gtx tob 01/2015
gtx tob 01/2015
The illustrated Walkthrough of PCB Production shows how I do it in my electronics lab since 2007. I heard of & tried the toner method years ago – but for a long time was less successful in reliably producing nice quality PCBs. I tried photo paper, telephone book paper, color pages of magazines, dextrin covered paper, the various catalogs of electronic supply companies. I tried different temperatures of the hot iron, higher & lower pressure. I tried different board cleaning methods, different chemicals, until I could cut “the chemicals” down to the etching agent & maybe isopropyl alcohol (a cheap vodka would do as well). Of the etchants I tried two: Ferrum-III-chloride (FeCl3) & sodium persulphate (NaS2O8) but want to try etching with copper-II-chloride (CuCl2) as an e-friendly closed cycle alternative. Now here is my recipe at the moment:
My Website´s logo was inspired by a traffic sign I saw on the Belgian Autobahn some years ago. This little graphic adventure is some vectorgraphic work I did about signs for ice, frost & snow flakes & permutations thereof. If the “real” Belgian ice warning sign looks like I recall it, I could not verify despite devoting some effort to searching for it.
The long awaited reader reader for the CTM 2010 workshop, includes schematics & layouts of the shown circuits:
Here I made a quite useful table of SMD (SMT) component sizes:
Even when using SMD, from time to time we need to drill some holes for through-hole-connections. I searched the web to find a better way before I start electroplating. Instead of using little bits of wire I found a novel way, which at least for me is 10 times faster:
This is a short intro handout for successful circuitbending (auf deutsch) :
This is the handout from a workshop in xxxxx, Berlin, a VGA-monitor as wave display with new Arduino VGA-Sync code:
In summer 2009 I had a discussion with a friend of mine Andrzej Wróblewski aka Bohomaz, stencil artist, about vector based pattern generation. This inspired me to do some visual studies in spirographics. Here you find some full resolution pdf – especially the first is a bit heavy in pdf rendering – but hey, it is A3 and easily can be printed out in A0 (all are Creative Commons 3.0 BY-NC-SA licensed):
This is the Schematic & PCB artwork for an STK4241 2x120W PowerAmp:
This is the Schematic & PCB artwork for a tiny LM4895 1W SMD PowerAmp (only chip + four 1µF caps!):
Today Sergio from Mexico found an old documentation of a sequencer I made in 2007 for Piksel Festival. It contained several small errors, so here is the updated version of this nice 8-Step-Sequencer. It is simple but yet has some extra features. With a 150ohms resistor as current limiter. the sequencer can drive up to 20 milliamps to give pulsed electricity to filter circuits or toys.
Give it a try! Please NO commercialisation – especially of the gate circuit, which really works despite being so simple. Contact me if you need a custom extended version, there are several more astonishing tricks..
Here is a simple Parallel Port Programmer for Atmega Microcontrollers (Schematic & PCB):
A KiCad-Tutorial – KiCad is an open source software suite for electronic design automation (EDA) – designing schematics of electronic circuits and printed circuit boards (PCB). KiCad is developed by Jean-Pierre Charras, and features an integrated environment with schematic capture, bill of materials list, and PCB layout.:
A Triac Tester – Had to work on a dimmer pack recently, I used & want to share this:
A Bat Detector – This thing works using only one chip, still EXPERIMENTAL:
A Making your Life as an Artist – An eBook by Andrew Simonet, worth reading from about page 50 on:
Since I wrote “Secrets of SMD” it is now more an a year. At that time a soldering iron was my favourite soldering tool. In between I tried table grill ovens and a pizza skillet (=electric pizza pan) as reflow device. Success rate varied a lot with molten dark parts (=high absorption) & unsoldered silver ones (=low absorption), no possibility to interfere in the running process & generally quite dark PCBs (overheated boards).
Today I mainly use SMD soldered on a clothing iron, use a hot air gun for the back of double sided or very simple single sided PCBs & the soldering iron mainly for small corrections or soldering connectors.
This is an overview of my recent process of DIY production of PCBs. I hope it is not too sketchy, cause I do this write-up in the pauses of board making with exactly this process – I am quite new to it – it´s fresh!
See it is a starting point for your own fast prototype-production or just a way to share some ideas.
First we have the hot air gun, a gas burner with catalyst, you can get it for about 20 EUR. Its useful for any re-work (de-soldering & repair) of multi-pin devices. Be careful not to overheat parts.
Next comes a hot plate, which can also be an electric pizza pan I used before. A clothing iron has the advantage that you can also use it for the toner PCB method or heating etching solution, it´s flat & heats up very fast. Fix it to a table with a clamp in a horizontal position:
Next we need hypodermic needles for the application of solder paste. I prefer the black 22G size, sometimes 20G. (Updated with this nice table)
For breaking the needle I scratch the surface with an electronic cutter, rotating the needle several times:
Then it is like breaking glass tubes, pull & bend at the same time:
After deburring on sandpaper or a sharpening stone we get a nice blunt shortened needle – as syringe I use a 1ml size with a silicone stopper, types with plastic stopper do not have a smooth action (the needed ones are sold as “latex-free” insulin syringes):
Against drying out of the solder paste I store the filled syringes in an old marmalade container with a tight lid, some drops of IPA (Iso Propyl Alcohol) or pure ethyl alcohol are added, the cap I shortened by melting with hot air:
Flux is used to cover the PCB before soldering. I produce it in two forms, as a liquid and with a honey like consistency by dissolving violin rosin in IPA. 1:10 volume ratio for the liquid, about 2:1 for the paste, which means dissolving as much as possible in warm IPA. Crystallisation forming the paste occurs on cooling. It´s more like making jam.
Here is the cheap Chinese solder paste I use. I guess it is meant for silk-screening. Adding IPA drop-wise may be necessary to get the heavy sticky liquid state we need for our purposes.
If you know some really good lead-free solder paste I may switch in the future. But most probably the 20..30°C higher process temperature needed will force us to use some different techniques.
I almost forgot the most important point: Try out things to get a feeling for the process & check your materials.
In this example you can see how I marked two cleaned areas, one with the excellent “Mechanic” solder paste to the far left, the other area with a more “professional” paste which needs a totally different process & does not work here.
In the right area I tried my first needle looking for dot sizes & how the paste performs on a oxidized/uncleaned copper surface:
A little drop of solder paste is squeezed from the blunt needle and applied as a small dot on the PCB. Pointed tweezers are helpful in placing SMD parts.
(BTW: I don´t like comments like “your fingernails need cutting” – no, they are useful for work!):
On complex boards, think strategically. Try to put fine-pitched parts (small distance between neighbouring legs) first, you may have to correct their placement a lot or even to wipe off all their paste. This is hard to do in a stuffed corner.
All two-terminal parts are easy – I just apply paste for the next two to five parts & drop them in position. I don’t know if that is necessary or even harmful, but I press them down a bit – I don’t want to loose parts when handling the board later.
Bulky parts come last. They are always in the way & it needs more concentration to not touch them erroneously.
First we need to bake the paste for some minutes. I use 10mins on “Silk” then heat up 20secs to “Cotton” – here the solder paste smokes a bit, the flux does its cleaning & flowing. Finally boost up to “Linen”.
Then reflow (= melting) occurs, stay there 15secs, eventually pressing down the PCB with some pointed instrument like a small screwdriver for to reach peak temperature. Then slide off to an unheated non-metallic surface to cool down slowly. Listen to the clicks of the iron´s temperature regulator, they give acoustic feedback of the temperature you reached. Use no steam 🙂
At the reflow point you may notice several wonderful things: surface tension of the molten solder pulls together smeared solder paste & parts start sliding a bit to “click” in position if unprecisely placed. This sliding effect depends on the board design – datasheets provide you with optimized PCB patterns for this to happen if you run into problems with your design program´s generic patterns. Don’t expect everything to work the very first time.
Some caveats: An over-rapid heat transfer can cause solder splattering and the production of solder balls, bridging and other defects. If the heat transfer is too slow, the flux concentration may remain high and result in cold solder joints, voids and incomplete reflow.
When I have to deal with through hole connections I put aluminium foil under the PCB.
Finally check for shorts between adjacent IC legs. SMD-probes are useful. In any case do a visual inspection with a magnifying glass, I use a 6x & have the feeling 10x is best for our work.
Because our flux is pure rosin, it is “no clean” and can stay on the the final PCB. Cleaning with IPA and a toothbrush is also possible.
As more and more is known about the physical background, electronics turned into a science where devices are digitally simulated and calculated according to mathematical models.
Especially digital electronics with its binary states´ precisely defined 0 and 1, is -nobody wonders- either working or not working. For the majority of digital hardware that means: One wrong bit will break its operation. What we will make from digital chips is always working – but never exactly.
By definition states between 0 and 1 are not allowed. In digital electronic databooks these states are called “not defined” or “illegal”. They are not usable to build a reliable, predetermined, deterministic machine (e.g. a computer) which produces exact, reproducible output within its enviromental parameters (i.e. the computer is functioning).
This exactness is remarkable, yet doesn’t fit to our known physical laws.
Where is the Heisenberg uncertianty (http://en.wikipedia.org/wiki/Uncertainty_principle) of modern quantum physics which should make such precision impossible?
The answer is astonishing: Digital electronics use symbolic states outside the physical reality! (http://computationstructures.org/notes/digitalabstraction/notes.html)
E.g. standard TTL logic gates (http://en.wikipedia.org/wiki/Logic_gate) operate with a 5 volt power supply. A TTL signal is defined as 0 or “low” when between 0V and 0.8V with respect to the ground terminal, and 1 or “high” when between 2.0V and 5V. States between 0.8V and 2.0V are “illegal”.
The experimental sound circuit SNU, which is the main device of the workshop, uses these illegal states and drives the digital chip it uses into this in-between world of uncertainty. What we get is complexity and uncontrollable behaviour. This workshop shows how building-blocks can be arranged in unusual ways. The SNU or the sequencer SEQ8 are just examples of how arrangements of the instruments can be build.
Not two of them will sound the same.
You get an alive instrument, different from a sampler that only controls premade sound. Like a violin, with many possibilities to create sound but also unpredictable moments, it requires constant judgement and adjustment of the player and at the same time has a live of its own.
or if you like..
„Einfluß statt Kontrolle“
Bei diesem Workshop erhalten die TeilnehmerInnen die Möglichkeit, experimentelle Synthesizer zu bauen, fantastische Noise-Maschinen. Jede ist anders, Gesetze werden auf den Kopf gestellt, IngenieurInnen kriegen die Krise & doch funktionieren sie:
Unkontrollierbar, launisch und eigentümlich wie nur wenige Instrumente.
Hier gibts reichlich Werkzeug, Lötkolben und Elektronikbauteile – Vorkenntnisse sind nicht nötig.
This is @ the tweak festival in Limerick:
The next seven video snippets are by Natalia Borissova, Munich curator & artist, learning to know the SNU:
Some of my own pieces where the main soundsource is a SNU – SNU-II-Soundtest uses the additional Seq8-sequencer, all the “Fulda” pieces have a delay & occasional ring-modulator added. “Heisenberg” has an additional sample of Werner Heisenberg speaking about quantum physics & a tanpura loop:
SNU-II-Soundtest 03.06.08, 04:02, 64kb, mono, mp3, 4.75mb
Fulda sucks 2009 21.08.09, 02:08, 256kb, stereo, mp3, 3.94mb
Fulda sucks 2009_2 21.08.09, 44:19, 256kb, stereo, mp3, 81,18mb
This Thing is Fukking Speaking I – mastered 27.09.09, 06:08, 256kb, stereo, mp3, 11.25mb
Here is a sound test with a recently developed Ultrasonic device (+ a SEQ to modulate the new Ultrasonic-Transmitter):
This Spider Loves Ultrasonics 16.09.10, 04:27, VBR, stereo, mp3, 7,14mb
After the lame Intro here is the 1st Soundtest of the Ultrasonic Pulser we actually built:
Ultrasonic Pulser 1st Test 26.10.2010, 11:27, 96kb, mono, mp3, 7,87 mb