No, I said I need some “juice”

One of the things I’ve noticed about electronic devices is that they all, unequivocally, run on electricity.  Microcontrollers and small motors and actuators and sensors and such are also fairly picky regarding the voltages they’re interested in using.  Primarily, microcontrollers require either 5V or 3.3V of power with very little tolerance away from their ideal supply.  The flow thus needs to be very regulated.

To do this efficiently a specialized power supply is often required.  Hobbyists with too much money at their disposal opt for bench supplies that have perfect little power nubs at different voltages and lots of fancy meters.

I don’t have too much money so I decided to see what I could do to make my own:

You’ll note that by checking out the back of it, it’s actually just an ATX power supply I’ve re-purposed from an old PC.  The wooden carriage for it is a few slabs cut from a desk that has long ago been retired.  One of those ones that you have to put together yourself alone in the dark of your apartment while you wonder about what might have been as you scrawl through useless misdirections and absent hardware.  The stain for the wood is called “Anything But Particle Board” and doesn’t match at all with the PCB mill I’m working on that’s positioned next to it.  The front piece isn’t attached yet, so the wiring is a little exposed at the moment.  Makes for a better picture/story.

The top of it has a power switch, a voltage supply indicator light, which is illuminated when the supply is able to maintain all regulated voltages given the current power draw on it.  It’s basically an “all is well” light.  Above that is a reset button, and to the left are 4 terminals, for 3.3V, 5V, 12V and ground.

A nice aspect of ATX supplies is that they are wired with a non-mechanical on/off feature that allows you to turn it on by sinking the power on one of it’s pins.  It also has a constant standby 5V output.  By grounding the power pin the supply is turned on with the switch I’ve installed.  By shorting that with the standby output, the reset button allows it to power down and restart.

Another bonus feature of the ATX supply is that it will cut off its voltage if it detects a short circuit or an unsafe draw of power.  This means that it will shut down as soon as it detects itself firing watt upon watt of electricity through your overweight, out of shape shell of forgotten dreams so that you can live another day to ponder the decisions in your life that led to this point.


Sleepless boys make all the toys

Not finishing projects is sort of my thing.  Finishing them and taking them apart when I’m done is also my thing.  But against all odds work as recommenced on the CNC mill.  Primarily because I really actually need it to get done so I can put it to work manufacturing parts for other projects.

Now all three axis are running on balanced threaded rod, powered by three 1.8°/step stepper motors:

Back side with cat in background offering look of approval.

The z-axis has a ridiculous gear ratio to ensure that lifting the mechanism is pretty strain-free.

Each motor was tested, everything runs, and holy shit the thing is loud.  Some dampening will be needed but for a prototype, no complaints.  The motor is being readied for mounting.  It is a 6.2v motor that will be cranked up to 12-14v for a little extra rage.  Unfortunately, the actual area it has to work with is 7″x5″x1″, which isn’t huge.  But this is a proof of concept and if it works I’ll make one that’s much (much) larger.  A greater z-axis is  going to prove critical I suspect.

To actually run this, I have a few stepper motor drivers from Texas Instruments.  Or I might just make it run directly from a dozen or so transistors and some serious power switching that will be coupled with a computer’s USB port and special drivers (that I’ll have to make) to stream bits of data directly to shift registers.

Bedtime now though.

Constructing a solar panel (part 2)

First, I’ve ordered 36 0.5V, 4.54A cells, tab wire, bus wire, liquid flux in a set from eBay for $47.99 after shipping.

When I received the package there was actually 40 panels in it as the seller threw a few extras in just in case any were damaged in shipping.  They were not.  Score.
Time to start tabbing.  If you’re going to do this ever yourself, make sure you use at least a 40 watt iron.  Tabbing wire melts beyond 680F and if done correctly, no solder is required to tab the cells.  
One by one until they’re done… note that the tabbing is happening on top of a 3/8″ thick piece of tempered glass.  The glass protects the cells as it dissipates the heat from the iron instead of injuring the silicon.
Done and done… The entire process took about an hour.  It definitely started going more quickly once I got the hang of it.  Note that the tabbing wire runs across the top of the panels and extends about 50% the length of the panel beyond the edge.  Next time I’m going to make it 100% and you’ll see why later.
Now it’s time to mount them.  I purchased a 4’x8′ piece of plexiglass from Home Depot for $18.  This gave me much more surface area than I needed, but I wanted to make sure that as this was my first panel built from scratch I’d give myself a little room between each to work with.  This turned out to be a genius decision later on.  Each cell is 4″x5″ so I laid them out in a 5×8 pattern with 1/4″ between on all sides with a 1.25″ margin.  Large margins,  extra space between lines.  How I survived writing papers in high school… Once positioned, white electrical tape was used to hold them in place.
An added benefit of the plexiglass is that it actually increases the efficiency of the cells because the chemical construct of the plastic actually deflects the photons towards the panels and also creates a sort of photon greenhouse effect with any that are deflected off of the cell on initial contact.
This is the poorly drawn gist of what’s happening.
Because the chemical makeup of the doped panels and the thinness of the cells to allow the field effect to occur, a very low voltage difference is generated.  However, a relatively high amperage is created because of the sheer number  of electrons getting flung around.  So each cell is ~2.2 Watts but 0.5V.  This ratio is going to make measuring the actual available amps a pain in the ass.  Later.
In order to build up the voltage to something usable I’m wiring each cell in series.  This puts a lot of pressure to perform on each cell as there’s no redundancy whatsoever, but it’s a prototype so who cares if it works.  To build up to usable voltage, the tab wire will fold from the top of one cell to the bottom of the next all the way down 5 columns of cells.  This will give me 5 rows of 4V.  Bus wire will be run up along the gaps between cells and continue the series, giving the upper left and lower right cells the ends of the series.  You can see the bus wire is under the black electrical tape:
Front after wiring completed:
Also, I’ve got wood:
It’s a 4’x8′ piece of plywood spray-painted white.  Any exposure of the panel that could reflect instead of absorb light is good for two reasons.  One, it might get somehow bounced onto one of the cells.  Two, cells are very temperature-sensitive.  As they heat up, the amperage declines.  I still have to crunch the numbers to see if it’s possible to addd peltier cooling that would be efficient enough to result in a net positive gain.
Mounted and sealed.  Note that solar cells are subject to oxidation when exposed to air, so making them as air-tight as possible is essential.  These are designed for a service life of at least 20 years.  I doubt I’ll let them live that long.
It’s alive!
And powering a simple inverter.   This is very lossy as there is no transformation/regulation to 12V so 8V+ is being lost to heat dissipation.  I should be able to push the voltage down to ~14V and increase the amperage 20-25%.
Next step, regulating voltage, smoothing spikes, building a base for it, and implementing a microcontroller-based light seeking/tracking.  Then battery charging/management.  Then probably something else.  Maybe a toaster bath.
So the goal was 100 watts for $100.  Total cost of materials:
Solar cells + tab wire + bus wire + flux:  $48
Plexiglass:  $18
Plywood:  $6
Spray paint:  $7
Electrical tape:  $4
Wood support beams:  $3
Screws and washers:  $5
Total:  $91

Fucking win.

The maths behind the sun powering my toaster while I bathe

Maths are a lot like karate.  It’s just cooler when an Asian is doing it.

But I needed to do some calculations or I wouldn’t have enough amperage to do anything useful.  Also I wanted to know just how the hell much this was going to cost.
First, I did want to calculate the amp hours that would be needed.  All of these calculations are done assuming 12v battery storage.
First, I’m going with the national average for household power consumption.  I know that I probably use more than this, but this will be the rough baseline.  The average consumption is 1000 watts at any given point in time.  So that’s 1kW/h.  At 12 volts and thanks to Ohm, storing 1kWh would require storing 83.3Ah.
83.3Ah * 24 hours = 2000Ah on the nose
As I’m only going to assume that I’ll have 5 hours of direct sunlight per day, I’ll need to come up with 2000Ah at a rate of 400A/h.
Incidentally that’s enough power to kill 20,000-40,000 people.  The collective State of Texas now has a boner.
I’m tired of going over this, it works out to me needing 400 square feet of grade A solar panels at a cost of $1700-$2400.  And a lot of  2 guage wire, which is somewhat impractical.  So small scale time, lets make around 80-100 watts.
The goal is to be able to generate 100 watts for $100.  
More to come…

My next attempt at accidental electrocution

We are like tenant farmers chopping down the fence around our house for fuel when we should be using Nature’s inexhaustible sources of energy — sun, wind and tide. … I’d put my money on the sun and solar energy. What a source of power! I hope we don’t have to wait until oil and coal run out before we tackle that.

– Thomas Edison (1931)


I’m not a dirty hippie.  I use plenty of non-renewable energy sources.  But I’m not thrilled with it.  So for an exercise in mathematics, electronics, and chemistry I decided to see what it would take for me to be able to reduce my overall usage.

After the maths that resulted in me realizing that a 6m x 6m solar panel would be required to power my home I decided to not jump right in to a multi-thousand dollar investment.  But I did, after playing with a simple little 1.5W panel from Radio Shack, decide that piecemeal and proof of concept would be a more interesting way to go about this.  So let’s at least make a little juice.

I started by purchasing 36+ 4.54A 0.5V manufacturing grade cells from an eBay seller.  This came with bus wire, a junction box, flux, solder, the whole thing… for $48 with shipping and such.  The jobbers came in and I immediately tore into the situation.  What follows is the start.