12/1/2025 - Setting the specification

As mentioned in the description, the goal of this project is to make a fully functional advanced flight computer.

I started by reflecting on my previous flight computer, which did not have the chance to make it to the skies. That was also my first attempt at making a flight computer.


Reflecting on the features I liked about the previous flight computer and what I wish I had, I am able to set a pretty comprehensive list of features I want on this flight computer.

I've been looking through datasheets as I complete this list of definitions to get a grip on what to use. There are many options, so proper research is a must for this.
From my previous project and experiences, it is important to define the features properly before starting a project so that we are able to manage constraints effectively, and that's why I decided to start this way.

Next, I'll start researching and selecting parts, creating the repository and setting up KiCad.

12/2/2025 - Research and select parts

I spent the past 2 hours searching for parts, checking compatibility and managing cost to performance ratio. Also created the repo and organised files in the recommended manner. Still not completely done with the selection process yet but at least we made some progress.

12/6/2025 - More sourcing and schematic work

Finally, after a lot of pain and suffering I have selected the parts I need to use and added them into the schematic. The power delivery part reamins a big headache as I am a bit confused by the datasheet so I'm working around that by using components I know will work based on online designs and other hack clubbers' work. I still have certain things to verify and hope to fully understand all the terms and knowledge in a datasheet so that I can pick components more independently.

I have also picked a cheap chinese gps module to save up on cost and the Ai thinker Ra-02 as the lora module.

Next I will be working on finishing all the power system wiring and components then work on component wiring.

12/7/2025 - Schematic work

A lot of work done today, got most of the components wired up. Although there are still certain things that I'm confused about, I'll take the time to learn them.

12/10/2025 - More component selection and schematic work

Added servo, selected MOSFET and double checked capacitor values. I am quite confused about how to route the P-channel MOSFETs. I burned so much time understanding MOSFETs, but I still don't understand them. Will spend more time understanding and locking in.

12/11/2025 12 AM - Schematic work

Almost done with the schematic, although there are still some studies to do for the MOSFETs and GPS, the rest are just labels. Finally, after all this pain. But this is only the beginning; there are more yet to come.

12/11/2025 11 PM - Assigning pins and a lot of learning

I assigned pins checking out whether a pin can be used for a specific feature or component. This took longer than I expected. I also had to learn about how transistor work as I needed it to operate my passive buzzer.

Then I had to study about PPS of GPS. Overall, there was quite a lot of learning but progress was less than I expected. Next, I will need to finish up the long delayed MOSFETs and get the rest of the pins sorted.

12/21/2025 - MOSFET rabbit hole and a lot of learning

Today I learned a lot from the KiCad forum, a guy named Emiel gave me some guidance on how current-limiting resistors actually work and how I should pick footprints for the components. These are no doubt going to be very important for this and future projects, as they are the basis of PCB designing with SMD components. Other than that, I also asked for help on MOSFETs, unfortunately he said he isn't sure about PMOS, and it's best for someone else to answer the question. So I tried making a few schematics that made sense in my mind and send it to the chat.

Attached below is how it currently looks, though I seriously think it's wrong because the gate pull-up isn't nearly enough to turn the MOSFET off, but if I connect it directly to the source, it might blow my MCU.

Just spent some time on it, IDK why it's so complicated after I implemented an extra NMOS to drive the gate, ensuring it turns off fully.

I am having a hard time finding someone who is good with MOSFET work to verify my current schematic.

12/25/2025 1 PM - Sourcing and schematic work

Although not 100% sure, I decided to take a leap of faith and just duplicate the pyro channel design I did earlier and finish up the schematic. I double checked the pinouts, and they match with those in STM32CubeMX, so that's all good.

Next, I started assigning footprints and did all of the resistor footprints and most of the rest. For capacitors, it is a bit more complicated as from earlier and recent reading for the buck converters, they might need bigger caps to reduce derating, which is frustrating and confusing. Other than that are inductors. I am unable to assign the footprint of an inductor, even after importing from LCSC. It seems you have to add the symbol and can't link the footprint to the generic symbol. I feel like I'm doing something wrong here so I'm checking with the Kicad community.

12/25/2025 4 PM - Starting PCB layout

Finally finished footprint assignment!

Now it's time to work on the PCB layout.

Took me a good few hours just to get things organised and laid out in order. (Other than pyro channels)

Next will be finishing the layout, setting size constraints, and then wiring them together.

I know it doesn't look like much, but trust me, those capacitors and resistors are tiny, and it's a PITA to get 50 resistor all put in one place to go to where they are supposed to be.

12/25/2025 5 PM - More layout work

Looks like we're getting somewhere!

By setting the board constraint, I was able to start packing things onto the board. And so far, it's looking good and seems like everything will be able to fit!

Next will be finishing the annoying pyro channels and start routing.

12/25/2025 7 PM - Layout done

Took me a bit longer than expected, but wow, is this beautiful.

I do have some extra space, and I think what I'll do is add a few places so I can have more sensors, like a force transducer or a thermistor, in case I need them. I am also thinking of putting maybe an SPI port so that if needed, I can have two flight computers for a single mission.

12/25/2025 8 PM - Adding thermistor port and making the board smaller

I added two thermistor ports and two JST connectors that provide power in case I need them in the future. Other than that, I also made the board smaller; now it sits at 65mm*110mm, which is pretty good in my opinion.

12/26/2025 1 AM - Routing

Holy freaking air ball. I need some rest.

Took me quite some time struggling to realise I need a 4-layer board, 6 layers would've been even nicer but I'm concerned about cost. So 4 will do, for now at least.

12/26/2025 10 PM - Routed the PCB

Let's freaking go! Finally got all the routing done.

I spent some time assigning the 3D models so it looks like the actual thing:

Apparently, I was using a lot of via in pad, and it's very expensive, so I had to change all of those. In the KiCad community, I was also getting a lot of feedback things like ground plane and EMI.

12/27/2025 3 PM - Ground plane fixing and cost evaluation

After evaluating my board, it seems that not having a continuous ground plane might be an issue. I tried compensating by having three ground pours, of which two are sandwiched in the middle with little traces. This makes it more like a typical SGGS board.

Putting it into JLCPCB, the cost came out to 335 USD, and after playing around with it, it seems I have to keep everything on one side to reduce the cost to around 250USD. which means more modification is required. But I'm really busy, so I have to put this project on hold. Hopefully, I can finish it by the deadline.

12/27/2025 5 PM - Modifications to layout

To reduce cost, I have put everything on the same side, which should save setup and stencil fees. After checking the cost sits at 285 with bare minimum specs.

12/29/2025 - Researched and changed MOSFETs for Pyro

Much fewer parts!

I swapped out the original pyros for a simpler, more efficient design uitilising less parts, which should reduce cost and also increase reliability. The way I got here was by studying the open source design of the telemetrum flight computer and noticing this particular model of MOSFET is used, although I'm not gonna use a comparator as they did for simplicity and cost reduction, I have physical arming to mitigate accidental firing.

By using something tried and tested, we reduce the chance of failure and can move step by step, as reliability and safety are very important in rocketry.

12/30/2025 12 AM - Routing and fixing schematic for pyro channels

I worked until 2 AM last night and still couldn't get it done, so I went to bed. It turns out I made some mistakes in the schematic yesterday, I managed to fix them and finished routing. Look how much better it looks now! Less clutter and fewer components will reduce the cost, albeit slightly, but also come with other benefits, such as reliability and simplicity.

12/30/2025 1 AM - PCB Fix

If you remember from the earlier journal, I talked about ground plane concerns; what I meant by that was that I did not have a continuous ground plane/pour. However, I have just come up with a solution which should probably work, this is by reducing the number of routes on this specific plane as much as possible and by using stitching vias. Not only that, but 3 out of the 4 layers have ground poursso this should pretty much resolve the issue of not having a continuous ground plane and EMI effect.

12/30/2025 2 AM - Blender Rendering Attempt

I keep seeing people render cool images for their PCB so I decided to install the extensions and learn how to do so. I am so rusty with blender so it took quite some time and a lot of help from a friend. Here's how it looks for now, no idea why it's so fuzzy but I'm sure I'll figure out in the future.

It's 2 AM yet again, so I'll get some rest now.

12/30/2025 9 PM - Coding + Routing

Finally got some time to write a really simple logging code for the flight computer, just for review's sake, because realistically we will need the board to code and troubleshoot.

I also reduced one more route from the "ground plane", so it only has 4 routes now. I've been trying different combos of routes to get past this, but it seems like they will add a lot of vias and have some pretty awkward routes. I think the stitching via strategy is still good for now, but I'll try to make it a complete plane if I can.

Furthermore, I made a more specific rule area for the antenna as well as changed the traces as GND pours help with RF signals. (reduces EMI)

12/30/2025 10 PM - DRC check and routing help from KiCad community

I was getting one unrouted net but can't find where it is, also having some clearance issues in DRC. So I resolved all the DRC errors other than those that are footprint specific which there's nothing I can do. (as well as some thermal relief ones those are not really an issue)

Turns out the unrouted net was between two pours and can just be resolved with a via. Following a community member's advice I added a bunch of vias randomly throughout the board so that there are more return paths for GND as some big areas are only connected with one via.

1/6/2026 12 AM - Added BOM and PCB fixes

I created the BOM and added links where possible. Not quite robust, but I do have like 10 exams coming up in a day, so I don't have much time.

I also added more via around the edges of the board as suggested by mpk.

If any of you remember that I have changed the MOSFET design to simplify things and save cost, I can let you know that it saved less than 1 dollar but a win is a win ;>

1/6/2026 9 AM - Update BOM

I just checked Digikey for lead time, and it's so long... I don't have any experience with soldering SMDs, that's why I wanted to order a PCBA. But I guess in this case, I'll just have to add a tube of solder paste and these components into my BOM and then redeem a hot plate with tickets, which I have none atm;w;

I also forgot to add in some parts that I chose to solder by hand to reduce cost.

1/20/2026 - Clarification to reviewer

Due to the nature of the ICM-45686 footprint, I am unable to hand-solder it like others. (The pads are underneath the package)

My current plan is to claim a heat plate from the store, then solder the IMU to it, as the pads are directly under the IMU, and I can't use a soldering iron to solder. I do not have much experience with SMD, that's why I'm ordering a PCBA. I would've ordered a full PCBA if I could, but because the IMU is out of stock, I am forced to do this.

Hope you understand, thanks!

2/5/2026 - Naming the project and silkscreen + reponding to reviewer

I've decided to name this flight computer Aries, as it is the first zodiac sign known for being courageous and ambitious. And I feel like that's what this project is about. I always wanted to use SMD and get features like this on a flight computer, but I always told myself it's too complicated. Now it's finally coming to life!

Will work on silkscreen next. Current idea is an Aries constellation pattern and something cool behind, since it's quite empty.

To reviewer:
For PCB, I think it's set to a minimum already, which is 2 PCBA and 5 PCB.

And shipping can be seen in the cart screenshot, but I'll add it into BOM if that helps. (seems like I need to reupload the order and recalculate)

2/8/2026 1 AM - Updates to the Front Silkscreen and some minor adjustments

I added the name of the board as well as a small graphic of the Aries constellation. I also added JST-XH for the 4 servos in case I need to extend it. Also did a quick render.

Moreover, I checked the price for PCB and updated the BOM for final review.

IDK why, but whenever I do the JLCPCB price check I always get it wrong a few times and end up wasting an hour on it.

Alright for now, it seems that the ICM45686 IMU is back in stock!!! so I'll need a quick approval before it goes out of stock again budget request will be the same, but I'll keep the raw IMU price there so you can see in case it ever goes out of stock.

(To reviewer: simple - 35 out of total and take one LCSC shipping and the two ICM45686 as cost in case it goes out of stock. for now it's in stock with 8 available. All of which available in BOM)

P/S: will work on the backside silkscreen whilst waiting for review.

2/8/2026 5 PM - Blender help

Got a lot of help from a friend teaching me how to manoeuvre in Blender and ended up with a render that I'm quite happy with! Learnt about shading and compositing to make things sharper, brighter, more exposed, etc.

Would like this to be my new banner but rn I can't cuz it's under review.

2/10/2026 - Clarification

Most expensive components:

These are total prices so no one component is more than 17 USD the rest are below 5 dollars with most being under a dollar as they are passive.

BOM of PCBA parts available here: (more in-depth on Slack)
https://github.com/kian-cheng/Advanced-Flight-Computer/blob/main/Hardware/production/bom.csv

3/1/2026 - Clarification to reviewer

Been quite a long wait, but I'll just put this here so that the reviewer doesn't miss it. It's also in Slack #blueprint-draft. Also, my IMU and parts are starting to rise in price.