STM32F042 USB-C Module — Schematic Review Before PCB Layout (2-Layer)
https://redd.it/1pzeyvx
@r_embedded

STM32U5 OctoSPI Quad Mode (IO4–IO7) + W25Q128 works, but XIP memory-mapped mode doesn’t

I’m using STM32U5 OctoSPI with Winbond W25Q128 QSPI NOR.

Setup:
OctoSPI in Quad mode
IO0–IO3 remapped to IO4–IO7
Indirect mode works fine:
JEDEC ID read OK
Read / write / erase OK

Problem:
Memory-mapped mode (XIP) doesn’t work
Invalid data / faults when trying to execute or read via mapped address

Looking for:
How to properly configure memory-mapped mode on STM32U5?

Any tutorials, app notes, or example projects
Any known gotchas when using Quad mode on IO4–IO7 for XIP.

Is XIP officially supported in this configuration, and are there any special settings (dummy cycles, opcode, cache/MPU config, etc.) I should be aware of?

Thanks!

https://redd.it/1pzj852
@r_embedded

where/how do you get small quantities of chips for new projects without shipping being the majority cost?

just that, any little project I look at doing now, I start to go , why bother, when amazon has the entire thing built for much cheaper. and the chips for a custom board, well then that's a different more expensive path.

I miss the cheap digikey days, with samples..or even free shipping aliexpress. I'm not one to plan for bulk orders

https://redd.it/1pzn375
@r_embedded

Very Confused about this module

Guys I recently i bought this ADC the datasheet of the AD7606 of Analog and the ADC module is different which makes me very confused , I surfed throught i only find the parallel connection only. If some used this module help me through it ,if you find this module schematic or any tutorial or you already used help out guys.

https://redd.it/1pzod7m
@r_embedded

Designed and assembled a camera controller board based on ESP32-P4 with PoE support (50x60mm). Also, optional edge TPU support. How does it look?



Note: I have to correct those solder blobs in the ESP32-P4 chip.

Note: Will correct those solder blobs in the ESP32-P4 chip.

https://redd.it/1pzw666
@r_embedded

Looking for a tool to measure overall consumption, inrush & peak currents and multiple analog voltages

Hey everyone,

I’m working on embedded/smart products with lots of on/off loads, soft starting, and switching rails, and I’m looking for a measurement tool that can help me with a few key things:

Measure overall power consumption (voltage + current)
High enough sample rate to capture inrush and peak currents accurately
Multiple (3+) analog inputs so I can monitor several voltage rails ramping up at the same time
Synchronous data (so voltage and current are aligned on the same timeline)

I’ve come across the Joulescope JS220, which is great, excellent current accuracy and sampling rate, but it only has one analog voltage input, which isn’t enough for my use case.

Do you have any other measurement tools you would recommend that are similar to Joulescope JS220 but with more analog input?


Thanks!

https://redd.it/1pzvkgi
@r_embedded

Foreign Computer engineer

Hi everyone,

I am moving soon to the US and I will be a green card holder so I won’t be needing a work visa. I have a degree in computer engineering from the Middle East and was wondering how hard it is to find a job at the US, if possible at all?

what process do foreign engineers follow in order to get a job?

https://redd.it/1pzy0yx
@r_embedded

SoM for wearable applications
https://redd.it/1q00oof
@r_embedded

Embedded Languages

What are all the programming languages that we can use to program a Embedded systems other than C/C++ ? I know Cpp and PY, did you use any other languages than C/C++ ?

https://redd.it/1q03hj9
@r_embedded

RPi 4 (4GB) edge face recognition (RTSP Hikvision, C++ + NCNN RetinaFace+ArcFace) @720p, sustainable for 24/7 retail deployments?

Hi everyone. I’m architecting a distributed security grid for a client with 30+ retail locations. Current edge stack is Raspberry Pi 4 (4GB) processing RTSP streams from Hikvision cameras using C++ and NCNN (RetinaFace + ArcFace).

We run fully on-edge (no cloud inference) for privacy/bandwidth reasons. I’ve already optimized the pipeline with:

Frame skipping
Motion gate (background subtraction) to reduce inference load

However, at 720p, we’re pushing CPU to its limits while trying to keep end-to-end latency < 500ms.

# Question for senior engineers

In your experience, is the RPi 4 hardware ceiling simply too low for a robust commercial 24/7 deployment with distinct face recognition?

Should we migrate to Jetson Nano/Orin for the GPU advantage?
Or is a highly optimized CPU-only NCNN pipeline on RPi 4 actually sustainable long-term (thermal stability, throttling, memory pressure, reliability over months, etc.)?

Important constraint / budget reality: moving to Jetson Nano/Orin significantly increases BOM cost, and that may make the project non-viable. So if there’s a path to make Pi 4 work reliably, we want to push that route as far as it can reasonably go.

Looking for real-world feedback on long-term stability and practical hardware limits.

https://redd.it/1q050vd
@r_embedded

How to implement my own custom sizeof operator in C ?

Earlier I was interviewing for Embedded Software Engineer position and he asked me this question and I still can't figure out how to do it since sizeof is a macro and not some C function and compiler calculates it at compile time.


https://redd.it/1q08hvj
@r_embedded

How do I use AI to ramp-up to a new SDK without engaging in a ton of bad practices?

I'm still a student and I recently got my hands on an nRF52 and am trying to learn the nRF connect SDK. I bought it to miniaturize a project I'd made before, and I've never used the SDK for this chip so I often use Gemini to explain things like say, creating a BLE service, or understanding how to dynamically register a service for this SDK. I'm also working in C which I haven't done in a minute.


I have no idea if what it recommends is considered 'good practice', and I don't know how to check. I seriously doubt anyone cares cause no employer is gonna read my github + this is just about the least impressive thing on my resu me but I don't want to teach myself rubbish yk?

https://redd.it/1q08vyk
@r_embedded

How do I learn nRF Connect SDK and Zephyr?

I've started a project working on a custom board with the nRF52820 chip. I'm developing using nRF Connect SDK because I'll need bluetooth as well, and it supposedly has good sipport. It's somewhat a pain, but I'm working through it.

I'm looking for some help with learning more about Zephyr and nRF Connect. I'm currently going through the devacademy Nordic provides, but if anyone has any additional resources, it would be much appreciated.

Thank you!

https://redd.it/1q09pb4
@r_embedded

How to install linux on an Allwinner A31 tablet?

How to install linux on an Allwinner A31 tablet?

https://redd.it/1q0blf1
@r_embedded

I have ~3 YOE and my major skills are Shell noscripting, linux, yocto framework and C. I want to switch companies using above skills and if need learn additional skills and make good use of it. Please guide me how to start preparing for job switching and searching.

Hello everyone,

I have almost 3 years of experience in a product based company. I work in a security operations team, majorly on yocto framework, shell noscripting, linux and C.

Now that I'm accumulating 3 years of experience, I would like to find better opportunities and switch from current workplace.

Please guide me how to prepare for interviews, if I need to learn any new skills or certifications. I'm learning data structures and algorithms, system design. Will that help me in looking for any job opportunities?

Please guide me.

Thank you and wishing everyone a happy new year.

https://redd.it/1q0cp6w
@r_embedded

Simple & cheap Eval-Board (with AI Accelerator Hardware) to learn Embedded Linux (Yocto)

I am looking for a simple and cheap evaluation board to learn Embedded Linux (Yocto) where the CPU has accelerator hardware for neural networks. It would be beneficial if the board also has a few sensors which I can use to feed a dummy neural network. The goal is more or less to show this on my GitHub and improve my market value as an embedded software developer.

https://redd.it/1q0ci9s
@r_embedded

Global synchronization time implementation

Hello everyone,


I'm working on a project, and was able to find an algorithm somewhere online where it synchronizes multiple devices over UDP. After implementing this algorithm on C using a linux microcontroller, and obtained the desired global clock(stored as a variable), I didn't know how to proceed with this. Is there like a command where I can replace the microcontrollers clock with this clock or is there something safer where I can have a virtual clock or something like that?

I'm new to this and would appreciate your help.



Thanks a lot :)



https://redd.it/1q0f0n6
@r_embedded

HELP ME choose a Logic Analyser

Hello All,
I am currently a Embedded Software Dev Guy who will be completeing his bachelors in another 6 months . Current located in Bangalore , India .

An intern at a robotics company for past few months . I currently have a 24MHz generic logic analyser (the 350 INR or $4 one) . They perform bad at their job whenever i try to debug something or work with them . Half of the times the driver issues pop up .

Currently i was exploring SPI in very detail and all my attempts to correct bit shifting in MOSI line basically failed in multiple methods . All through that i was trying to see the actual bits through the line and i was pretty much running the SCLK at 1.6 MHz . Still my logic analyser was doing shit .

Now i was planning to get the Digilent Digital Discovery with a 15% student discount which would cost me around 28,000 INR or around $310 after all import duties and taxes.

I was also looking at the Saleae Logic 8 where i could avail a 50% student discount and get it for $250 and after taxes and import duties and delivery would be like $360 .

But i was more aligned towards the Digilent beecause of 800 Ms/s high sampling rate given the Saleae does at most 100 Ms/s but given the overall support and development environment of Saleae is better , i am actually confused .

My original budget though stands at 23000 INR or $250.



https://redd.it/1q0elio
@r_embedded

Embedded Linux firmware and software update strategy

I'm trying to implement firmware and software update for my board, but have some problems on how to approach it.

As a second stage bootloader (that gets loaded by the BootROM) I'm using BL2 from TF-a. Its purpose is to load FIP (Firmware Image Package) that contains BL31 (EL3 Runtime Services), BL32 (OPTEE), BL33 (U-BOOT). Then, once the u-boot is loaded it is supposed to load Linux kernel. I have requirements that I need to be able to update almost all components namely: FIP (thus BL31, BL32, BL33), kernel image (including bootnoscript) and rootfs.



TF-A supports something like PSA Firmware Update that allows to select FIP bank, just ordinary A/B scheme where there are FIP-A and FIP-B copies, and the selection happens through so called metadata. It also supports boot counter that allows to roll back when the boot counter reaches max boot attemp. Up to this point everyting is quite straightforward.



The things get more complicate when boot process reaches the u-boot stage. As I pointed out I need to have two copies of kernel image and boot noscript thus I need the following:

\- bootnoscript-a

\- fitImage-a (kernel image)

\- bootnoscript-b

\- fitImage-b (kernel image)

I am aware that I can embedd the bootnoscript in the fitImage however, my fitImage is stored on the disk in the encrypted form, so the bootnoscript is used to decrypt the kernel image before loading. I decided to follow A/B paths for each component thus the TF-A is the source of boot index.

The information about boot index selected by TF-A (A or B) is propagated to u-boot, so u-boot indeed has such information. The boot index can be then used to select bootnoscript-a/fitImage-a or bootnoscript-b/fitImage-b. However I'm not sure how to do this. I mean it's not technically difficult, it's rather matter of "how it should be done correctly". What's more I want to make use of Standard Boot instead of old distro boot. Looking at the Standard Boot methods there is e.g. bootmeth_noscript available however I don't see how this could work in my case. This bootmeth_noscript just simply looks for bootnoscript and if it finds it, then it loads it. As I said in my case I have bootnoscript-a/b, not just single bootnoscript. What's even worese I cannot make a separate boot partitions for boot-a and boot-b as the only paritioning scheme available on my SoC is MBR which allows to create up to 4 partitions. I already reached this limit (boot, root-a, root-b, data), so the only reasonable solution for me would be to put the A/B artifacts in subdirectories under /boot part.



To sum this up - the only sensible solution that comes to my mind is to create my own boot method that will use the TF-A provided boot index, then will scan /boot partition looking for e.g. boot-a/boot-b subdirectories (depending on the boot index) and the required blobs (bootnoscript, fitImage). However, to be honest I'm not sure if it's the right solution. Would be grateful for some suggestions. Other solution would be to creata shim noscript loaded by bootmeth_noscript that will just source another noscript under /boot/boot-a/b/bootscr.scr but this shim noscript would not be duplicated and risky to update.

https://redd.it/1q0exbg
@r_embedded

Interfacing Teltonika 1-Wire RFID with STM32: A deep dive into safety, protocol analysis, and UID extraction.
https://redd.it/1q0iopw
@r_embedded

I want to get into the IoT/Embedded fields but I don't know where to start or what beginner project to choose

Hello everyone and happy new year!

I recently got a Computer Science degree and I worked as a Python backend developer in the past. I'm now looking for a way to get into the embedded programming field, because I like the idea of writing code for smaller devices that interact with the physical world, and in particular I like the whole concept of IoT.

The big problem is that I have no idea where to start. I'm thinking of choosing a project to work on it and start learning but none of the beginner projects (such as a home weather station) I saw are interesting enough to make me want to start buying a Raspberry or any microcontroller and work.

My main interests are telecommunications (I even thought of signing up for a Master in Telecommunications Engineering) and coding, in particular I would like to work on something that uses LoRa or BLE. I was thinking of buying a raspberry to create a mobile home surveillance camera that I can move from my phone (I don't have access to a 3D printer so I would have to use lego pieces for the mobile part) but I don't even know if this is too advanced for a newbie like me.

Any kind of idea, advice or material is incredibly appreciated. Thanks in advance!

https://redd.it/1q0ibgh
@r_embedded