
Sign up to save your podcasts
Or


(Michai is too lazy to make shownotes now, check again later. :-)
Hi, and an extremely good 2021 to all of you.
We start off this brand new year talking about some totally unrelated and slightly more related new year's resolutions, a miniscule car fix, and glueing with a 3D printing pen. All in not much more detail than stated here.
Continuing on last episode's VGA generator board, I added a DMX512 interface as suggested by some people - thank you for the idea. I found Q Light Controller Plus (qlcplus) a nice and so far solid piece of software for controlling DMX devices from a laptop + USB/RS485 converter cable.
A potential issue (or at least in my head) when making interlinked devices like these, is potentially different ground-potentials depending on which power outlets you use to power networked your device. (See what I did there?) We talk a bit about how to guesstimate the possible magnitude of this difference.
Just for fun I made an xmas LED star-shaped decoration using an existing DIY LED driver board. Reuse everything, recycle success. Also, made a modular transparant acrylic home LED decoration: shine light into the edge of an engraved piece of acrylic. Of course there was failure, this time in the form of hygroscopic LEDs + laziness + oven = damage.
Finally, I made a small power switch PCB, to be placed in between USB power source and USB device: automatically switch the device off when the environment goes dark.
And remember: if YOU want to make this podcast suck less, please send comments, rants, complaints, threats and/or praise. :-)
Relevant links:
Pics of things talked about:
DMX512 VGA generator boards.
Engraved acrylic LED decoration PCB in action.
...in its base, open...
...and closed up.
LDR USB power switch board...
...and schematic.
Welcome to a pretty boring episode - just as you're used to by now!
Prepare yourself for epic or less epic talk about a mains test jig that proved very useful, some awesome Probe Master DMM probes well worth the money, and both me and occasional cohost Domen fixing household stuff, and the subtle psychology of starting a repair.
Homebrew-wise, the only thing I made was a smooth green LED under-plant decoration. For fun, I visited an informal light-art hangout and talked about adding DMX512 to my VGA generator board.
Relevant links:
Pics of things discussed:
The test-jig from episode 41.
Clumsily made LED strip using paper, glue and cursewords...
...and a simple PWM/MOSFET driver to control 3 such strips.
Let there be Light, and let there be a Plant!
Domen's epic Fireplace Fix: worn out film cap.
The thrilling tale of 4 months of doing pretty much nothing interesting is coming to a podcast near you today!
After hearing about the hilarious exploding electrolytic, we continue with some lame or slightly-less-lame tools one can just buy off-the-shelf: using H-field probes for injecting magnetic fields into a PCB, the Aim-TTi I-prober 520 for measuring magnetic fields and current, and a simple but superhandy USB power meter.
(I guess it's just a matter of time before all this deteriorates into an unboxing-podcast...)
Various IRL events were canceled, so a YT livestream from a friend of mine got me thinking about 3D printing again. I settled on using a commercial 3D printing-service instead of DIY, and that suits me fine for now. This episode lists some criteria that may help you trying to decide between using a printing-service or getting your own printer, and if so, which printing-method suits you best.
Lo and behold, I actually made something as well: a LED planet decoration thing using 3D printed "planet" and acrylic rings around it, that light up when LEDs shine onto them. For simulating visual effects, I tried and liked the Cairo 2D graphics library (for C).
Links in order of appearance:
And some fluffy pics:
QCAD: 2D cross section of planet decoration's "planet sphere". The shape itself, and dimensions therein ("A", "B" etc.) can be read and used by OpenSCAD (3D) for extrusion. Pretty cool.
OpenSCAD: extruded 3D shape, resulting from rotating the 2D cross section 360 degrees.
A transparent engraved disk of acrylic fits between the 2 halves of the sphere.
Planet decoration in action.
(laptop shown, to get an impression of the size of the decoration)
A piece of LEDstrip sitting snugly inside the bottom half of the planet-sphere in the center, shining outwards into the transparent acrylic sheet.
LED driver, old board with flaws...
...and new version, with fewer flaws.
You DESERVE one of these if you make USB-powered crap.
Shoot your zapper at uranus! A dirty mind is a joy forever, and so is having a NES.
In this time of global turmoil, the CBA Podcast once again proves to be your one-stop source of distraction and meh-grade information.
We start off by a quick list of how to keep busy in a sort of useful way, followed by a short talk about a 230 Vac test-jig and designing for test-jig repairability.
Already mentioned in a previous episode, I finally got around to building a digital driver box consisting of push/pull FET and open-drain drivers for a low-voltage digital input-signal. And that's not the only thing that got finished: I also got around to make mechanics and software for a spiky icicle LED ball decoration thing, which worked out quite well.
A fuzzy talk about getting to know your electronics equipment more intimately concludes this episode. If you are looking for useful things to do there, perhaps this is worth a shot.
Relevant links:
Some pics of the things mentioned in this episode:
230 Vac test-jig, using DIN-rail components and a big block of MDF wood
digital driver/amplifier box, innards
LED icicle ball, innards. The small PCB is an AVR micro and 2 16-channel constant current LED-drivers.
painted with acrylic paint and spraypaint. Should probably just have 3D-printed this in the first place...
LED icicle ball, full brightness, sitting on desk
Welcome to the first and so far best episode of 2020, the third year of this podcast. Thanks for sticking around and making this happen, or alternatively, thanks for just tuning in - hope you like it.
After a quick word on hot glue applied to mains wires, yours truly talks about being surprised by a SOT23 500V constant current driver (BPS BP9918C) in an LED lightbulb.
Super-exciting, all of this.
A cheap RGB LED strip decoration revealed a 5-24V LED driver (Greeled Electronic Ltd SM16703P) as higher voltage drop-in replacement for e.g. WS2811.
Apparently not all christmas magic is done with those newfangled LED things - oldskool still has its place in an incandescent light decoration with smooooth ramp up/down effect.
Perhaps being of more practical use, we talk about how the body converts food energy into mechanical energy, and show that exercising for the sake of losing weight is not the best approach. #spoiler
Projects I actually worked on or even - *gasp* - completed, are a double pole mains switch, an ongoing hack including driver board for an LED icicle decoration, and a big clunky resistive load that also turns out to be inductive. (Turns. Windings. Inductive. Get it?)
Links to relevant stuff (links to datasheets are probably volatile):
Pic of hot glue ghetto-enclosure:
BPS BP9918C example circuit:
Double pole mains switch:
32-channel LED driver board (30x30 mm):
Resistive load:
4.5 months is still not enough to make you forget this podcast, is it? (And also, sorry about the long gap.)
Another mixed bag as usual. 2 meetups with the CBA Quality Team for some Quality Time.
Eindhoven Hackerspace Hackalot moved to a new location, and I put some LED decorative stuff there, making use of a light intensity curve for a better visual result.
Furthermore, a trip down memory lane with an old elevator mock-up project, tests with 2 potting-materials, a hot glue meltdown, and taking a look inside an older anti-theft clothes tag.
Inside the anti-theft tag:
At the bottom right is the mechanical lock. A pin is locked inside the brass-coloured cylinder by metal balls. Applying magnetic force will unlock it (although I didn't try here).
No RFID with this one, just an LC tank circuit. I think these were similar tags they already used when I was a tiny kid.
Hi there! Thanks for tuning in (or reading this text and then deciding not to tune in at all). Another episode with a mixed bag of semi-interesting topics.
Thanks for the feedback - the inter-segment jingle will be a bit less loud from now on. Poor ears.
We start off by counting pass & fail on this year's new year's resolutions, followed by mentioning a small inline PCB for driving LED-strips and so on. The PCB was used for 2 light effects so far - links to videos of both are given below.
Ran into a "gotcha" when dealing with LED-effects: intelligent LEDs stuck in "full brightness" mode, causing the supply to shit itself and the system to become unresponsive. (Quickfix: small delay between boot and start-of-effect at power-on.)
For the LED-effects, I measured rand() execution times (avr-gcc, avr-libc, ATmega168 @ 8 MHz) and was pleasantly surprised by its speed and predictability. Will use again.
Finally, some fuzzy guesstimates / measurements about what kind of voltage drop to expect when running current through long cables.
Relevant links:
Snapshot of my inline PCB:
You thought that we finally learned from our mistakes and abandoned this podcast, and you thought wrong. Once again it's time to lower your standards, and enjoy a new episode of the CBA Podcast.
We start with some PCB failures and flaws, followed by a very brief review of the Hackalot lasercutter & vinylcutter workshop. We attended T-DOSE, which is a Dutch open source event hosted in Eindhoven.
Then for something slightly out of the ordinary, we talk about back-of-envelope calculations about whether a climbing-rope can actually break if I fall. (Spoiler alert: it cannot.)
Some clarifications and errata about this climbing-rope segment:
To calculate maximum force on the rope, I used the harmonic oscillator formula given here , where the actual force is calculated from fall factor, mass of object, but also the rope's rated impact force. (I forgot to mention that).
Also, I made an error: I mentioned a maximum actual force of approximately 85% the rope's rated impact force, but this is in fact less than 70% - nice.
For those special people who like this: standard EN 892 deals with dynamic ropes like the one discussed, and EN12277 deals with harnesses ("belts"). My harness is tested with a sustained load of 15 kN. The type of carabiner I use is rated for 27 kN load.
Finally, some relevant links:
After more than 1 month of laziness, let's kickstart this episode with some failures and apologies. Next up is a summary of Hackalot's part in the Open Hackerspace Day 2019, and a teardown of a cheap and somewhat boring Aldi 230 Vac LED lamp. Finally, some very-voodoo and somewhat-less-voodoo methods of measuring the percentage of bodyfat in humans are briefly discussed.
Links to topics discussed:
Here's the front and back (or back and front) of the 230 Vac Aldi LED-bulb's PCB:
...and here's what a scraped-off LED looks like under the hood:
You saw it here first, folks.
From the publisher's feed