hpr-knowledge-base/hpr_transcripts/hpr3497.txt

Episode: 3497
Title: HPR3497: Jankilators.
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr3497/hpr3497.mp3
Transcribed: 2025-10-25 00:26:25

---

This is Haka Public Radio Episode 3497 for Tuesday 28th of December 2021, today's show
is entitled, Junkilators. It is hosted by One of Thunes, and is about 21 minutes long,
and carries an explicit flag. The summer is, follow the way of the Junkis
and Monster Into the Mubby Flux on head to watch more of the lens.
Hello, this is your host, One of Thunes. I would like to get into Hacking Power Supplies.
I'm going to be broad with my definition of the terms Hacking and Power Supply.
I might oscillate between the Starlit surface of a faraway land
and the inner workings of 20th century technology.
Let's go with that oscillator. Let's say 20 years ago a megahertz processor.
Let's say that wasn't to be laughed at in those days,
and a million cycles per second could get some numbers crunched.
Why would you need to crunch numbers? Well, that's a good question.
This example from a book by David Harrell called Algorithmix, The Spirit of Computing,
addresses the conundrum of whether it's worth even trying.
The example puzzle was abstracted to a jigsaw of 25 pieces,
which was simplified into a square, two-dimensional area, with a picture on it,
divided into 25 equal squares, so a five by five grid.
The picture was further simplified by coloring the edges of the component squares,
such that each square was divided into four triangles,
and the color of each edge would extend to the point center of each square.
In other words, the edges of the squares matched to the adjacent square.
You should assume that there is any one solution, so a complex photograph would fit the bill.
It might seem that this is a simple task. Certainly for a human,
you could find the picture fairly quickly if you've done a few jigsaws.
Had you reduced the picture to just colored edges, you might struggle a little bit more
because you don't know what it is you're trying to reconstruct.
Surely this is a task for a microprocessor.
If you were just beginning, you might think, well, I'll put one square next to another square
with a matching color face, and then I'll move on to the next square until I can't go any further.
And then I'll record that sequence in a bin of failed sequences not to be repeated.
Then I'll try a different sequence, perfect boring territory for a computer to deal with.
Well, sure enough, you might get lucky, but what's the worst case scenario here?
Let's say you've got 1 million cycles per second, that's our oscillator,
and let's call that 1 million comparisons.
If you're being fussy, let's say it's an 8 megahertz oscillator,
that 8 oscillations get us 1 comparison, so we're still going for 1 million comparisons per second.
Because we're going to work out the answer in seconds for our time elapsed.
Well, to cut a long story short, if you're quite unlucky,
and the very last sequence you try is the successful sequence,
which you must allow for if your life is depending on it.
Then you're looking at the order of 25 factorial comparisons.
Factorial being the operation whereby a number is multiplied by the number lower than itself,
repeatedly until the decremented number is 1, not 0.
You might have seen factorial represented by an exclamation mark after a number,
so 3 factorial would be 3 multiplied by 2 multiplied by 1.
Apparently, more seconds than had so far elapsed in this universe.
I'm going to trust David Harrell on that figure.
Now, does that mean you've got a terrible algorithm,
or does it mean your microprocessor is not fit for the task?
I think the point to that author was making was that there are ways that you can decide how long a thing will take before you ask the processor to do it.
My takeaway was that to step back and assess a situation within context is often beneficial,
especially when preparing to commit oneself to millions of cycles of anything, especially human stuff.
Let us return to the Starlit Cloud layer of a small English forest.
A couple of witches had noticed some unusual activity in the neck of the woods.
They reported the felling of a tract of trees to an ecologically-minded group of local humans.
They, in turn, associated the tree loss with the run-up to a village-sized hole in the ground.
In the distant past, that area of forest was mottled with cavernous bell pits from which coal had been dug out by hand.
The coal wasn't very good quality, but it turns out where you might not be able to build on a greenfield site,
where you can build on an ex-industrial site.
Furthermore, if your need is infrastructural in the energy sense, then you can get permission to dig up an ancient forest.
So my friend Eleanor popped into the woods aware of Ponchi found an abandoned trailer, a relatively small trailer,
which no one was using, been lying there for a while.
She had an idea of how she might tidy up a bit, so she took her hacksaw to the axle
and removed one of the wheels.
So when I say axel, we're talking about something like a scaffolding tube with a wall of steel about 4mm thick.
And in the end, each end of this tube was a stub-axle, which held the bearings of a wheel.
If you've never cut through a scuff-olding pole with a hacksaw, then you might be surprised just how quickly that can happen.
Sometime later, I was presented with a similar stub-axle without any attached wheel, and asked if I could use it to carry the weight of an axel flux generator.
That sounds a bit fancy, but axel just refers to the relative arrangement of coils and magnets, and as most of us know, flux is the normal kind of magic.
So having previously traced the distracting flicker of my monitor back to some wires in the wall of my house, I had taken an interest in the generation of electricity.
So for that reason, I cobbled together a three-phase generator of alternating current.
In the form of a wind turbine, in this case, I can't say I believed I was saving the environment by milling and welding a lot of 6mm steel.
Shall we jump into the theoretical aspect of that welding plasma to avoid a nasty burning?
If you've looked into electronics, you might have found a few ways of generating alternating current.
I am fascinated by those methods which use electronic components, like inductors and capacitors for LC oscillators, also transistors of various kind.
However, I must remain mindful that our bodies are currently bound to the dark shale of the open cast coal mine, where we do not have a substantial source of direct current to start with.
Therefore, I will hold my attention on the mechanical generation of some amount of current which we can harness or store.
Plans and pictures do exist somewhere in my scattered databases, but it's been 12 years since I pulled them off the internet, so there might be some show notes.
I am trying to avoid non-necessary technical description of commonly known objects.
So as a brief verbal outline, imagine a scaffold pole pointing into the sky upon which is a shorter piece of larger scaffold tube,
cupped over the top of it to hold the previously mentioned stub-axle to which is attached to steel discs which bear opposing magnets.
The robust nature of this design is worth noting, so imagine the rotor discs are 8mm thick, there are two.
Apart from mechanical strength and rigidity, the thickness of the steel should contain the flux of the magnets.
As a rough guide, you'll want the steel disc to be at least half as thick as the magnet.
The aim is to contain all of the magnetic flux and have none leak out of the back. You can test this.
Imagine looking at one steel disc and sticking magnets onto the face of the disc around the outside edge, alternating north and south facing outwards,
and then doing the same thing on the other disc, such that when you put the two discs together, the magnets align with each other and stick to each other, so north sticks to south.
In practice, you would not allow these two discs to come together.
Apart from serious injury, it would be very difficult to get them apart again.
Therefore, when you attach these magnetor rotors to the stub-axle instead of the wheel, you separate them with threaded bolts, you know, nuts and bolts to hold them apart.
And you want to hold them apart just far enough to put the coil stator in between them to hang the stator in space between the magnetic fields.
These stator brackets, in this case, are also attached to the stub-axle.
My stator was three sets of coils of copper wire cast into a disc of polyester resin.
Imagine a disc with wires poking out of it and a big hole in the middle, but the disc is of a greater diameter than your magnet rotors,
such that you can bolt it in space without any of it interfering with the moving rotors.
The moving bolts of the magnet rotors, the wheel, go through a hole in the middle of the coil stator.
So for mental abstraction, we have two permanent magnets attracting each other across empty space.
The wires to coiled wire relatively moves through those magnetic fields.
So there are no brushes involved here, and the only wearing parts are the bearings of the stub-axle.
Well, I said I wouldn't describe common things, and then I did describe a wheel, so be warned.
Here comes the stator. A picture really does paint a thousand words, so I refer you to all kinds of libraries.
The abstract, the three sets of coils, we'll call them three individual coils, each wound from a single piece of wire.
So have these phases been defined already, or is this one of those coding protocols where objects are defined upon mentioning?
Each of the three phases has two ends.
If you trace the outputs of each of the phases on an oscilloscope, you will find that they are out of phase, just as a result of the physical arrangement of the magnets in relation to the coils.
The voltage in one set of coils is peaking at a different time to the other coils.
If you recall the magnets stuck to the rotors were alternating in polarity the way they were stuck on.
I've got half an idea that as one of these magnets passes over one leg of a coil, you know, over one side of a coil, and then passes into the hole in the middle of the coil, which is the same size as the magnet, and then moves into the second leg of the coil.
At that point an opposing magnetic field is moving over the first leg of the coil, and what to the coil being a fairly circular spiral.
The current is pushed from one side to the other, up one leg and down the other. I should zoom out.
The flux path is taunting my semantic translation capabilities.
Our equipment interface consists of six wires poking out of the resin.
You can wire the ends of those three phases in different ways. The Y or star formation involves joining one end of each phase altogether in a point.
You might join a line to the center tap of that star configuration and call it neutral.
Then you could attach a separate line to each of the three loose ends and use the three voltages which you would find between the neutral and each of the loose ends.
Looking forward to the grammar episode now aren't you? Alternatively you could not attach a line to the neutral and you could just take voltages between any two ends of those loose three ends from your star configuration.
There exists an alternative configuration called delta where you connect your three phases into a triangle.
Imagine one phase is one side of the triangle such that you have three corners on your triangle where from you can choose any two corners as your electrodes.
Each of those configurations will give you different benefits relating to higher voltage or higher current across those phases.
That might be relevant if you're trying to engineer an electrical situation otherwise you might just measure and see which best suits your need.
I think in the forest they just left those three phases open not joined together at all. Then took the outputs from the three rectifiers and connected those in series to get the largest voltage.
As a low tech quick and dirty solution they just used the average wind speed to effect their regulation.
Three phase generation also gives you better return for your copper investment than some of the methods.
There's a lot more for somebody else to say about three phase generation. They can explain to you why if you've got a triangle of coils, somehow Kirchhoff's laws,
describe how the disturbed electrons do not find themselves in a useless triangular short circuit.
I'm not fluent in anything much, maybe English, colloquial English.
I did weld another scaffold tube to the floor end of the tower pole in a T-shaped joint so that we could hinge the whole tower down into an A-frame or an X-frame such that the business end of the turbine could be held at a height where the blades could be taken off
and the gear sprocket of a mountain bike which had been welded onto a steel plate so that could be bolted on instead of the blades and the thing could be pedaled when there was no wind.
A note on the blades they were curved and planed out of scaffold boards so about 12 inches wide by maybe an inch and a half thick.
And when that thing, which worked really well actually, you can plane even just with a saw and a chisel, then plane and rough sanding.
Believe me, if you did not put the brake on and you lift it off the ground and those blades start turning, it will be a 3 metre propeller, a 10 foot area effect knock out zone.
It is a scary situation if you are close enough say you were lifting the pole when someone was pulling on the rope, exciting.
You can put a kind of electronic brake on by shorting the stator coils which will cause a magnetic field resistance to the turning of the rotors or resistive load effectively.
The turbine up to the forest camp was intended to charge batteries for communication devices and lighting mostly.
Eleanor and company used three big bridge rectifiers which would handle amps, you know, maybe three amps each, maybe ten amps each.
We're getting specific now to generalize depending how much power you put into your system, you need to deal with how much you are getting out.
And the voltage is going to depend on how quickly that wheel is turning, which has obvious implications to what you can attach it to because if you just stick a voltage onto a battery to charge it, obviously there are limits.
Why does it only get so hot before they glow a melt and burn?
Interestingly in the wind turbine case, if the wind blows too hard, that's going to melt your stator which is an expensive and time consuming thing to recreate.
This particular turbine had a tail firling mechanism, so when the wind blew hard enough, the tail would be lifted and force the blades side on into the wind which would reduce the torque substantially, reducing the potential current flow.
But at the other end, which is probably more interesting to this audience, imagine a voltage of anywhere between 0 and 30 volts but with a current quite a high current, that still means your battery is either not charging or is burning or exploding.
We really could use a geeky bit of electronic engineering, voltage and current regulation.
Surely it's switch mode boost and buck by now. So we've got those nati square things for wires at least, two of which are you're in for your alternating current and two of which are you out for your direct current, bridge rectifiers, they're great.
But let's be optimistic and say we don't get the sky load of discharge down our tower for attempting the gods of physics in a Terry Pratchett style.
I didn't look for a really big zen a diode, but that might do in a pinch. Nowadays you can take your pick of voltage regulators and current limiters from, call it the internet.
The nice thing about the bigger stuff is that you can find applications for old and now maybe cheap components just because they're clunky, miniaturization isn't everything, is it?
In the context of opportunistic harnessing of power supplies, then insulation will probably be your main concern.
I think we're getting to the spicy part of the recipe. Shout out to the low text of compressing air or lifting a weight or pumping water up somewhere or turning a flywheel, loving the magnetic bearing flywheels myself.
Thinking back to that ammeter where the wire is deflected by a magnetic field, you could maybe turn the end of that into some kind of switch to a discharge excessive power.
Sounds like a relay. I can feel my mind trying to limit me to things which I can somehow make with a pair of pliers.
Voltage regulation is going to involve an inducted field collapsing and being sampled essentially at the voltage which we require.
Of course the planet is now littered with modular modules which will do this job for us very easily and it's probably much easier.
Maybe even cheaper in terms of energy, spending energy to produce things to just buy a voltage regulator, but we know we can do that.
I want an oscillator to feed my inductor and to switch my inductor and can I do without to transistor or a huge motor for that to matter.
Are the joys of a gentle stream? This janky sysamonster was faithfully anchored into stumps and loose-failed trunks of sufficient mass offering its clanky howl to be already toxified and receding forest.
Then I've run out of time to discuss the fancy magic. Sure I missed bits and made mistakes. I would complete the Christmas scene of abandoned tents, tree houses and sleeping under a very tiny haystack in the frost of the forest.
Bit too far out of scope. This was one of spoons not getting to the point. Maybe tune in next time. I'll give it another try.
You've been listening to Hecker Public Radio at HeckerPublicRadio.org. Today's show was contributed by an HBR listener like yourself.
If you ever thought of recording a podcast then click on our contributing to find out how easy it really is.
Hosting for HBR is kindly provided by an honesthost.com, the internet archive and our sync.net.
Unless otherwise stated, today's show is released under a creative commons, attribution, share like 3.0 license.