Whether you are multiplying 1200 by 4, or 120 by 4, it is all represented as 1.2 x 4 on a slide rule. As the result is read out to be 4.8, where to place the decimal point?
Except for simple cases that you can estimate the results, it’s often difficult, or counter-intuitive to place the decimal point correctly for slide rule calculations.
In a hand calculation, n x m = 3.141 x 5.926 = zzzzzzzz, we start with a 4-digit number multiplying another 4-digit number. Then at the end of the multiplication, we try to place the decimal point.
I just saw a fun card magic trick by MLT Magic Tricks (on youtube). The trick itself was well executed and explained, but I still screwed up on my first try. And there are other people in the comment area experiencing issues as well.
First of all, my problem was that I used a 54-card deck, instead of a 52-card deck.
If that doesn’t quickly fix your issue, please allow me to explain the math behind the magic trick. …
One interesting form of string art starts from a circular or square canvas with nails on its perimeter. The strings/yarns are only wound around nails on the perimeter. The repeated straight lines can sketch out a portrait nicely(link to Petros Vrellis).
Here are bunch of other examples on Hackaday.
This is a cool algorithm implemented in Processing language (github link). Basically, you input an image, the program will output a list of sequential nail locations that the string needs to traverse, which is also called a weaving pattern.
(Additional Hough transformation may reduce the thread count, to be explored https://rosettacode.org/wiki/Hough_transform#J)
Electronic modules are getting smaller and ever more powerful. For example, modern MCUs are packed with capabilities, and at a very low cost. In addition, making a custom PCB (via JLCPCB for example) is often more time/cost efficient than breadboarding. These are all good news for DIY electronics fans.
I still remember the old times when I had to save lunch money to buy a single transistor, and then went sad for days after unknowingly damaged it, probably due to ESD.
Although you can now DIY electronics cheap and easy, one thing still remains costly. …
In Electronics PCB world,
0.1" is a standard spacing or pitch for ICs, connectors etc. for many, many years.
In LEGO world,
8mm pitch and
4.8mm diameter hole is its gold standard. (LEGO brick dimension drawing link)
Here is a useful trick to unite the two together with LEDs.
You simply spread several
0.1" through-hole LEDs (5mm OD LED on LCSC) on your PCB, with
8mm spacing center-to-center.
Once you solder the through-hole LEDs, the PCB can plug into any LEGO piece or technic beam very tightly.
After that, it’s only limited by your LEGO imaginations to e.g. …
In an effort to make a minimalist digital clock, I was looking for an easy to construct enclosure for all the small electronics modules, such as AAA battery, display modules, and MCU boards etc.
Yes, 3D printing is still outside of my skill set.
Interestingly, I’ve seen origami paper boxes, as introduced in this youtube video (link), which can be traced back to Tomoko Fuse’s earlier books (Amazon link). In her recent 2018 book, she used the term “modular origami”, which roughly speaking is a technique of stacking cleverly multiple identical origami pieces, called modules, to form a complex 3D…
HT16K33 (datasheet link) is very similar to TM1637. It is also widely used in display modules such as Adafruit 8x8 LED matrix (schematic link and LED hardware link), or this 4-letter pHAT module (link).
Again Adafruit provides nice Arduino/Python language support (link) for its modules and HT16K33. But we can certainly experiment on HT16K33 with a FT232H module as well.
Comparing with TM1637, in terms of its I2C interface and digital commands, HT16K33 has the two main differences: the opcodes and that its I2C byte order is MSB first. …
Rust embedded software tooling is always an interesting alternative to MCU vendor specific IDE software, such as those provided by TI, Microchip, STM32, Nordic, etc.
With the advent of probe-rs and subsequently probe-run software, Rust for embedded coding can take a different path from
openocd + gdb, and it is getting more convenient.
There are already lots of cortex-m quick-start guides, especially of STM32F103 “Blue Pill”. Unfortunately, Rust library such as
stm32f1xx_hal is changing fast. For example, the
Timer now is initiated differently from the past.
Here is my latest try get it up and running:
In my earlier article two years ago, I mentioned Bob Peace’s comment of Dr. Middlebrook’s “differential amplifier” book. (link)
Here is my answer to four of my own questions. (Sorry for the long delay in between.)
Electrical circuit analysis for human designer is a bit like solving Sudoku puzzle.
In Sudoku, there is the local rule or constraint: no duplicates in each 3x3 square. Then there is the global constraint: rows, and columns across multiple squares can not have duplicate numbers. You have to satisfy both the local and global constraints.
In circuits, the local constraint is the device/component f(I,V)…
A “low-entropy” equation that depicts the circuit behavior can help analog designers work backwards, finding values of the circuit elements in order to meet the specification.
Let’s start from the output impedance of a simple linear circuit.
A common procedure of determining the output impedance is by: shorting input to ground, and look into the output port, to “see” the output impedance. In this simple case (Fig. 1), we can “see” R₁ ∥ R₂
When the circuit becomes more complicated, it is difficult to “see” the output impedance directly. One way is to inject a current and calculate the resulting…