Great job overall - I like that you showed what did, but also what didn't work in your designs. If you want, feel free to send over your design files - I'd be more than happy to feature your board in a design review video on my channel.
I would love to have you review the current board! Please feel free to tear it to shreds, I want to improve the next design as much as possible. Thank you so much for watching, files are coming your way! ;)
I have just finished developing my first flight computer and I am currently waiting for the JLC shipment to arrive. This video would have been so helpful to me a couple of months ago. Great content
You might want to use a switching regulator instead of a Linear regulator. If the difference between the output voltage and input voltage is high, an LDO will simply dissipate that as heat. A switching regulator is more efficient in this case but you have to be careful with filtering, because it also introduces some noise. Great project.
I do agree about the in-to-out difference if the currents being drawn from an LDO are high, however SMPS regulators introduce noise that typically is not good for analog and sensing. A better approach would be to have a SMPS feed an LDO. Make the SMPS be 10 to 20% higher that the LDO dropout on the input side. Also, one should not use LDO in high load situations given that they are normally used for analog anyway.
I disagree, LDO's are better for low noise requirements. Its better to do a 2 stage LDO stepdown for the voltage. M7806 for first stage and a 1A version of the 1117. There are larger packages for LDO's that deal better with heat dissipation and a small copper heatsink directly below the LDO on the bottom side of the board will provide better thermal stability
@@sjwpcbdesign 10 to20% is vague. All LDO's have a rated dropout voltage usually ~ 1.3 V for 1117 series and 2V for 780x series. Its best to step down voltage close as possible to minimum dropout without ever going below it. That way you are not dissipating much heat.
@@FriendlyIntentions First, 1.3 to 2 volts is not a typical drop out. That sounds more like a traditional regulator. LDO should be more on the order of 0.2 to 0.5 volts. The 10 to 20% is with respect to the Vin + drop and add a 10 to 20% margin of the drop out. For example if you want 3.3 out of the LDO, your Vin would be Vout + 0.2 + 0.04 = 3.54V assuming 20% of drop out. Therefor the out from your switching regulator is 3.54V. If needed you can add a pi filter in front of the LDO. However I have found this not necessary in most situations.
@@sjwpcbdesign ams1117 spec is 1.2V dropout . You are thinking low current LDO's, those suck. A 3.3V ams1117 in sot 223 with a VIN of around 5V can easily supply 500ma with it only getting to around 20-30C above ambient
Wow, this is truly impressive! Developing a flight computer for a rocket with limited knowledge is an incredible achievement. Your explanations are clear and easy to follow, and it’s fascinating to see the step-by-step process you went through. It’s evident how much time, effort, and passion went into this project. Respect for this extraordinary work, and thank you for sharing your expertise with us! 🚀🔧✨
This is awesome work and documentation dude. The presentation is the cherry on top to the great engineering work. I do avionics for a living and am very impressed man! I hope you apply to the top places cause they'll definitely be lucky to have you!
I would have watched this even if it was 1 hour. Great explanation and very impressive work. I want to create my own PCBs for my rovers but I find it challenging. This was a great motivation for me to learn more about pcb design.
Very interesting work!! Sharing the evolution through 3 versions was a great choice. That 1062 board was probably overkill as flight computers don't need that much computing power.
This is awesome keep up the work i am currently working on designing my own first board but i am having problems with the SD card and esp32 but this video will help a lot especially with the voltage switching regulator i am using for my board that is the hardest part but keep up the good work.
Very nice overview. I work as a DC-DC Power Engineer, and might I suggest the TDK FS1403 uPOL it is a little overkill for your application but you should avoid the thermal problems you mentioned in your video.
Increasing measurement sensitivity have little impact about noise. You need to look at accelerometer and gyroscopes mathematical model. In there you will see bias terms and random walk values. Also temperature is a significant figure in noise. Keeping steady temperature is important. If you understand statistics and mathematics behind this, you can create simulations of your sensors and rocket model and get a greater result. Keep going, very nice work. 👏
It’s worth knowing that the board that it can plug onto I understand that this is your one and that someone else can make another one but I would highly recommend trying to consent to one board board, multiple board boards and the connectors especially in some rocketry things can simply add too much weight
This is excellent work! What are you using to display all your telemetry data? I can tell it's a custom interface, but did you use some kind of Java framework or something?
Being an ece(electrical and computer) student and not quite grasping all this, it seems really cool, but for our robotics club we just buy off thr shelf boards😭. I hope I’ll figure out pcb design and python(code makes me wanna dsewerslide).
Pretty impressive. Thank you for sharing you project. Did you have to earn an engineering degree to gain all your understanding in flight control, or is this a hobby for you?
Nice to see your design iterations! I would use a flash instead lf the SD card, and definitely look into kalman filters for the sensors. I see that your power module is a bit problematic. Why don't you make a 3rd board, and make your design more modular? This way you could iterate faster or certain issues while cutting costs on other parts
Thanks for your feedback! Versions 2.0 and 2.5 included both an SD NAND Flash chip and an SD card slot. The next design will definitely feature more advanced sensor fusion, like Kalman filtering, and a better board architecture.
if the buzzer is a piezoeletric device, it's a capacitor. so the bjt firstly charges it to the voltage, and as the capacitor wont have any path to discharge, it maintain its voltage and doensn't make any sound. you need to add a resistor in parallel to the piezo in order to make it discharge and charge it again on next bjt turn-on.
Hi Kofi, Nice work! it is really amazing the way that you are developing both code and hardware by your own! I've been looking for the "software" version of this video but NULL was found ejejeje, would it be possible to get a Github link or a video explaining how are your developing the code? Thank you so much!
hi i am also getting started on my journey of working with more advanced microcontrollers, can u suggest any tutorials on learning how to code an stm32?
Hey Kofi! I designed a custom STM32L4 PCB, but for some reason it is not recognized via usb on my computer. Did you use the SWD peripherals and utilize STM32 cube programmer? Any info helps, it’s my first time using anything stm related. Thanks!
You need a fast GPS receiver, and I do not mean fast serial port I mean fast GPS coordinates updates, milisecond or better ideally. This plus some fancy 3axis + time trigonometry and you can get this thing to track a path. On that subject, I am saying really fast GPS coordinate updates because, rocket science... A normal GPS receiver on this board should more than suffice for drone airplanes, copters etc unless you plan to go supersonic or something hehe
I am interested in your firmware/software that you did for this controller. Normally IMU's have sensor drift. Did you implement a form of sensor fusion to mitigate this problem? Also, I am curious about the board design after seeing some of your chip sets damage. This should not happen unless you missed some design steps in the selection of parts.
The next video will be all about the firmware/software dev! I think the primary reason for the dead MCUs I experienced was due to my poor handling causing things like shorts, transients, ESD, etc. I don't recall any "random" failures though.
Is it too expensive to use Altium for personal project ? Your references are also my favorite channels too 😂 Im not pro into EE or Electronic/PCB, wonder if it's possible to design my own board with self-learning 🤷♂️?
I am unaware of Altium's exact pricing but I find it to be quite unaffordable without university/employer sponsorship. That said, I've been hearing great things about Kicad and will certainly be exploring it for future projects(especially after school ;)). Also, you can absolutely self-learn to create your own PCB, especially with those resources. Good luck!
@@KofiAsare0 KiCAD perhaps is a little more than the web app EasyEDA I think. Tried to use it but still prefer easyEDA for small project. The amount of mistakes that I myself can't verify really made me to always ask peopel to verify it for me 😂
The current design has a 30.5mm x 30.5mm mounting hole pattern and can run a modified version of something like Ardupilot, INAV, or Betaflight. However, considering the large form factor and lack of vibe resistance, it may not be optimal. I hope to vastly improve on this in the next design! 🙂
Hi I am a begineer in electronic Did you have or know any website to build my rocket? I ever built my FPV Fixed wing using F405Wing as my flight control. I use Inav to control my flight but I no nothing about programng
The gold border is a ground ring added primarily as an easy way to clip oscilloscope ground probs. It also looks pretty and stiches the internal ground planes together.
Great job overall - I like that you showed what did, but also what didn't work in your designs. If you want, feel free to send over your design files - I'd be more than happy to feature your board in a design review video on my channel.
I would love to have you review the current board! Please feel free to tear it to shreds, I want to improve the next design as much as possible. Thank you so much for watching, files are coming your way! ;)
@@KofiAsare0 Could you please share it with us the viewers as well? I'd love to take a look at this beauty myself! :) Thanks looking forward to it!
@@KofiAsare0 no response, guess Ill have to do it myself
Wonderful overview of your flight computers. Thank you for sharing why you made all of your design choices. I’m excited to see your future projects!
Thanks for watching, I'm absolutely looking forward to seeing your future projects as well!
I enjoyed your descriptions and lack of background music. Very professional.
I have just finished developing my first flight computer and I am currently waiting for the JLC shipment to arrive. This video would have been so helpful to me a couple of months ago. Great content
You might want to use a switching regulator instead of a Linear regulator. If the difference between the output voltage and input voltage is high, an LDO will simply dissipate that as heat. A switching regulator is more efficient in this case but you have to be careful with filtering, because it also introduces some noise. Great project.
I do agree about the in-to-out difference if the currents being drawn from an LDO are high, however SMPS regulators introduce noise that typically is not good for analog and sensing. A better approach would be to have a SMPS feed an LDO. Make the SMPS be 10 to 20% higher that the LDO dropout on the input side. Also, one should not use LDO in high load situations given that they are normally used for analog anyway.
I disagree, LDO's are better for low noise requirements. Its better to do a 2 stage LDO stepdown for the voltage. M7806 for first stage and a 1A version of the 1117. There are larger packages for LDO's that deal better with heat dissipation and a small copper heatsink directly below the LDO on the bottom side of the board will provide better thermal stability
@@sjwpcbdesign 10 to20% is vague. All LDO's have a rated dropout voltage usually ~ 1.3 V for 1117 series and 2V for 780x series. Its best to step down voltage close as possible to minimum dropout without ever going below it. That way you are not dissipating much heat.
@@FriendlyIntentions First, 1.3 to 2 volts is not a typical drop out. That sounds more like a traditional regulator. LDO should be more on the order of 0.2 to 0.5 volts. The 10 to 20% is with respect to the Vin + drop and add a 10 to 20% margin of the drop out. For example if you want 3.3 out of the LDO, your Vin would be Vout + 0.2 + 0.04 = 3.54V assuming 20% of drop out. Therefor the out from your switching regulator is 3.54V. If needed you can add a pi filter in front of the LDO. However I have found this not necessary in most situations.
@@sjwpcbdesign ams1117 spec is 1.2V dropout . You are thinking low current LDO's, those suck. A 3.3V ams1117 in sot 223 with a VIN of around 5V can easily supply 500ma with it only getting to around 20-30C above ambient
Wow, this is truly impressive! Developing a flight computer for a rocket with limited knowledge is an incredible achievement. Your explanations are clear and easy to follow, and it’s fascinating to see the step-by-step process you went through. It’s evident how much time, effort, and passion went into this project. Respect for this extraordinary work, and thank you for sharing your expertise with us! 🚀🔧✨
Loved that detailed explanations. Great job Kofi
This is awesome work and documentation dude. The presentation is the cherry on top to the great engineering work. I do avionics for a living and am very impressed man! I hope you apply to the top places cause they'll definitely be lucky to have you!
Great job, well documented...
I would have watched this even if it was 1 hour. Great explanation and very impressive work. I want to create my own PCBs for my rovers but I find it challenging. This was a great motivation for me to learn more about pcb design.
Maaa, new engineer dropped!
Nice Work 😄
Really looking forward to the next video about the software overview. So good man youre amazing!
Keep up the good work! Loved the details about choices of each component 👏🏻
Yeah 👍
The purple and gold is an aesthetic!
Very interesting work!! Sharing the evolution through 3 versions was a great choice. That 1062 board was probably overkill as flight computers don't need that much computing power.
very nice man, learning fast and keep going
This is awesome keep up the work i am currently working on designing my own first board but i am having problems with the SD card and esp32 but this video will help a lot especially with the voltage switching regulator i am using for my board that is the hardest part but keep up the good work.
Great breakdown, can't wait to see more progress!
Great work mate. I can see your efforts .👍
Phenomenal - Just subscribed :)
Very nice explained. I liked you way to present the information. 🎉
Great job 👏👍 will await for the 3.0 🎉
Great video once again!
This channel deserve moarrr subscriberrrrrrrrrs
Very nice overview. I work as a DC-DC Power Engineer, and might I suggest the TDK FS1403 uPOL it is a little overkill for your application but you should avoid the thermal problems you mentioned in your video.
Increasing measurement sensitivity have little impact about noise. You need to look at accelerometer and gyroscopes mathematical model. In there you will see bias terms and random walk values. Also temperature is a significant figure in noise. Keeping steady temperature is important. If you understand statistics and mathematics behind this, you can create simulations of your sensors and rocket model and get a greater result. Keep going, very nice work. 👏
Will certainly look into IMU temperature stability and noise models, thanks for watching!
Awesome work dude
Good work, definitely will follow the progress
Cool! What did you use to make that gui? Looks really good
Yeah 👍
Processing IDE Java, more in the next video!
Looks good nice bro
It’s worth knowing that the board that it can plug onto I understand that this is your one and that someone else can make another one but I would highly recommend trying to consent to one board board, multiple board boards and the connectors especially in some rocketry things can simply add too much weight
This is excellent work! What are you using to display all your telemetry data? I can tell it's a custom interface, but did you use some kind of Java framework or something?
Hello! Nice video!
Could you discuss the graphical interface?
Exciting stuff.
Yeah 👍
Thank you for watching, the next video will be all about the software!
@@KofiAsare0 Nice! Hope to see it soon.
Greetins from Brazil
Really impressive thank you for Sharing your journey
Thank you for watching!
Great Job, Kofi
Uau!! estou impressionado!!!!!!
nice work
Being an ece(electrical and computer) student and not quite grasping all this, it seems really cool, but for our robotics club we just buy off thr shelf boards😭. I hope I’ll figure out pcb design and python(code makes me wanna dsewerslide).
Amazing project!! (Also a fellow Ghanaian 😝)
Really good video information. Do you have schematic of the design.
Great project good job !
Pretty impressive.
Thank you for sharing you project.
Did you have to earn an engineering degree to gain all your understanding in flight control, or is this a hobby for you?
Yeah. I’m also expecting answer of this question . .
Thanks for watching! This is mostly a hobby for me, I am also working on an undergrad degree!
Nice to see your design iterations! I would use a flash instead lf the SD card, and definitely look into kalman filters for the sensors. I see that your power module is a bit problematic. Why don't you make a 3rd board, and make your design more modular? This way you could iterate faster or certain issues while cutting costs on other parts
Thanks for your feedback! Versions 2.0 and 2.5 included both an SD NAND Flash chip and an SD card slot. The next design will definitely feature more advanced sensor fusion, like Kalman filtering, and a better board architecture.
if the buzzer is a piezoeletric device, it's a capacitor. so the bjt firstly charges it to the voltage, and as the capacitor wont have any path to discharge, it maintain its voltage and doensn't make any sound. you need to add a resistor in parallel to the piezo in order to make it discharge and charge it again on next bjt turn-on.
Hi Kofi,
Nice work! it is really amazing the way that you are developing both code and hardware by your own!
I've been looking for the "software" version of this video but NULL was found ejejeje, would it be possible to get a Github link or a video explaining how are your developing the code?
Thank you so much!
Thank you for watching, the software video is on the way and will hopefully be out relatively soon!
amazing !!!
nice video!!!by the way, how do you develop your own flight control application(C# or qt)and process your data you get from your board?
hi i am also getting started on my journey of working with more advanced microcontrollers, can u suggest any tutorials on learning how to code an stm32?
Phils Lab is definitely my goto for the advanced dev!
Amazing, but missing one crucial detail: Your logo!
Awesome ❤
Hey Kofi! I designed a custom STM32L4 PCB, but for some reason it is not recognized via usb on my computer. Did you use the SWD peripherals and utilize STM32 cube programmer? Any info helps, it’s my first time using anything stm related. Thanks!
You need a fast GPS receiver, and I do not mean fast serial port I mean fast GPS coordinates updates, milisecond or better ideally. This plus some fancy 3axis + time trigonometry and you can get this thing to track a path. On that subject, I am saying really fast GPS coordinate updates because, rocket science... A normal GPS receiver on this board should more than suffice for drone airplanes, copters etc unless you plan to go supersonic or something hehe
Bro which is the software used for GUI. Is that Processing ?
Yeah 👍
Yep, more in the next video ;)
@@KofiAsare0 im waiting...
I am interested in your firmware/software that you did for this controller. Normally IMU's have sensor drift. Did you implement a form of sensor fusion to mitigate this problem? Also, I am curious about the board design after seeing some of your chip sets damage. This should not happen unless you missed some design steps in the selection of parts.
The next video will be all about the firmware/software dev! I think the primary reason for the dead MCUs I experienced was due to my poor handling causing things like shorts, transients, ESD, etc. I don't recall any "random" failures though.
fire!!!
Is it too expensive to use Altium for personal project ?
Your references are also my favorite channels too 😂
Im not pro into EE or Electronic/PCB, wonder if it's possible to design my own board with self-learning 🤷♂️?
I am unaware of Altium's exact pricing but I find it to be quite unaffordable without university/employer sponsorship. That said, I've been hearing great things about Kicad and will certainly be exploring it for future projects(especially after school ;)). Also, you can absolutely self-learn to create your own PCB, especially with those resources. Good luck!
@@KofiAsare0 KiCAD perhaps is a little more than the web app EasyEDA I think. Tried to use it but still prefer easyEDA for small project.
The amount of mistakes that I myself can't verify really made me to always ask peopel to verify it for me 😂
@@KofiAsare0 did you try CircuitMaker ?
I was curious about what hardware would be fast enough for something like this.
Será que isso teria uma forma facil de se adapitar para ser usado em aero modelo? Uma pergunta que veio na minha cabeça.
The current design has a 30.5mm x 30.5mm mounting hole pattern and can run a modified version of something like Ardupilot, INAV, or Betaflight. However, considering the large form factor and lack of vibe resistance, it may not be optimal. I hope to vastly improve on this in the next design! 🙂
parabens!
Hi I am a begineer in electronic
Did you have or know any website to build my rocket?
I ever built my FPV Fixed wing using F405Wing as my flight control. I use Inav to control my flight
but I no nothing about programng
interesting i would love to have a look into the board can you send the files ?
what software / frame work you use for this real time data visualizations is it QT or something else ?
I used Processing IDE Java, more info will be in the next video!
@@KofiAsare0 Ok i will definitely check it out
Hi, what are the name of the conectors you used?
Board to board connectors used are the Harwin M58-3800442R(main) & M58-2800442R(carrier)
do you have a github aur something where i can have a lookat your project?
this would be fun if i had the resources
How did you built the telemetry GUI?
The GUI was written using Processing IDE Java, which will be covered in the next video!
What is the reason for the gold border around the board?
The gold border is a ground ring added primarily as an easy way to clip oscilloscope ground probs. It also looks pretty and stiches the internal ground planes together.
@@KofiAsare0 does it in any way help with EMI/EMC?
Is a open source project ?
NASA hire this man already
Is this open source?
9:47 DAYMNNN YOU LOOK PRETTY MY CHOCOLETTE MEN