The AVA Flight Computer
ฝัง
- เผยแพร่เมื่อ 21 ก.ย. 2020
- Lemme know if you spot the error on the board. There is at least one... :)
Help support BPS.space: / bps_space
Optimal State Estimation - Dan Simon: amzn.to/2HoCYlV
Contact BPS.space: bps.space/contact
Second channel, mostly for KSP: / musicmakr
For more info:
/ joebarnard
/ bps_space
/ bps.space
/ bps.space
www.bps.space - วิทยาศาสตร์และเทคโนโลยี
Dude this whole video is ART. I don't even fly rockets, but the board is so pretty I want one just to frame it on the wall, but wouldn't waste one like that. Better to leave them for people making cool rockets. This is awesome, great work
U watch bps space?? You 2 are awesome it’s cool y’all know about eachother
You're both awesome channels, I was about to watch your spider silk video!
This is what happens when "somebody who went to film school" learns engineering.
Didn't expect to meet you here
@@jeffvader811 SAME
10 years later: "How i landed a model rocket on Mars."
using only AVA
Hopeful for that!!😍
It wouldn't be a model but he has said he wants to land something on the moon
I hope that happens too
With this size of rocket. It's possible but it would need boosters to get into orbit and for way to Mars it would need ion engine that doesn't exist, but maybe in future.
The fact you've taught yourself all of this is absolutely remarkable. Big kudos.
Most professionals were self taught. They just happened to go school at some time during their learning process (usually in the middle) and had to do most of their learning in the evenings, in the weekends, in the holidays or when they were blowing off class...
@@peterfireflylund that's my story I'm in 9th so I've to study theoretical physics on my own coz it's not taught in schools..I have a book TH-cam n ofc Google that's enough🌚🌚
@@tanmaydeshmukh3517 best of luck to you! Don’t skimp on the math. Complex analysis, linear algebra, and differential equations are *everything*! :)
Lots of the “self taught” people actually did go through formal education and may even have doctorates. Most of their actual learning was self directed, though. The official diplomas and the connections from a university make things a lot easier. Very few people can be Ramanujan and be recruited to Cambridge on the basis of their own research.
If you are from India, going through IIT or similar will probably be the easiest and safest path for you.
Learn to program as well. You can do so much with Python and Numpy...
late to the partry, but before the people later... Um, idk how to really say it in a way that people will listen. it's like Rick: School isn't a place for smart people.
not that people that go to school are all not smart, but smart is not what you become by attending. you become trained. it's not the same. when you're smart, you LEARN things. when you're trained, you do what you're told. Stay in school, smart people can learn from it. just don't fall into the mental trap of thinking it's teaching what you need. it's teaching what it wants you to know. if you have passion and dreams, that will really bother/drive you.
it's not evil(probably) it's just the way it IS.
maybe it was understood maybe not, but the point being that if you REALLY wanna learn stuff, you're gonna have to do the legwork yourself. i mean, let's be real, he's not starting from zero in any of the fields necessary... but he is ahead. this is one of many strong reasons open source is fought for.
“There’s a nonzero amount of aesthetics to be considered”
Spoken like a true engineer.
...except for the part where one needs a public audience to justify it.
Take a look at the Apple II motherboard. And the original Apple Macintosh motherboard. Beautiful.
If it looks good, it flies good.
Dont undervalue aesthetics which for sure has its roots within the *millions of years* of observation.
It is even symmetrical, looks better than the Arduino UNO board tho
Reminds me of the Marble Machine X process
so you're telling me the missile just got better at knowing where it is and isn't?
The missile knows where it is at all times. It knows this because it knows where it isn't. By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is greater), it obtains a difference, or deviation. The guidance subsystem uses deviations to generate corrective commands to drive the missile from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is. Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't.
In the event that the position that it is in is not the position that it wasn't, the system has acquired a variation, the variation being the difference between where the missile is, and where it wasn't. If variation is considered to be a significant factor, it too may be corrected by the GEA. However, the missile must also know where it was.
The missile guidance computer scenario works as follows. Because a variation has modified some of the information the missile has obtained, it is not sure just where it is. However, it is sure where it isn't, within reason, and it knows where it was. It now subtracts where it should be from where it wasn't, or vice-versa, and by differentiating this from the algebraic sum of where it shouldn't be, and where it was, it is able to obtain the deviation and its variation, which is called error.
@@Chezburger8 listen ive done a lot of complex stuff but you made my brain hurt.
Y'all are now on a watchlist for using the term missile instead of rocket lol
Shh!! It's not a missile! It's an advanced guided, rocket propelled UAV!
@@Chezburger8 I see your a wikipedia editor as well.
I love how this guy just graduated music school then said, fuck it, I wanna build rockets.
And here we are. Awesome.
People often don't realize it's actually a fairly technically orientated degree. Aidan Millward is another guy in the UK who has got something similar and it is surprising what his channel is about.
Ropponmatsu2 I don’t care what music school you went to, “quaternion non-linear control systems” aren’t on the syllabus.
@@BradyBegeman Audio engineering (which Joe did iirc) does include an awful lot of signal processing which is the backbone for most control systems. It's not a degree in aerospace, but it's a relevant background.
I don;t doubt it...But your point is what? I tried to explain it probably does include concepts like feedback, interference, and being able to solder together highly detailed, electrical connections.
I've done work with high powered rockets and space access and I have no qualifications either.
“quaternion non-linear control systems” are probably not relevant at this level that he is working at - Landing model rockets - and that is why this is interesting.
Wow
"I'm sure there's someone already in the comments angry with something I did."
Oh man, I'm an engineer and currently doing my Ph.D. in computer science, and I couldn't even try to criticize the work you do. That is a LOT of work in creating your own board. Communication patterns, I/O, electronics, hardcore programming... your work is admirable. Wish I had resources here to start learning how to fabricate my stuff like you do.
Keep on the hard work.
you will have one day
That's kind of what hes referring to, he did alot of "extra" work using 3 microcontrollers instead of just using something like a broadcom arm chip with potentially 100s of times the processing power, taking up less space, being easier to design the traces in the board and more. But at the end of the day its a hobby so he can do whatever floats his boat.
Perhaps one day a student will share the practical experience while you taking care the literature and novel theories, this is something common to happen here
Love the design! Your video stood out to me because I happen to have been one of the inventors of the NASA AVA (Affordable Vehicle Avionics) computer.
Oh dear... I had no idea, I would have picked a different name! Certainly didn't mean to copy it 🤦
As someone who has worked on drones, its funny to see you working on a similar path that drone control followed. Started with basic PIDs and slowly got better with full state constol, FOC motor control, fancy filters, better and faster protocols ect.
Same thought here. I remember building quads when you had to rip the guts out of Wii remotes to get your sensors and such then cobble everything together and spend a long time tuning the PIDs manually. Within a few short years commercial off the shelf offerings were essentially plug and play at all scales.
Why not simply use a matek F7 board XD
@@marco9409 that's actually a pretty valid question.
@@marco9409 I would suggest the SP Racing H7 extreme, and strip betaflight leaving the filters, PID control, blackbox and other useful things.
Started with simple PID, simple drones e.g. mikrokopter, 8 bit atmega and ends in Ukaine where DJI carrier and drop missiles.
As an student finishing his degree on engineering, I congratule you. You are one of those guys who get things done instead of just talking.
I've learned a lot of concepts from you.
Dont care about the critics, they are probably from the ones who does not do the things you do. Jist keep it going!
Greetings from Paraguay!
So I don’t know how I stumbled on your channel, but I’ve stuck around a little to see your embedded design. I’m an aerospace engineer and used to work on engine controllers and think it’s awesome you’ve integrated your own processors and sensors. It’s probably not necessary for your requirements but I think the next “steps” for your design would be to create an enclosure for the PCB. This gives you an opportunity to tackle more environmental conditions like shock/vibe, sand/dust, flammability, temp cycling etc like what aero industry does with airborne hardware via DO-160G. Use connectors instead of terminal blocks to decouple your avionics from the main rocket system and allow you to potentially hermetically seal your enclosure (which is the easiest way to tackle many of the environmental conditions, especially full humidity!) This will also evolve the design of the rest of your electrical system as it will likely require you to make a wire harness to consolidate the signals into a single or couple connectors. Anyways I’m sure none of this is necessary but I appreciate you sharing your lessons learned and in the spirit of community contribute mine to you or who ever reads this. I hope you make a video on how you modeled your system for the kalman filter predictions and how you handled the continual accl. drift. I think that would be a very significant contribution to TH-cam.
The thing about the ardunio platform is... the forums are massive, a random person has probably already had a similar error, cant say that about some other microcontrollers
That build/reflow montage in the beginning is just pure eyecandy. Uuuugh so smooth
Perfect timing! Just released my Teensy-based flight controller for electric VTOLs too :)
Nicholas Rehm =O
Will your FC be able to work with the AVA for your projects?
you also do great stuff bro. Keep it up.
As a model rocket enthusiast, I absolutely love this channel and seeing the depth that it goes into. You’re doing great work and inspiring many of us to dig deeper into the hobby.
The BEST video I’ve seen all year! I really enjoy your logic and engineering perspective. Good on ya for hand making that board and amazingly high quality video presentation.
This is some really high quality video .... Big up man!
Yaad man? Lol
@@thebirdhasbeencharged ?
@@invisibleimpostor7311 where you from? The Caribbean?
@@thebirdhasbeencharged I don't know...
Finally putting that music degree to use in that intro I see.
Nice.
He writes like 3 new songs for each video
I have two modes when watching this channel:
1. Rocket go brrrrr
2. Nerd out on hardware and software
Keep up the good work, love the content!
Really, REALLY good video on your board, and wow, what an incredible amount of work that has gone into it's development. You captured highlights, lowlights and general commentary, all with a good sense of humour. Nicely done!
>16g, have fun
Wot? Wdym?
Lol exactly what I was thinking
Incredible. Seeing that sprint flight with all the data around it makes you realize... Joe has really built his own Space Agency from the bottom up. So cool.
Simply amazing! It is so great to follow your journey and to see the progress in your engineering!
I build quadcopters with hardware that, while not custom like this, is similar in component usage. The crazy thing is I used to be very active in rocketry (I still have about 20+ in my workshop) but never had the idea to do stabilization, guidance, telemetry, etc. It makes perfect sense to me now :)
You have to stop uploading these videos right when I want to do my homework. Now I'm going to procrastinate for the next hour again
He uploaded this 1 minute before my google meet class started 😂
Same lol
Hi Joe! A video I think people would find interesting is discussing the details of the Kalman filter. Im learning about it a little in school but the functionality you have implemented seems very interesting! Love the work you've done!
This is the first video I've seen on your channel and it's SO well made! I'm definitely taking a look through your account, thank you for sharing your passions!
This video might be one of your best!
Nice balance between being informative and funny!
Great work man!
Wow. Just wow. You can be damn proud of how far you've come. You're so smart. Big congrats king.
the aesthetic, its beautiful, keep it up
I can see how much you love what you do. I’ll look at your kits to see what you offer. The idea of building this is like brain Candy. Thanks so much.
What a beautiful flight processor, it's layout makes sense, its really nicely put together and it looks like it works well from later videos. I can't wait to see what you do with it.
Okay, so this is the most complex PCB made by a TH-camr I have ever seen. Great job on the PCB by the way.
I just realized what a golden age of electronics we're in. 10 years ago having a commercially printed privately designed PCB would have incredibly expensive.
Love this stuff...
And now primitive kits to print your own PCBs at home, or office.
This is beautifully shot and the editing and post was done exceptionally well.
this is amazing, just starting to play around with robotics and I never even thought about designing a custom board to coordinate all the circuitry even though that's kinda what we're doing when we integrate all the various components (pwm, gpio, sensors, telemetry, etc. etc.). this is a beautiful board that is wonderfully and carefully thought out and a fantastic balance of price, performance and extensibility. sorry I'm late to the party but thank you so very very much for sharing your creation. just discovered your channel so I can't wait to see what else you've been up to. cheers!!
"I think joes cheating on me, whenever he goes to make videos he keeps talking about this girl Ava."
As an electrical engineer. I find your channel amazing. Keep it up
I love custom stuff like this. This board, the layout, even the logo is sexy AF! Nice work man.
Love your work.the board is clean and the sensors and stuff is so complex
Dude you are probably better than a lot of senior engineers!!
Oh my god he sounds like a true computer engineer now, great job!
You are doing so well my friend! You make everything look so easy, keep up the good work!!
Your passion for building these videos is absolutely beautiful.
*Technology at its finest*
Keep it up dude :)
For a next iteration of this computer:
You might be able to combine the MK20 and bluetooth chip with something like an nRF52 family processor. Next to being a SoC with an MK24-like processor, it has bluetooth on-board, which saves the energy overhead of a separate chip. nRF chips are also a lot more flexible when it comes to pin muxing: (almost) any pin can do anything.
A good SDK (written in C) is available with a lot of examples.
I've use MK24s and nRF52 series for work and was pretty impressed with the latter.
In a way this channel is so wholesome.. Just a creative, intelligent and passionate person inventing cool stuff and literally doing rocket science as a hobby. I'm super impressed by this guy!
Awesome build quality and video production! Inspiring! Keep on keeping on!
Notes...
4:55 - Slight, passive/aggressive "Woo-Tang Moment".
6:34 - "...YET!"
7:05 - High-order "Woo-Tang Moment" born out of frustration (ODB raises an eyebrow from the afterlife).
9:51 - In other words, "zero f%cks given...WOO-TANG!"
13:18 - slow, yet steady, build-up to frustration-based, nerd rage-onset of "Woo-Tang Moment". (deep breath, Joey...we're with you. Suggestion: Bourbon helps).
More to follow.
The error on the board is visible at 2:54: “G RX RX 3V” -> should be Rx Tx ?
Your content is so good and your depth of knowledge is so great I simply had to subscribe!
Excellent....i know how hard it is to make this type of flight computer and u done it alone all the coding is done by you ,the board is made by u .... inspirational ..salute to your hardwork
1:23 two rx pins marked?
And black wire on data instead of ground 😔
As he said: "there are better options".
@@falconidaes lmao
That's a UARR connection
And he had to mount the connector the wrong way around, whichmakes it a pain in the a** to remove that cable, as you can't reach the latch with your finger.
2:26 nice save
Porter Campbell lmao
Love your video man every second is full of experience and information
Haven't watched in over a year had tons going on. Slowly catching up. Awesome work and progress. Gonna start following better and maybe try a few of your tricks.
Any suggestions on how to start developing own PCBs without much electronic background or never having it done before? How did u do it?
There are tons of tutorials in the internet. What i find easy to start is "toner transfer" method using laser printed paper. You can even draw circuits with a permenant marker and have a pcb.
UV fluorescent lights are cheap. Photo PCBs (glassfiber + copper + a layer of photoresist) are cheap. The chemicals are cheap. You can print the pattern on a normal laser printer. You can use normal paper if you like but the exposure times are dramatically shorter if you use drafting paper (because the white clay in normal paper blocks most of the UV light). You can tape a piece of drafting paper on to a "carrier" sheet of normal paper and put that through your laser printer.
Lots of tutorials, both in text form (with photos) and youtube videos.
The chemicals you need are whatever you use to etch the copper (lots of options for that), whatever you use to remove the non-hardened photoresist after exposure (a little drain cleaner works fine -- but it only takes a few seconds to remove and if you wait too long, the hardened photoresist also goes), alcohol works fine for removing the hardened photoresist after etching away the copper.
Getting a high resolution is surprisingly easy. Getting repeatable results is also surprisingly easy.
Drilling holes in the PCB is actually harder. Sawing/breaking the PCB into smaller pieces is harder (and the dust is quite nasty, so be careful).
Don't wait for someone to teach you. Don't wait for the ideal teaching materials that are suited perfectly to you. Just start doing it. It is far faster and easier than trying to be perfectly prepared first.
We are back....!!!!!
Nice to see the upgrade after the sprint flights.
Shame about the humidity though, but mistakes make us better man....
Amazing Work.
Awesome work! I hope to get to flying stuff soon, but really AWESOME work! Keep it up, I would love to see more.
This is The greatest video i saw till date. Thanks a million. Keep doing the greatest work.
also maybe next time we'll see an fpga implementation of all this, and then a shiny new asic for the next iteration 🤞
Board error: your serial port connector has two "receive" pins -- or at least so says your silkscreen. ;)
And one of them is swapped with the ground line (black wire).
@@vannoo67 to be fair cable color doesn't mean anything other than ease of maintenance, so is the silkscreen.
@@thomasesr But to also be fair - they are both mistakes.
@@vannoo67 also, to be even fairer. I know, that is why I mentioned the silk screen as well.
Every engineer knows that RX and TX are ALWAYS swapped, it is a serial Murphys law...
Fantastic all the way around Joe. I've experienced some of the same conditions working on my own automotive CANbus based logger/power/interface. Always fun hearing someone else talk about design hurdles and solutions because there are so many ways to solve the issue. I learn something new everyday. I'm still in ardi-land too but now my justification is "well, Joe Barnard can do it, so I can too" :p
terrific work on that chip mate, keep up the great work
I was like "Wow Joe doing a serious video this time"
Untill 7:00
Got me by surprise and cracked me up
i rewatched that it’s fun seeing not so serious joe
Who wants to tell him I2C is a serial communication protocol?
Ik he's referring to other low data rate protocols and I get his frustration, I'm currently looking for a solution to bypass my Prusa Mk3S serial connection and directly write to the SD card.
your channel is unbelievably satisfying. good shit
At 7:00 and at 13:29 I laughed SO hard, my friend wondered why I was laughing that hard, and when I showed him he laughed just as hard as me. Your channel is sooo underrated, needs more subs for sure
you could sell these as kits or already assembled & configured/programmed **edit: oh i guess they are for sale! (mentioned near end of video; but not in a mass produced, off the shelf capacity - yet)
Clearly a good display of an ongoing project. I noticed some things that I would do differently, as you stated, but it’s really about the learning process, the way we face difficulties and overcome them that makes up for the immense pleasure of creating something. Whatever way you chose to do things, whether in a monolithic fashion or using a clear separation of concerns when it comes to hardware, it makes no difference as long as the most important thing happens; that it works. Optimisation comes after. A single MCU that is powerful enough to run all processes would have a smaller footprint and weight, but still, unless you are entering some competition, it’s not the most important thing. As long as you have fun and learn along the way, it’s never wrong.
I like how clean and satisfying the video is
Bro how you get so information from youtube or reading books
What?
@Lemon Pasta in music
Looks like the ATLAS logo from advanced warfare. Giving me weird vibes.
Great project presentation and very cool looking board as well
I was sent this by a friend. Really interesting video and fun to see the hardware-software mix.
I love the hardware stuff but when you get into software man my brain cooks. I wish I was your neighbor so I could make a hybrid liquid solid rocket for you.
If he looks like Elon and he acts like Elon, he IS Elon.
awesome content cool to see this channel also give bits of computer engineering along the way it inspired me to embark on the journey of making my own custom linux single board (flight) computer with a RK3399 SoC
I love your videos Joe!
I’m working on a rocket rn, I am gonna see what happens if I replace the fins with grid fins that are in small motors
Wondering who the "we" is he refers to constantly :) Maybe he's already done the AI Neuralink ahead of Elon :)
Beautiful work, this is great, this is art.
Truly amazing work!
And i watched 18 min ad.
Dude, you ROCK! And your this design, RULES!
Awesome board, awesome video! Thanks Joe!
I have no idea how this ended up in my recommendations but this is awesome. What an awesome rabbit hole this whole thing sent me on.
Im astonished as to how much thought went into the whole design. Especially the redundancy part with skipping multithreading and doing it with 3 processors in hardware. But one quick change personally do when I have multiple chips that use a USB connection, is to put USB hub IC on the board so there is only one port on it, that connects to everything.
Your content is very interesting and inspiring. Thank you so much for being creative.
What a gorgeous board. Nice work.
Love seeing more use of the MS5611. Definitely my favorite!
I like the admission of feeling embarrassed about sticking with the Arduino platform and realizing it doesn't matter. An amazing project you have going on and kudos for sharing with the rest of us.
I have absolutely no idea how I landed here but watched through the whole video as I was just fixated on your explanations.
Amazing work from what I can tell from my limited understanding.
Legend in the making, keep up the amazing work :)
Love it! Keep going! Well thought and designed!
I didn't understand over half of what you said but this video was fantastic. I am a child of the 1950's and was one of those early model rocket builders. My one dream was to make a rocket that would do more than fly high and return with a chute. My Dad was an and electronics expert and built a simple circuit board to include in a rocket to measure air temperature but the only thing it would do was change the frequency of the tone it transmitted back to the ground. It never flew as it was too heavy with the board and 9-volt battery. So your work here makes me think of those early days of cardboard tube, balsa wood fins, and lots of paint to make them look good.
Absolutely brilliant work!
I’m just learning how to code now, so I’m on the raggedy edge of being able to understand anything that was said here, but it’s fascinating and I’m thankful for the recommendation.
Lovely work, beautifully presented and explained. Big thumbs up from Tassie. 👍😀
Honestly I'm really happy that I had an internship recently as an embedded software dev because it was really nice to have a lot of context for everything on AVA. Nice job and I feel you on UART.
Embedded dev is dead. Nowadays it's supposedly about knowing how to parse XMPP stanzas and fiddling with device tree files on a Raspberry Pi. Oh, and about squandering about a hundred or so instructions at a time on a DigitalWrite that can be done in one.