smockett4924
    @smockett4924 3 ปีที่แล้ว +2

    lovely video jacob. pretty pog ngl

  • @jameshughes3014
    @jameshughes3014 3 ปีที่แล้ว

    congratulations on your success so far, this is really amazingly well done

  • @kazi_rageeb
    @kazi_rageeb 3 ปีที่แล้ว +1

    very cool the design and control systems and wonderful video editing.

  • @winter-survivor
    @winter-survivor 3 ปีที่แล้ว +1

    Wow. Looking forward to see more upgrades of your mini Starship. Awesome.

  • @brianblasius
    @brianblasius 3 ปีที่แล้ว +2

    That's a cool design. You deserve more subscribers..

  • @teslababbage
    @teslababbage 3 ปีที่แล้ว

    Wonderful video Jacob! I can't wait for an update on how your research is progressing ...

  • @glutenfreegam3r177
    @glutenfreegam3r177 3 ปีที่แล้ว +1

    When I was a kid building rockets..I was just happy when the parachute deployed properly and if it didn't land in the trees, that was a bonus!

  • @dariodalcin5177
    @dariodalcin5177 3 ปีที่แล้ว

    I love how small it is, nice job man

  • @elmariachi5133
    @elmariachi5133 3 ปีที่แล้ว +3

    It's interesting to see the corrections the servos perform, because of the slow controller speed. Maybe this would be a nice use for slightly raising the controller's clock? We could directly see the improvement.

  • @medienmond
    @medienmond 3 ปีที่แล้ว +4

    Make this open source and you will have a bright future ahead of you...

  • @senorjp21
    @senorjp21 3 ปีที่แล้ว

    This is a sophisticated project! Excellent work

  • @TZeroSystems
    @TZeroSystems 2 ปีที่แล้ว

    Excellent build and walk through!

    • @jacobthornhill
      @jacobthornhill  2 ปีที่แล้ว

      Thanks Rob! More videos coming (hopefully) soon

  • @CommonApathy
    @CommonApathy 3 ปีที่แล้ว

    Damn man, I'm glad youtube recommended me this, brilliant job!

  • @MikeTaylor-tw5wb
    @MikeTaylor-tw5wb ปีที่แล้ว

    This is very impressive. I really want to see where you take this. Thanks for sharing

  • @Shapshapper
    @Shapshapper 3 ปีที่แล้ว

    Can't wait for more in this series!

  • @djrocketry
    @djrocketry 3 ปีที่แล้ว +8

    That's really cool. You definitely earned a sub and a like!

  • @Leo99929
    @Leo99929 3 ปีที่แล้ว

    BPS.Space struggled with this. This is an achievement to be proud of! Well done Jacob.

    • @floodgates6210
      @floodgates6210 3 ปีที่แล้ว +1

      He doesn’t struggle anymore bps space is amazing

  • @ilSySTeMli
    @ilSySTeMli 3 ปีที่แล้ว

    Nice work , hope to see more in the future.

  • @Drew_TheRoadLessTraveled
    @Drew_TheRoadLessTraveled 3 ปีที่แล้ว

    Wow. very impressive, I thought I was doing well to program P.I.D. on a KK2 quad control board. You my friend take it to a new level of complexity.

  • @andrewweishaar2150
    @andrewweishaar2150 3 ปีที่แล้ว

    Your TH-cam will explode if you keep doing cool stuff like this.

  • @ahmethamdicelik1277
    @ahmethamdicelik1277 3 ปีที่แล้ว

    this is great and very inspring, thanks for sharing!

  • @marcellammerse7519
    @marcellammerse7519 3 ปีที่แล้ว

    Impressive and very well explained! This is really inspiring and makes me want to build rockets too. Would love to see more of your videos Jacob!

  • @rachitbundiwal3355
    @rachitbundiwal3355 3 ปีที่แล้ว +1

    Really Impressive and well explained 👏👏

  • @adeeperbluegreen
    @adeeperbluegreen 3 ปีที่แล้ว

    Much better than that BPS channel. Also much more realistic objectives. Keep up the good work .

  • @CottonTailJoe
    @CottonTailJoe 3 ปีที่แล้ว

    Well don’t stop man keep up the good work!

  • @takashitamagawa5881
    @takashitamagawa5881 3 ปีที่แล้ว

    This is a great achievement, congratulations. I've been waiting to see a video of someone succeed with this in model rocketry. Considering how high a thrust to weight ratio there is with these solid model rocket motors, getting stabilization through thrust gimbaling is not an easy feat. The high thrust is typically needed to quickly get directional control with fins after launch.
    It would be great to see this concept applied to scale model rocketry for models of the Atlas, Falcon 9, Gemini-Titan, and multi-stage vehicles such as the Saturn rockets.

    • @jacobthornhill
      @jacobthornhill  3 ปีที่แล้ว

      Thanks! High thrust to ratio was a challenge, but it was somewhat unavoidable due to the limited selection of motors available in the UK. I’ve recently discovered the Klima D-3 motor (high impulse and low thrust), so I will probably use those in future rockets.

    • @mk6315
      @mk6315 2 ปีที่แล้ว

      Multi stage thrust vectoring makes my brain all hurty

  • @alexwang007
    @alexwang007 3 ปีที่แล้ว +7

    The servo has a huge lag overall, and if you tried properly optimizing PID parameters you'd know that even the slightest amount of delay in your loop will heavily impact your loop bandwidth. (Will way more easily cause underdamped oscillations and could even drive it to become unstable, forcing you to drop the P value)
    Your curve already seems smooth, but a general practice is to put a lowpass filter after the derivative term because they expose noise. A simple and cheap implementation is a digital infinite impulse response filter: out = out * 0.95 + in * 0.05; and put this the control program loop. This difference equation is already normalized, just adjust the proportion of the weightings to change the respose of the filter.
    I am VERY interested in your project, if there are things I could contribute please let me know. I could provide my two cents on control stuff, actuator stuff, and sensors if you'd like!

    • @jacobthornhill
      @jacobthornhill  3 ปีที่แล้ว +1

      Thanks! This project was the first time I had attempted to implement any sort of control system, and I was learning most of it as I went along. I didn’t even know what a PID controller was before starting, so my response to the un-dampened oscillations I encountered in my simulations was to simply reduce the derivative and proportional gains (far from ideal as you mention). Your low pass filter suggestion is a great idea, and I will implement it in future projects along with other concepts that I was unfamiliar with when developing this rocket. I am currently working on a quadcopter, which I am developing all software for, including autonomous features which require a lot of sensor fusion and more advanced control systems. I am currently trying to learn how Kalman filters work, so I can implement one to fuse GPS and accelerometer data to provide position and velocity estimates. If there is anything you can offer to help me grasp and implement this, I would greatly appreciate any help. I am struggling to wrap my head around that maths needed to code this. Thanks in advance!

    • @alexwang007
      @alexwang007 3 ปีที่แล้ว +1

      @@jacobthornhill ooooooooooo respects for attempting kalman filtering!
      If that was your first time doing a control system then WOW this is beyond impressive.
      For kalman filtering I think you need a decently accurate model of the system you are trying to simulate to be able to make good predictions, MATLAB is *fantastic* for this or any sort of control systems, including advanced stuff like linear quadratic regulators, do you have access for it? The low pass filter is an infinite impulse response filter, having the advantage of having very low resource comsumption (just one extra variable to achieve very strong low pass filtering), and you can fully characterize these digital filters using the Z transform, and get frequency responses right away. I could difinitely help with these more basic math stuff; for the kalman filtering, it is also something I am greatly interested in, especially when you can double integrated the accelerometer output and correct drifts with less frequent GPS updates, it is awesome.
      You could add me on discord, or which ever you prefer!

    • @alexwang007
      @alexwang007 3 ปีที่แล้ว

      @@jacobthornhill AlexAlexAlexAlexAlex#4507

    • @Unmannedair
      @Unmannedair 3 ปีที่แล้ว +1

      I second the Matlab part and would add that Matlab has support for hardware in the loop and Arduino. I would also recommend upgrading to a teensey since it is a full 32bit controller. It will give you much higher control resolution.

    • @alexwang007
      @alexwang007 3 ปีที่แล้ว

      @@Unmannedair absolutely agreed, and simulink is great. also tensy 4.1 is brutally fast, 600MHz vs 16MHz. you could also get the arduino portenta which is powered by a STM32H747

  • @DangerousStuffEJP
    @DangerousStuffEJP 3 ปีที่แล้ว

    Love the vids, and also I think I'm your 1000th subscriber

  • @edabean007
    @edabean007 3 ปีที่แล้ว +4

    With content like this, its not long before the algorithm hits 😊

  • @fuzzyboi1721
    @fuzzyboi1721 3 ปีที่แล้ว

    Pretty cool! I hope to see your next video soon!

  • @EgonSorensen
    @EgonSorensen 3 ปีที่แล้ว

    2:59 - Abort function
    - with the engine firing full throttle, I would imagine this would create some potentially dangerous situations if the rocket starts oscillating around itself horizontally near the ground.
    Great build, even better video and explanation!
    Edit - The parachute is attached to the top, so the center of gravity will change unpredictably if/when deployed with a firing engine.

    • @jacobthornhill
      @jacobthornhill  3 ปีที่แล้ว +1

      The centre of gravity would remain the same, but the centre of pressure would change, making the rocket aerodynamically unstable. It would go out of control, you’re right, but it’s velocity would be limited by the chutes. This is better than a scenario where no chutes are deployed as the rocket can dangerously fly off fast in any direction. Worth noting that this is hypothetical, I have not tested the abort function in flight.

    • @EgonSorensen
      @EgonSorensen 3 ปีที่แล้ว +2

      @@jacobthornhill - sorry, you're right - Centre of gravity should be the same, it is pressure I was thinking about.
      I imagine the parachute also would be prone to be set on fire, as it is being unfolded. (I live in Denmark, buying rocket motors isn't something I can do in the local hobby store - so no testing for me)
      Fun to watch on TH-cam, and quite educational too :ø)
      - hope you'll never have to test the abort feature LOL

    • @jacobthornhill
      @jacobthornhill  3 ปีที่แล้ว +1

      Glad you enjoyed!
      You’re right - the parachute would likely meet its demise. It’s not an ideal system but I figured I might as well implement it for human safety (not that that is a priority lol)

  • @siamparagon9250
    @siamparagon9250 3 ปีที่แล้ว

    Fantastic! Wishes all the way from Bangkok, Siam

  • @Myname-il9vd
    @Myname-il9vd 3 ปีที่แล้ว

    I have a feeling I’ve stumbled upon a gem on TH-cam.

  • @EricMBlog
    @EricMBlog 3 ปีที่แล้ว +1

    I'm would be bet you know this, or somebody else mentioned it already, but you can get motors without an ejection charge, so you don't get that explosion at the end. They have a P at the end of their model number for 'Plugged'.

  • @MartinBogomolni
    @MartinBogomolni 3 ปีที่แล้ว

    Well done Jacob!

  • @skysurferuk
    @skysurferuk 3 ปีที่แล้ว

    Looking forward to the update...

  • @JohnTarbox
    @JohnTarbox 3 ปีที่แล้ว

    I would love to see an updated video.

  • @TomTrones
    @TomTrones 3 ปีที่แล้ว

    Impressive build!

  • @beanieteamie7435
    @beanieteamie7435 3 ปีที่แล้ว

    Can't wait to see more.

  • @cerealru1069
    @cerealru1069 3 ปีที่แล้ว

    Absolutely brilliant

  • @robertbrander2074
    @robertbrander2074 3 ปีที่แล้ว

    Pure Genius ! ... Where did you learn to do all That ? ... That's an amazing amount of figuring ! ... Excellent !

  • @Mawukoattivor
    @Mawukoattivor 3 ปีที่แล้ว +2

    Nice rocket and video production. Are you willing to post the schematic and rocket stl?

  • @proto_hexagon5649
    @proto_hexagon5649 3 ปีที่แล้ว

    wow i like that easy and complex at same time.

  • @dmitrynuzhdin
    @dmitrynuzhdin 3 ปีที่แล้ว

    Cool! I am going to test my TVC rocket soon! My hardware is different but the overall approach to the problem is the same.

  • @ryanbrown172
    @ryanbrown172 2 ปีที่แล้ว

    Stubby and Stabilized. Also Stupendous

  • @panaisjean-baptiste3797
    @panaisjean-baptiste3797 3 ปีที่แล้ว

    Awesome work !

  • @vauhner81
    @vauhner81 3 ปีที่แล้ว

    Excellent work! Keep us informed. Got my sub. Cheers.

  • @goatsinker347
    @goatsinker347 3 ปีที่แล้ว

    Very awesome. Wish you had a super fast camera to slow the take off scenes down.

  • @salehali3400
    @salehali3400 3 ปีที่แล้ว

    Wow that’s a really good job

  • @noahsantoni7621
    @noahsantoni7621 3 ปีที่แล้ว

    Awesome video!

  • @mikeoliver3254
    @mikeoliver3254 3 ปีที่แล้ว +1

    Very cool and very well done, any chance you will post your code and STL files?

  • @perryrhodantheonly
    @perryrhodantheonly 3 ปีที่แล้ว +1

    Well done!!

  • @petewright2283
    @petewright2283 3 ปีที่แล้ว +1

    Very very cool. Don’t use delay charged motors if you can. You could pulse C6-P which is plugged or you can make your own.
    Lightly scrape the clay cap.
    Remove the granulated BP.
    Apply an epoxy cap and let it cure.
    I’d love to try and build one these. Have you thought about faster servos?

    • @jacobthornhill
      @jacobthornhill  3 ปีที่แล้ว +1

      I would buy plugged motors, but being in the UK limits what i can buy. Faster servos would be nice and I only used the ones I did because they are cheap.

  • @tomkuttler
    @tomkuttler 3 ปีที่แล้ว +4

    Im Building a similar rocket too and im looking for a example software that I can use as a base for my rocket. Did you publish the code somewhere?

  • @tinkmarshino
    @tinkmarshino 3 ปีที่แล้ว

    Damn interesting son.. I am gonna have to go look at your new ones now..

  • @Nifty-Stuff
    @Nifty-Stuff 3 ปีที่แล้ว

    Nice! I was excited to see the new version (now that it's been about 2 months since you posted this), but not yet? I'll subscribe and hope to see more!

  • @Pancake3000
    @Pancake3000 3 ปีที่แล้ว +1

    This is awesome.
    Reminds me of a certain joey b..

  • @scottaye9999
    @scottaye9999 3 ปีที่แล้ว

    Nice work

  • @NeilStansbury
    @NeilStansbury 3 ปีที่แล้ว +5

    This looks very similar to what BPS have been doing for the past few years, which I am sure you must be aware of. Is there any synergy or any learned lessons from their achievements that you have taken on board?

  • @rjmorpheus
    @rjmorpheus 3 ปีที่แล้ว +1

    This is good stuff!!

  • @general_isaac
    @general_isaac 3 ปีที่แล้ว

    This is amazing!

  • @Bob_too
    @Bob_too 3 ปีที่แล้ว +3

    QUIZ: What is the morse code message at the very end of the video ?

  • @jesperwall839
    @jesperwall839 3 ปีที่แล้ว

    Impressive! More of this 👍😀

  • @MakeItFlyofficial
    @MakeItFlyofficial 3 ปีที่แล้ว

    This is really good 👍

  • @Zrzyck
    @Zrzyck 3 ปีที่แล้ว +1

    unpopular recovery system. I like it. I'm using pyrotechnic due to weight, but you are clearly not focused on that aspect :)

  • @fab231242
    @fab231242 3 ปีที่แล้ว

    Amazing. Bravo

  • @herbert633
    @herbert633 3 ปีที่แล้ว +1

    nasa: you can't thrust vector solid engines, it can't be done.
    jacob: hold my beer
    (also credit to india for a different method but thrust vectoring nonetheless)

    • @oglordbrandon
      @oglordbrandon 3 ปีที่แล้ว

      The shuttle's Solid boosters had thrust vectoring.

    • @oglordbrandon
      @oglordbrandon 3 ปีที่แล้ว

      en.wikipedia.org/wiki/Space_Shuttle_Solid_Rocket_Booster#Thrust_vector_control

  • @davidbauer1485
    @davidbauer1485 3 ปีที่แล้ว

    Time to get a more powerful plugged, long burn composite motor!

  • @shere_kan8329
    @shere_kan8329 3 ปีที่แล้ว

    Great work !!

  • @sock501
    @sock501 3 ปีที่แล้ว

    This is amazing. However, the ‘chute ejection charge on the motor turns your rocket into a bit of a cannon at just above head height… that’s bad. Add some way to vent the gas that charge produces at the top of the motor mount unless you want a black eye!

  • @projectlongrange5942
    @projectlongrange5942 10 หลายเดือนก่อน

    hello, would you mind sharing the 3d files for the tvc mount? Your video inspired me and now I also want to build an tvc controlled rocket. But i dont really know how to design such an tvc mount...

  • @fignutron
    @fignutron 3 ปีที่แล้ว +1

    Very nice vid! I would like to see the design and code for this

  • @holemajora598
    @holemajora598 3 ปีที่แล้ว

    Awesome video! I’ll be looking for more. new sub!

  • @ahmethamdicelik1277
    @ahmethamdicelik1277 3 ปีที่แล้ว

    I believe if you improve the passive aerodynamics of the rocket you will get a more stable flight.

  • @guatagel2454
    @guatagel2454 3 ปีที่แล้ว

    Thank you!

  • @Leos-World
    @Leos-World 3 ปีที่แล้ว

    Nice. You might want to look into using Klima D3-P long burn motors for your project.

  • @masonmiller9792
    @masonmiller9792 2 ปีที่แล้ว

    i know im a little late but could you go in to more detail about the code and the PID controler

  • @ЕвгениГеоргиев-т1я
    @ЕвгениГеоргиев-т1я 3 ปีที่แล้ว

    Nice project. BPS ( Joe ) did a similar one. Wish you all the luck.

  • @anishchakraborty568
    @anishchakraborty568 2 ปีที่แล้ว

    Plz plz plz make a full tutorial on it
    It would be very useful

  • @alexcook4851
    @alexcook4851 3 ปีที่แล้ว

    Damn that’s cool !

  • @mkase4697
    @mkase4697 3 ปีที่แล้ว

    Nice setup you have there. I am sure in time you'll find ways to simplify and improve it. Perhaps by creating a custom board with integrated MCU, sensors and transmitters. No doubt you've already thought about this. A couple of bits I thought you might like to think about - disabling the parachute deployment charge [you'll need to work out how to do this but perhaps you can simply buy a motor without the charge] and using a solenoid latch instead of a servo for the parachute deployment door. Other than that, brilliant effort. Keep up the good work. I look forward to seeing your progress.

  • @matthewrichman7777
    @matthewrichman7777 3 ปีที่แล้ว +2

    could you provide the step or stl files for the tvc/rocket please! I would love to build one and modify it myself!

  • @knallpistol
    @knallpistol 3 ปีที่แล้ว

    How come you don't have millions of subscribers? This is sum fine work here mister. How did you learn this? Especially the coding part?

  • @lawson1033
    @lawson1033 3 ปีที่แล้ว

    This is really cool! Could you please post the STL files? This is something that I would love to try out.

  • @DynoRC
    @DynoRC 3 ปีที่แล้ว

    We shall thank algorithm gods for bringing us together once again

  • @rudraagrawal1988
    @rudraagrawal1988 3 ปีที่แล้ว

    Great work, i will suggest you using 29mm motor to reach more altitude and to get more flight time to get more data...

  • @raketenolli
    @raketenolli ปีที่แล้ว

    Very cool! Did I get that right that you're simply using the gyro output (rate of rotation) for each axis independently and feeding it into the control loop? You're not using the MPU6050's internal DMP that e.g. outputs filtered quaternions for orientation?

    • @jacobthornhill
      @jacobthornhill  ปีที่แล้ว

      Thanks! Yeah I am just using the gyro rates and integrating each individual axis to get orientation. It’s not a great solution, so on my current rockets I am using quaternions that are calculated by the flight computer.

  • @adventurfish
    @adventurfish 3 ปีที่แล้ว

    Very nice

  • @reeceburns5681
    @reeceburns5681 2 ปีที่แล้ว

    do you plan on sharing files? i’d love to build one too

  • @iamthebeastyt1518
    @iamthebeastyt1518 10 หลายเดือนก่อน +1

    Give us the code for the tvc

  • @adrianinvents
    @adrianinvents 3 ปีที่แล้ว

    Well done!