Long "Holiday" ;

To all my friends, i'll be leaving Singapore for Taiwan tonight.

Will be there for three weeks for military training.

Returning on the 4th of October (woo...hell long time for me that is...)
in the morning.

When we're over and done with this, it'll be a month left before our passing out parade when we don our new set of stripes.

Still have lots of projects and stuff to work on, and these won't change; i'll continue to devote effort and time to them. there's just too much of this stuff i need to finish. much of them are a backlog of projects with histories stretching all the way back to the start of my secondary school life, and with pre-histories well, originating from my scientific prehistory. Even so, many of them are important milestones and enabling technologies. Who would imagine a crazy more-than-decade-old idea of laser trapped and bouncing around in a multifaceted crystal in a complex manner as the core module of an artificial brain (sketched in that wonderfully childlike manner) can now point to the study of complexity and emergence and chaotic phenomena as fields leading to insights and a new nascent paradigm of artificial intelligence?

the future is uncertain until we receive our postings, where we will likely remain for the rest of our NSF lives. I can only wish real hard that they give me some room to work on my interests.

this is a piece of shit. don't go ard telling ppl i write poems. please.

Should i continue working my way down this seemingly endless path that seems
to lead to nowhere?

this doubt about pursuing electrical engineering, electronics, robotics and artificial intelligence, and a whole host of other crazy ideas is recurring with greater frequency these days. yet the passion i have for these things is undeniable. i cannot relinquish them.

i have witnessed another idea collapsing before my very eyes this afternoon. i dont seem to be able to salvage it from the heartless process of engineering decision. And that decision appears to write off all the effort spent on and off this thing over the past 4 yrs or so.


Doubt
------
The craziness of those ideas make them flimsy
nothing to stand against
the winds of skepticism
blowing from the inside, and the outside

doubt has been lingering around every dark corner
something that appears everytime i hold up the mirror
in self-scrutiny

a spectre resides within me
it reminds me of a bad past
it threatens a terrible future
if i choose to keep bashing through this uncertain path

uncertainty and doubt
the two horns of the devil
that devil's inside me
and i can't get it out

i don't know
don't know if i should keep making these silly patterns
patterns that consume my mind
and sap my energy

and when a nasty revelation appears before me
the shock waves of the ringing death knell
smash my house of cards
all the hearts and gems come crashing down
so does faith

the joker sneers at me

all for nothing
wasted energy
dissipated into the nothingness of the universe

if that is what lies ahead,
a disappointing retreat i must make

should i go on?

A tale of two methods

There was once the method of circuit fabrication, which took a piece of aluminium foil, stick it onto a piece of paper (for backing), print or draw on (with liquid paper) the circuit traces onto the foil, then dip it in concentrated saline and let it be the anode (did i get it right? hiyah the positive end of the power source lah) of the electrolysis of salt solution.


Supposedly, the chemistry goes like this: in the presence of chloride ions, the aluminium oxide layer that is so omnipresent on the aluminium surface will sorta dissolve. When the electrolysis occurs, the chloride ions are also discharged and highly oxidising chlorine gas is produced. Which attacks the raw aluminium metal whose protective oxide layer has been stripped. So the aluminium metal gets oxidised and forms aluminium chloride, which dissolves into aqueous solution.


This basically means that the parts of the aluinium that are exposed to the
saline solution will dissolve, leaving the parts protected by liquid paper. Which we can remove with alcohol and connect metal parts to.

Sounds perfect, but here's the snag(s):
1. because the whole sheet of aluminium is only linked to the power source at ONE particular point, what happens if etching occurs such that an "island" of aluminium is left behind? Since the aluminium is no longer connected to the voltage source, the electrolysis does not occur, so the aluminium isn't etched away, and remains behind. Leaving a really really messy (i call it 'noisy') circuit board. And it totally sucks if these things interfere with the operation of the circuit. But tests showed they didn't matter too much.

2. i totally chose to totally ignore the fact that aluminium cannot be soldered using the conventional tin/lead solder. All thanks to the irritating oxide layer present on all aluminium surfaces. And all previous attempts to temporarily eradicate the oxide layer for the solder to stick first, failed to work.

There exists solder that allows for soldering onto aluminium. It is no doubt expensive, and difficult to get.

Another alternative may be treatment by a chemical, called the tinning solution. okay maybe its called something else, nvm...anyway this is an organometallic compound, whose metallic component is, well, tin. Basically it liberates the metallic tin onto the metal surface. Since we can solder on
tin, we've solved the problem. Of course i do wonder if the tin sticks to the
aluminium in the first place! Such a solution is used in the conventional PCB
fabrication process, to coat with tin portions of copper onto which we want
to solder our electronic components. This is because tin makes a better medium to solder on, due to its lesses reactivity as compared with copper.

Once again, the hindrance to solderability is the bloody oxide layer!!!

As for Snag #1, we could conduct the electrolysis with a set of inert electrodes, and then place the aluminium to be etched really really really close to the anode...hoping that the chlorine produced would dissolve all the exposed aluminium nearby. Of couse chlorine is a very powerful oxidant, so how to find an 'inert' piece of metal 'impotent' enough to resist the temptation of losing its electrical virginity to chlorine? Can't possibly use mum's gold jewellery for this purpose can i?! (yes i was actually audacious enough to think of that.) Okay actually pencil lead can do the job (sentence added on 10th Sept)

The investigations into this method happened too long ago for me to recall if i've tried the remedy stated in the paragraph above. I must have made tons of notes on countless sheets of surplus "Industrial and Marine Electrical Contractor Tax Invoice" papers, but they are now nowhere to be found in my new home...and i've unfortunately forgotten the numerous minor details of my previous exploits. dank. (btw, 'dank' is the only word my schoolmate of 6 years considers 'clean' enough for usage as general expressor of anger/disappointment.)

Lone exploits sometimes get quite tiresome. You conceive it, you try it, you fail it, you try it again some other way, you get yourself out of your blind alleys and nasty snags, you kick yourself out of depression and disappointment, you debug everything yourself, you be the worst and harshest critic of yourself, you judge yourself, you screw yourself, until someday it (hopefully) works and you have something to be proud of.

IF that day EveR COMES.

Meanwhile, no one's gonna test your crazy ideas, so you've gotta test them yourself, and sometimes you've got so many tests which involve so much repetition that you get tired of doing them and too lazy to bash through that dense vegetation of drudgery to get to that lone tree on the empty patch.

**The Hand of Hang Jian's Rationality Department pulls him back from his
natural-inclination-driven digression**


Woo...okay (insert slightly dazed look containing a tinge of bewilderment)Anyway, i was thinking about this method which i had long abandoned, and compared it with the current one involving electroplating. The electroplating method creates really fragile circuit traces, but allows for soldering. The other method has rather robust circuit traces, but resists soldering. BOTH face the same problem of messy, noisy circuit patterns. (oh boy...circuit patterns...i do sound metaphysical to the electrically uninformed human don't i now?) Which i think i have solved by distributing my electrical contact points. BUT if the proposed solution to Snag #1 (see above) works? Then i have no more noisy circuit traces to worry about anymore!
(by the way, i will not extend the proposed solution to Snag #1 to copper clad boards, because they don't allow me to pass the circuit through a printer so i can only hand draw my circuit patterns...with liquid paper!!! no way...)

Which leads me to the question...Any other compelling reasons for using the electroplating method instead of the 'other' aluminium foil etching method?
Other than solderability? And supposing the noisy circuit pattern problem in this method was solved?

Reasons FOR the plating method
Environmentally friendlier because unlike etching, we don't need to dispose
of the plating solution.
Soldering is possible. Can't solder on aluminium without special solder which
is hard to find.


Reasons FOR the etching method
Don't need to hunt for zinc cation containing solutions...just add salt to
water
there's a good thickness to the circuit trace. Furthermore, the resistance of
aluminium is lower than that of zinc...(i think...you think, i thought, who
confirm? later...)
If the solution to Snag #1 works, we also don't have a problem with noise.

There is also a snag with the plating method regarding soldering. Due to the
tiny thickness of the plated metal, the circuit trace is extremely fragile. During soldering, the heat of the soldering iron will be passed from the circuit trace to the adhesive below it. The adhesive will melt under the heat, thus losing its tack. There is also cohesion between the circuit trace, the soldering iron, and the molten solder in between. When the soldering iron is lifted from the trace, this combination of factors causes the trace to dislodge from the substrate of the circuit board. This condition occurs in both the plating and etching methods.

The difference is made by the thickness of the trace. Aluminium foil is certainly thicker and more able to handle the physical strain due to the above situation than the plated trace. Nevertheless we do not discount the possibility that the aluminium foil trace may tear during soldering too. In response to the certainty of this occuring in the case of the plated circuit (as indicated to much disappointment by most recent tests...but also long anticipated AND ignored) 2 methods were developed to reduce the possibility of tearing.

They are, in short, to sink the heat rapidly, and to reinforce the traces by prudent and effective coating of the traces with quick-drying adhesive.

So i think the main issue wiith aluminium foil lies with its unsolderability.

Provided our solution to Snag #1 works...(okay i know the repeated reference to Snag #1 is getting irritating), unsolderability is a serious problem.
if we didn't use soldering, what could we use? a room temperature solution?

Powdered metal?

---Date line. the above was written on 9/9/05. the below is written today.---

Anyway, with reference to that issue of uneven etching of the aluminium foil in the 'other' method, a solution has just occured to me. It would be possible to press a matrix of pins onto the aluminium surface. These pins are all connected to the positive end of the power source. Which means that we now have multiple contact points on which we apply our positive potential. Which means most large islands created are probably in contact with one of these contact points and thus will eventually be etched away. Which leaves us with the very very tiny pieces of aluminium micro-islands that makes up the bulk of the noise. I can imagine using sharp pencils as positive electrodes, press them firmly on the micro islands such that electrical contact is made, and see them dissolving away. This was attempted before, but i think it somehow didn't work out...will investigate again. The micro-islands, as mentioned a couple of paragraphs ago, do not quite affect circuit operation...unless we've got really close circuit traces. The benign-ness of

these noisy things are based on the assumption (okay make that a fact) that most of the islands are electrically isolated from each other. But we should have a way to remove them if we need to. So i hope the proposed method works...though i think it did not. We'll see.

To do now:
1. make sheets of paper-backed aluminium foil
2. print copies of a test pattern onto the foil with my laser printer
3. try out the etching method again with the new recommendations

Alternate Circuit Fab Process (For prototyping purposes)

Aims and Objectives

The conventional PCB fabrication process is tedious and expensive, and involves many chemical compounds.

The exposure process takes some time to happen. Then there is the developing, and then etching. Etching probably takes the most time, and expensive etching tanks are required to do the job.

For those working at home, such a process is expensive; the chemicals are costly, toxic and difficult to obtain, and requires most importantly, an etching tank if the etching is to be completed in an appropriate length of time. If insufficient funds are available, as always in the case of self-funded or poorly-funded projects, an etching tank is a painful expenditure. I've been reduced to immersing a nascent PCB in a plastic container, agitating the mixture with my hands and a pair of chopsticks...

Etching is also an environmentally unfriendly process, and the chemical waste generated is not something that should be washed down the sink at home.

These disadvantages of conventional PCB fabrication make it unsuitable for small scale uses, such as those in makeshift labs and homes.

However, there is also a desire to fabricate printed circuits for regular prototyping purposes, because they are more reliable and less tedious to make than soldering them onto prototyping boards. Don't you find flipping a proto-board on one side to insert a component, then flipping to the other side to solder the connections, then flipping back again and getting confused something good to avoid?

Prototyping on printed circuits also means the prototype is more like the possible production version, which makes the prototyping process better.
There're some things breadboards aren't good for too.

with the above, we've come up with a method ourselves. Its by no means a production method. Just meant to make prototyping easier, faster.


Principle
Print out a mask of the circuit pattern on aluminium foil. Electroplate the foil so that only parts not covered by the mask will be plated with a thin layer of metal. Using something sticky, press onto the electroplated aluminium foil, and peel. The electroplated layer will be peeled off from the foil and onto the sticky piece of thing i want my circuit to be on.


The Process

1. Creating the pattern
This can be done with PCB creating software. What should be printed out onto the foil is a negative of the actual PCB circuit pattern.

To do this in PowerPCB:
After creating the circuit pattern, select the "drafting" button.
Click on the "Copper Pour" button on the drafting toolbar.
Create a copper pour outline area along the board outline.
Click on the "Flood" button and click on the copper pour outline.
ALL THESE ARE DONE IN THE SAME "TOP" LAYER, OR THE SAME LAYER AS THE LAYER WITH A CIRCUIT PATTERN.
Then go to File=>CAM
When dealing with the settings, select "Copper" to be printed, in Black colour.
Preview to confirm, and then print.
SELECT PRINT OPTIONS, AND CHOOSE "MIRROR PRINT"! OTHERWISE, THE CIRCUIT PATTERN PRODUCED EVENTUALLY WILL BE LATERALLY INVERTED!!!


2. Printing is to be done on a laser printer. The aluminium foil should be made to be free of creases and stuck to a piece of paper. This will be fed into the laser printer. Printing settings should be set for max toner and darkness and resolution. Parts not meant to have copper MUST be fully covered with toner.

Do multiple passes thru the printer to ensure this, if possible. Otherwise, hand-correct with liquid paper.


3. Trim and clip the masked paper-backed aluminium foil onto a hard flat sheet of material. Immerse and electroplate.
Electroplating was done with "Baker's Soldering Fluid No.3", which is actually zinc-chloride. Or rather contains zinc-chloride. (So what exactly is its composition??) And so we use zinc plate on the other end of the electrolytic cell. well...the zinc plate isn't exactly zinc...contains some iron too. Well that's fine.


4. Once done with electroplating, transfer onto a sticky surface.


Things to note:
If the sticky surface is flexible, do not allow it to flex too much, cos the very thin electroplated layer with break.

Electrical contact is below 20 ohms...but as yet unsure how much resistance there is. And yes i know anything even at 1Ohm and above is rather unacceptable...i know i know.

After soldering, take care that the components aren't pulled off the board. That'll wreck the circuit traces.

A sketch of my thought process

Last night I thought i could write in prose about my personality, but when i failed to do so, i thought i could write it in a poem. that failed too. writing assignments simply don't work for me. it just seems that i can't write about some topic on-demand. Not even when GP A levels demands it, nor even when i demand it from myself. it just doesn't work that way.

so i proceeded to work on my alternate circuit fabrication proces project, and wondered how i abandoned an earlier process for the current one. These things happen all the time in engineering...or at least in most of my own pursuits. Millions of (okay less than that...) ideas come and go, and the surprising bit is that after you've killed an idea, its ghost comes after you and you're forced to consider it some time later. The ghost might be resurrected in physical form once again, for the same purpose, or becomes the seed of a new idea for a new purpose. Which is very good. So its good for engineers to keep ghosts as pets. They might be useful someday.

Anyway, i wondered about the reasons for abandoning that previous idea for the current one. So i wrote a document last night trying to address that issue.

Turns out that document seems like a good way of presenting myself. It shows the haphazard approach that is present on all levels of my pursuits; things come to mind at random...i may lose interest in one thing and pick up something else to chew on for a while before returning. i've got lots of projects in mind, and they all compete for brain-time. i learnt that i can't complete one thing at a time, nor read a book from cover to cover. Maybe that's why i haven't read fiction in a long time. So i won't force myself into things i feel i cannot make progress for the moment.

hmm i thought i had a lot to talk about, but i've forgotten. can only remember the above...which can be summarised as haphazard and messy.

The next two documents are about this whole circuit fabrication affair. The first is background about the latest method, which i thought would be the final version. which is why you will see an introduction that suggests its some kind of paper introducing a new method for circuit fab. It is followed by an outline of the method. I will call this method the 'plating method'.

The second document is what i wrote last night. The issue surfaced in my mind recently, so i had to address it or risk lack of peaceful sleep. that document is like a 'real-time' recording. The thought process really flows that way. Sometimes the terms used are rather incomprehensible. This is not because they belong to some highly technical vocabulary. Its just that i have mental images of some things, and i just give a short label that will immediately refer to these things. I think scientific literature is like this...of course the compactification of expression leads to incomprehension by the so-called 'layman'. Think of it as a painting. Its sometimes very difficult to describe the subtle portions of a painting. You can't just say "The Potato Eaters" is a painting of some people eating potatoes. Doesn't tell much abt the painting. You just refer to the title, and assume that the reader has seen the painting and understands the label is tagged to the painting. Hence its difficult to describe some things, so we just tag a label to a commonly-accepted-previously-tied-down meaning. Just that i'm alone in this, so no one may understand what i mean.

Still, they make interesting collections of words, albeit incomprehensible. No?

Enjoy the peek into my rotten numskull (its spelt that way??!! i dunno, just assembled it phonetically.)

Saturday

saturday was a long day.

despite the fact that i slept at 1am the night before, i woke up early into 8 o'clock. Turned on the computer, noted down the stuff i had to do on notepad (which was previously scribbled on 4 small pieces of paper back in camp). The usual stuff on a weekend morning (ie computer and internet, breakfast with mum, piano, television, newspapers) took me to 11am faster than i would have liked it to. hmm actually for us life-deprived NSFs, we sometimes wish time stops for us on weekends...then again, if time stops, how to run our schedules?

(oh boy this is turning into a boring events log for the day)
was watching Air Crash Investigation when my mum got me off the couch and into the car to Parkway Parade where i would get part of my million things done. They were: getting a haircut, go to hardware shop and bookshop to look for stuff required for the circuit fabrication process project, going to Giant hypermarket to get many cups of cup noodles to pay back the amount i ate last week in bunk.

And then we had our lunch at the basement food court.

And we went home, and i was supposed to leave for raffles place at 2pm to once again search for stuff i needed for the circuit fab project in Meng's Store, before making my way down to City Hall MRT. But i took a nap which lasted a tad too long, and so my mum brought me straight to city hall where i met Zi Jian and we were supposed to walk to the Starbucks at Suntec and meet up with Dequan. For a good conversation together. But this didn't happen due to the Comex and the huge crowd it created. Instead we spent some time walking ard the exhibition, and had a look at the food fair too, where we met Dequan instead. Then more walking. And i picked up some leaflets at Apple's booth and have since gotten so smitten by the iBook that i have decided to get it. Seriously. Then we went to Mac's for a drink and some fries, and after some talk abt the military, we left. I wished we could have dinner together, but Zijian had a class gathering and Dequan was going to have dinner with his family. So i went about town all alone.

went to HMV, looked for some CDs...remembered this Virgin Classics CD compilation of Mozart Piano Sonatas. Couldn't find it. Went to Esplanade Library, got some stuff, went to Meng's Store at Raffles Place but found it closed, then went to Taka, had dinner at Mac's (ate McSpicy Double, a meal that brings forth nostalgia and many painful memories of the times at VJC robotics) and then got Kueh Lapis for mum, then went to Cold Storage to look for the circuit fab stuff (later it turned out that the contact adhesive was not suitable after testing at home later at night) then went to HMV at the Heeren (wow i actually found my way there, by a stroke of luck. i never knew where it exactly was along Orchard (if it really was along Orchard Rd) just followed my friends blindly in the past) That was where i saw the CD compilation, selling for $20+ great deal but i didn't buy it then when it was on clearance or sth...was placed in this box with a mish-mash of other classical CDs(last yr i think). Now its gone, and i can't find it. Tried Borders, but the place was messy, couldn't find it...though one CD had some sonatas on it, thankfully the digital sampling showed it was a bad performance...in my humble opinion. I didn't like that interpretation.

Left at 11+, returned home at 12. Slept at 2.

T'was a long and lonely day...