Engineering Endeavours

"Anyone who has never made a mistake has never tried anything new."

-Albert Einstein 


Seems I havent updated my website in a while. Dont ask why. I have been busy making stuff though. I did make A HUGE solar panel like Ive said, Infact more than just one.

This is the larger panel I made with the cells shown below. It puts out around 12V@3 Amps. It was enough to melt the solder from four 1 ohm load resistors, as they got VERY Hot. Theres no Doubt in my mind this could charge a 12V battery if needed. Id love to have more than one of these, its been sitting in my basement though without any plexy on top. I dont feel safe enough letting it sit outside unguarded :/

This is a smaller, recent panel Ive made specifically for charging a cell phone incase power goes out (or if Im in some remote location). It can provide up to 3V@3 Amps. 3V isnt much, which is why I have to boost it in order to use it.I will also be making a mount for it. Yes its also made from somewhat broken cells I got from ebay. 

LED Flashlight I was working on for the past month.There was a lot to do with this project, I had to figure out how to charge Lithium batteries, balance them (I have two in series), design a low battery indicator and a boost converter to power LEDs. Seen here is the LED's, boost converter and Low Battery circuit. By My Calculations I will have to charge it once every few months because of the quiescent current.


Finally got around to building my boost converter based on the 555 Timer and also built a 3.6V NiMh Battery pack from batteries I had from a "dead" laptop battery. It  should still prove useful for a rechargeable flash light. I noticed I have to use a higher duty cycle for the same output voltage  when using a lower voltage battery. Still is pretty bright but because i used 20 degree LEDs the spot is more concentrated and isnt as wide as my other  one. The benefit of this flashlight is the fact I can charge it via Solar Power or Even by a hand cranked generator. Speaking of solar..

I finished up two panels, one I had been wanting to make and another where I bought some cells through ebay. The cells are VERY fragile though and broke a bit too easy for my comfort :/.

This is the Larger Panel, made with 5x5 cells. Yes many broke..from accidents..but they are way thicker than the others ones I have. Can maintain a output of 5V at 300mA or so..good for charging a phone or batteries even. Dunno why the low output. I suspect the breaks have something to do with it :/. The holes on the bottom allow for airflow. I looked up the price for these and a few years ago I paid 67 bucks! Can you say Rip off!?

This is a smaller panel I made, made with 12  3x3 solar cells. They are rated 2amps each but that  seems rather optimistic. It gives about the same output as the one above, but this one is sealed up. I left it outside for a day to see how well it can survive..and it survives rather well. The metal tape might come off though, so I probably should have used silicone to seal it up, but silicone doesnt like to adhere to wood/paint..atleast the bathroom kind.  These cells were very fragile...paper like much thinner than the ones above. Paid like $15 bucks  for these.   One fully broke after I dropped the panel (again an accident) and couldnt be saved but it still puts out a good 5V at 300mA.

Then for earth week I jumped on a good deal for 36 3x6 solar cells. It was a kit for $47 and included free shipping. Came with Tabbing wire, buss wire, a diode and flux pen. of course it wouldnt be complete with out solar cells.

The Money shot :D Showing off the awesome blue color of the cells. Supposedly these can output 3.6A Each at 0.5V. 36 cells will give 18V open circuit, enough for a 12V battery. The whole panel should output 60W or so.


Ive been trying to make Ferric Chloride for a while and I think I may have finally found a way. Using Fe2O3 and reacting it with HCL didnt work.  Reacting Ferric Hydroxide with HCL may work with an excess of HCL but It seems better to use Copper Chloride or just to buy FeCl3


I finished my proof of concept for making a boost converter using PWM. I used a simple 555 Circuit to make the duty cycle adjustable and attached it to the circuit below. I am pretty satisfied with the results I got. At first I was getting odd problems with efficiency and such (was only getting 38% with a MOSFET and 73% with a NPN transistor). I was also using a silicon diode that I thought was a schottky type, turns out it wasn't. Guess that's what I get for using parts out of a cheapo computer PSU. The upgraded circuit performed rather well too! I am currently getting a 95% efficiency from the unit! Wow! I never expected to get something like that. 

It lights up 12V Lights pretty brightly. My breakthrough came when I used a 4.8V NiCd Battery from a old RC car as the power source. Its also big enough to fit into a Altoids Tin :). Seems 6V Lantern Batteries aren't really suitable for Flashlights. 

The reason why I used 4 182 Ohm resistors in parallel was because I dont have a 45 Ohm resistor. When I used a 1k resistor the MOSFET was being driven to softly and was heating up because of it, so a few people suggested driving it harder by using a smaller resistor. I mostly used parts I had around, 555, a pot, a cap and diodes I suspect everyone should have. But the MOSFET and Switching Diode are harder to come by..Unless you order them from Mouser. The Inductor I used is from a radio shack assortment, the color code on it says its 105 uH and has 2 Ohms DCR. Sure it could be better but at 95% efficiency why complain? Of course thats then the battery is fully charged. Efficiency will drop off as voltage gets lower. I forgot what the lowest voltage a 555 (TTL) can withstand, but We can probably use the CMOS version and get the same results.

The Load is 21 LEDs in series/Parallel and need 12V @210 mA, but their pretty bright with the circuit outputting 9.93V @ 72mA (Ive seen the circuit put out up to 11V).


Power Supplies are everywhere. Their in our computer, cell phone, TV..everything. If its powered by Battery, its mostly likely got a Switch mode power supply in it. Same goes for anything wall powered. SMPS's are in practically everything. Their more efficient than a linear converter because power is only being consumed for a portion of time. Now, there are a ton of Specialised IC's out there for controlling PWM for SMPS's, sure you can buy one off the shelf for a few bucks (most are pretty expensive, about the same as a PIC Micro) but where's the fun in buying one off the shelf?

Ever since I heard of the Idea of using A PIC micro as a SMPS controller Ive been intrigued by the idea. PIC Micros can output up to a 10 bit resolution PWM signal. PIC Micro's have a wide voltage range as far as I am concerned, they can operate from 2V up to 5.5V, which makes a PIC the Ideal candidate to boost voltages from a battery (2.4V to 3V). Of course, it still can control things. For example, boosting up a 6V Lantern Battery to 12V to Power some LEDs.

The program I used is simple. It Outputs a 50kHz Signal where the duty cycle is adjusted via a analog voltage that is read from a pot. As the Analog Voltage is Varied, the Duty Cycle is also Varied. 

This is the Boost Converter Basic Circuitry. PWM is the output from the PWM from the PIC. Vin is the input voltage you wish to boost. Note there is no Feedback..yet, since duty cycle is pretty much set by a variable resistor. Since the load is constant (in this case), the duty cycle can also be constant once its set for the proper output voltage. In the case of powering the LEDs the output voltage is around 10V, I can either A) Find out the resistance of the VDR and implement it via hardware OR Find out the exactly duty cycle and implement it in software.  Implementing in Software uses less hardware, resulting in a smaller footprint. As it stands now a PIC18F1320 Is used to provide PWM but I am curious to know if a PIC12F can be used since most of them have PWM..the only problem would be software. The type of basic I use doesnt support the 12F procs, only the 18F. I could just get the includes and edit them so they will work with the 12F, shouldnt be hard. Since I already know the duty cycle I can just hardwire it in via software. Of course if worse comes to worse I can always do ASM *grumble*.


Lately Ive been experimenting with home made batteries. The Idea came to me when I was thinking  about fuel Cells. I wondered if Tungsten, the element found in light bulbs could be used to make a cheap Fuel Cell. Light bulbs can be found at the dollar store, mine sells 4 for $1, Add Salt Water and another a electrode and youve got a cheap Fuel Cell. Thats the theory atleast. In reality its probably Harder. To Test this out, I charged 100mL of water with H2 and O2 using 2 Graphite electrodes and a 5V Power Supply that had some salt (a pinch) added to it. Let it bubble for a few, inserted the Tungsten electrode along with a graphite electrode and measured voltage. Got somewhere around 300mV, and didnt measure current, however I am wondering if its because of the salt in the water and not because of the actual H2 and O2 recombining.

So then I figured, well what if I make a 20% solution of NaOH and Stuck aluminium in there to give H2 (of course I know realise I dont think that reaction gives H2...). I still got the same reading, However, measuring voltage from The Peice of aluminium to the piece of graphite, I got something Much greater. Thus Sparking an Idea. I made a list of Electrodes and Electrolytes one would have in the house. Electrolytes are:

  • NaOH
  • Copper Sulfate
  • NaCl
  • KNO3
  • HCl
  • KCl
  • H2O2
  • Lemon Juice
  • Baking Soda

Electrodes Include:

  • Mg
  • Zinc
  • Steel/Iron
  • Aluminium
  • Graphite
  • Copper

It seems that anything with Zinc or Mg, will have the most power output. If One needs the lightest Power Source, Aluminium/Mg in a solution of 20% KNO3 Works Pretty well. No Corrosion as far as I could see on the terminals. Surprisingly, the lightest and Most Powerful Power Source is Graphite/Mg in a solution of HCl, However, the Mg strip is Corroded heavily by the HCl, so this battery probably wont last long. The only problem with the KNO3 battery is the fact KNO3 is a oxidiser, which is used in black powder and rocket fuel. It doesnt get along with Aluminium, but only if you have it mixed with Sulfur and such.


I have successfully found a way to measure inductance by measure frequency. It works for most of the inductors ive salvaged from PSU's, even flyback transformers.  However, I cannot say how accurate it is, because it wont measure the one good inductor I have (4.7uH) So I will have to buy a few parts. Strangely, I dont have 47k Resistors..I have no clue where they went.  Its pretty much a adapter for my DVM, just have to use my meter on Freq Mode, and do some math.  Sure I could get a PIC micro to do it, but why overcomplicate things?

Heres the link:

I  used a 1000pF Cap  For all the Caps in parallel with the inductor. Thats the only change Ive made to the circuit (and I didnt include the buzzer).


Im currently trying to find a way to measure inductance by measuring frequency of a colpitts oscillator. However, it seems my meter cant measure the frequency when I use certain Inductors. Either its not oscillating with certain Inductors (needs to be amplified) or my meter cant measure it (which is impossible-since it can measure up to 10Mhz). 

I have also made the RS232 Module for the data logger, but havent tested it. I used a new method of etching, that involves using a sponge and ferric chloride. It works pretty fast, probably the most fastest and efficient way of etching I have ever used. Link to new section

I also want to come up with a way to make Ferric Chloride. I think It should be as easy as mixing HCL and Red Iron Oxide.

I also forgot to add that I made some adapter boards for the parts I bought. You can view them here:

 *UPDATE* I reacted 136g of HCL (20% concentration) with 10g of Red Iron oxide. It seems by the color of the liquid I have made FeCl3! I am letting it settle as  so I can get an idea of how much yield I have.


Added Data Logger Project. Schematic will be posted soon. 


Ive gotten more parts in, this time from Mouser. I plan on doing some experimentation with the I2C Bus. For this I needed a better PIC with I2C. I found two PICs that suit my needs, a 28 PIN and a 40 Pin. I always get a 40 Pin in case I need more I/O's. Heres the parts list:

  • 24LC256 -256k EEProm 
  • MCP3221 - 12 bit ADC
  • MCP23008 - I2C port expander
  • MCP9801 - I2C temp Sensor
  • TC1321 - 10 bit DAC

All the above parts are SMD. I have some other stuff coming from Futurelec. They are:

  • 24FC512- 512k EEprom
  • 24256  - 256 EEprom
  • MAX232 - TTL to RS232 Converter
  • AT45DB041D-SU- 4Mb Flash Memory SOIC (SPI)

Granted, I dont need 512k, or 256k or even 4Mb, but I want to make a remote data logger with all this memory.  Im also getting a SOIC to DIP adapter from futurlec to make interfacing to the Flash memory easier. Here are the two PICs I also bought. They have SPI and I2C hardware support.

  • PIC18F2410
  • PIC18F4520

I also got some 3.3V regulators for the 18FxxJ series, so I can use those for low voltage applications.


Google will be soon moving my site to google sites, Im excited by this! Im hoping the editor interface will be a bit more expanded (I love lots of options) and maybe even a down-loadable interface would be nice :). 


I had put that PSU I made up on Instructables, since Ive been wanting to put something up there to share my knowledge. Turns out, it had more of an effect that I had predicted. 



Link to instructable


ive been experimenting with my new PIC programmer the Junebug whenever I have time. Ive gotten it to display stuff in BCD and then Display info on a 7 segment display. Ive also gotten it to display A two digit number on one display. For some odd reason I can only display the 2nd number first and then the first number 2nd. Not really a problem but it seems odd.  I also recently bought a 8X2 Character LCD from here. I didnt want to use a 16X2 LCD because it seemed too big if all I needed to display was two characters.

I also made a breadboard PSU, which can be found here:



I am adding stuff as I go along,  I recently got a package from mouser with a bunch of stuff in it, which I'll be using with my PIC programmer that I just got in the mail.  Some of the stuff Ive gotten includes: 

  • PIC18F44J10
  • PIC18F24J10
  • 74HC147N    
  • M74HC4543
  • MCP9700T

The two pics have more RAM and more I/O than the PIC18F1320 that I have, incase I need it. The 74 Series Chips are going to be used to power a LCD and for a home made BCD keypad. The MCP9700 is a Centigrade temperature sensor that I will be using to sense temperature. A Small description can be found in the PN junction temp sensing link to the right.


Ive found that the wind generator concept works. It would have given 20W at 600 RPM. Note I say "would have", this is because the resistance of the wire started to be too much, and reduced the voltage I would be getting from the generator. As A Concept, It works =D. So I just need bigger wire. 

 20W Isnt Much, but you can probably charge a battery or something from it, but its still usable considering I got the parts for nothing (well, other than time salvaging wire). The sad thing is, I havent come across a Microwave in a while :(. 

 Ive begun to experiment with PN junctions as temp sensors, as they have a very linear curve, 2mV/C. See New section for More.


Added Wind generator Project and will add pictures of completed PSU. 


Ive updated the Charcoal Sifter pages and other pages.  I know Im missing pictures, I'll get to those sometime at the end of the week.


Created New Page. Welcome To Engineering Endeavors! I re-did my old one to better fit my scope of things I do and have done.

Added Home-brew PSU Page.