Not legally an Engineer Yet, Call me an EIT
A really great option when you’re looking for something cheap and quick that can be programmed at a plethora of levels. The options that are primarily supported are Micropython, C, and C++; however, I’ve seen work with rust on it which is always good fun. The dual M0 cores make it suitable for light levels of multi-tasking, especially when paired with a good RTOS to handle the timing itself. The PIO cores can allow for simple state machine like behavior which allows for easier responses at the system clock speed. Generating things like VGA signals, or even DVI signals are then possible.
Part of the reason I like recommending this chip to students is the USB programming which eliminates the need for any sort of external programmer. This makes it ideal for students who are just getting into embedded systems, or for projects that need to operate on shoe string budgets without access to more experienced individuals.
I saw a link saying that this chip will be dropping sometime soon. It’s a combination of the RP2040 with the addition of Flash and Ethernet. Making it the ideal single chip solution to create low volume projects which connect up to ethernet. It’s definitely something I’ll be checking out as soon as I get the chance and I’ve signed up to be notified once they’re available for purchase.
Hype, Hype, Hype, Hype. This is an upgrade from the earlier RP2040 chip and has packages that act as drop in replacements, simply better in every spec.
I’m quite hyped for these chips to become available on platforms like JLCPCB so that I can start including them in my designs. Although I’ll have to use up my existing stock of RP2040 chips before I buy them in bulk I’ve got a couple samples from Raspberry Pi which I’ll be integrating into some thesis related boards shortly.
While the larger amount of ram, faster base clock, and lower power consumption are nothing to turn ones nose up at. I’m quite excited to see what comes of the Hazard3 Risc V cores which are included on the processor. These will likly not do much for most hobbiests, but I hope represents a potential future IC which removes the ARM cores and goes fully in on Risc V.
So I used to use AVR microcontrollers fairly extensively in my various small projects. Coming in DIP packages back when I soldered all my projects myself I ended up quite liking these ATTiny chips for their greater number of IO as I seemeed to always find myself IO limited in most designs. I’ve used a lot of the AVR series at this point with Micros over 100 pins; however, during the early days when TQFP packages made me nevous there was definitely a lot of cases where the DIP 20 package costing ~1/2 that of the lightest ATMega chips was an ideal solution for me.
This seems like a good all in one solution. Although coming in at the eye watering cost of $35 per chip! The propeller microcontrollers from parallax are 8 core microcontrollers with enough peripherals and connectivity to choke a horse. I hope I get the excuse to embed one of these into a project at some point. The datasheet quite interestingly has its revision history showing that the chip has undergone multiple revisions since its original release in 2018, with the most recent REV C release coming in 2020 helping with ADC performance marginally. Several of its features seem promising for various applications, although the lack of sufficient details somewhat stops me from pulling the trigger on certain solutions. For instance the Smart I/O pins brag 64 identical pins, but I somewhat have trouble believing that the Triangle/ SawTooth/ SMPS PWM output can be activated differently for every output.
Well It’s official I’ve begun to dig into embedded linux and microprocesors. Although it’s hardly a surprise. Building my own motherboard has been on my must do for literally years at this point. Recently I’ve been using the F1C200S a chip which will likely make its way into my thesis in the form of both a guide on doing board up on designs containing it, and as a nice chip for creating an integrated measurement platform. I’ve since been looking at the other AllWinner Microprocessors on the market and I must say that while they have truely pitiful documentation they offer a wide range of options suitable for a lot of possible applications.
Pin compatible Sister chips they offer 32MiB and 64MiB of ram respectively. Suitable for running simple lightweight linux systems in low-stakes applications. The 4 different required voltage rails can apparently be reduced to 2 by merging the 2.5V and 2.8V rails into a singular 2.7V rail (Don’t quote me on that one). However the total support ciruictry can likely be squeezed down to an area no bigger than the chip itself. I originally purchased 30 of these chips for a project, and proceeded to sit on them for well over a year before putting together the board to gain familiarity. I’ll likely move to a second board soon with substantially improved features. Check out my repo with the I2Sense lab on github for what should be a simple, fully functional design using the chip.
Interestingly there are apparently a series of F1C500S and so on which are simple rebadged copies that advertise their linux capabilities; however, I haven’t been able to see them up for sale on any of my normal sources.
I’m quite the Xilinx Fanboy when it comes to FPGAs. I still have yet to integrate one into my own board design; however, that day is certainly coming at some point. I originally learned FPGA programming with the Digilent Basys 3 board containing a Artix 7 FPGA on board. I then got to work with the Zedboard during a chip up and explore the joys of having both ARM cores and FPGA fabric on the same device.
I just put in the request for some sample parts of this 1K LUT FPGA device. If the parts get approved I’ll do up a test board for them and see what I can acomplish. The contain a one time programmable non volatile memory for their configuartion. This means they act like very simple ASICs, with a per part cost presumably under a couple of dollars, this makes them a lifesaver where they’re required. The 3x3mm package size is ideal for tight spaces where larger custom logic woudln’t be doable. Hopefully I can come up with a good project to create with this part to validate a workflow with it once it arrives.
Discrete parts rarely are what one would call interesting; however, every so often I have across a part or two that is intereting simply because it’s rare that you get to use it.
I got to use these parts for a couple of projects and they are certainly interesting as far as discrete parts go. They’re matched N-JFET transistors with some exceptionally low noise performance paramters. Down at 3nV/sqrt(Hz) at the 10Hz point. I needed some astonishingly good leakage and flicker noise parameters which is exactly what these transistors provided me. Now if only they didn’t cost quite so much at $13/part for the A variant (Best Performance) and $8/part for the C variant (Worse Performance), that being said the specifications for the lower variants include higher-current testing which may be of value. These would make a great low noise analog frontends for a lot of different applications, especially given the GS breakdown voltage of -40V.
I really think I want to make a toaster from scratch, and honestly this would be the way I’d want to do it. The chip is relatively speaking fairly simple coming in a DIP 8 package, it is a great way to run a toaster. At $0.76 per chip at small quantities, it is a little questionable if one would be better off saving the passives and using a simple microcontroller to drive some heavy fets to acomplish the same purpose; however, the specific nature of this device likely makes it substantially more durable than it would be if it had to encorporate some form of complicated digital logic.
The chip is simple enough with reasonably functionality that It gives me serious 555 timer vibes having the potential to to acomplish the same level of versatile applications if given the chance.
A reasonable looking sensor for detecting lighning within a 40 kilometer radius. I’ve considered designing this sensor into a camera trigger system for doing better lightning photography. Triggering the camera whenever lightning is detected. Although the sensing would be difficult it would make a great method of capturing awesome photos of lightning strikes assuming that the system can detect fast enough. At $14.43 per chip it’s not an insignificant expense; however, the reference design, and built in lightning detection algorithm makes it well worth it.