Bike Light Build: Part Two

I got around to playing with the light and the bench power supply last night and it appears that the light shows little or no dimming until the voltage drops below 12V. This means that in mild/warm temperatures I could possibly use NiMh batteries two, without going to the added trouble of buying two 8.4V packs instead of the standard 7.2V packs. At least that is what I have gleaned from wikipedia's information on NiMH batteries. The only other concern is the amperage draw, but I think that the sub-C size cells found in RC battery packs can handle the 500mA draw of the light, and then some.

Today I started building the body of the light with 1" square aluminum tubing that I bought at the hardware store down the road. I started with a 2" long section, which I then notched out so that a 7/8" long section of square tube could be held inside the notch.

I did this by first using a hacksaw to slit down the sides of the tube adjacent to the inner wall, up to a line that I had scribed one inch down the tube.


Next I drilled holes in the tube along the side of the scribed line where I wanted to remove material.


I then used a pair of vice-grips to break off the material at the line of drilled holes.


The next step is to file the bottom and sides flat. Then a piece of square tube is cut to 7/8" and filed to fit into the notch.

I was going to use a piece of aluminum extrusion intended to be a shelf braket as heat sinks, but then I remembered that I had some PC heat sinks that I never used on the PC in our office. I proceeded to cut those to two inches long and added a chamfer on one top corner. The picture below shows the pieces placed together as I intend them to fit.


I also placed an order with Newegg.com for some Arctic Alumina adhesive(AAA). AAA is thermally conductive, commonly used to attach heat sinks to CPU's and such. It has a good reputation for being a tough adhesive too, so I will at the very least use it to adhere the LED and heat sinks to the body. I will probably use JB Weld for more structural joints.

Next up we will finish constructing the body...

Bike Light Build: Part One

Just a few quick note about my latest project. I'm hoping to actually document this one in a useful manner, so bear with me. I've done a little bit of research about DIY lights for mountain biking and finally decided to get around to making one. Fall is upon us and soon there will be little to no chance that we will see the sun for months on end. I will therefor need something to light my path while biking.

I had considered going the cheap route of using a halogen light, but the run time is abysmal compared to other options, such as LED and HID. HID is simply stupid expensive, so I decided to go with an LED. After reading a website all about such lights here, I decided to go with a Cree MC-E emitter from ledsupply.com. The MC-E is actually four emitters closely packed together under one lens and it comes pre-mounted on an aluminum plate to help dissipate heat. The emitters are wired in series which means the forward voltages must be added together, resulting in a Vf of 12.8-13.2. The current is shared by the emitters and I will supply that with a Luxdrive Buckpuck with which I can provide a regulated 500 mA. I chose the wired model, with the potentiometer to vary the current from 0-100%.

To test the Buckpuck and LED I soldered some speaker wires that I scrounged from a box of stereo stuff I had in the closet. I put some alligator clips on those to aid in testing as well. I wired the Buckpuck to my regulated power supply and set the PS to 14.4V which is the voltage that I can get from two 7.2V Ni-Cd battery packs that I bought for my remote control car, when they are wired in series. I then attached the clips on the LED to the Buckpuck outputs and turned on the PS. Voila! Blinding frigging light. Seriously, don't look directly at this thing. At 350 mA this LED throws approximately 430 Lumens and I was driving it at 500 mA which should give it a 30% increase in flux. That is truly bright. I didn't leave the light on for long as it was not attached to any heat sinks and even an LED will heat up when driven at 6.5 watts. I just wanted to verify that everything worked anyway.

Next step... check to see how low a voltage this thing can be driven at so that I can be sure that the battery packs I have will work even when low, or if I need something with more than a nominal 14.4V.

Really Bare Bones Board Build

I purchased an RBBB(Really Bare Bones Board) from Modern Device Company quite some time ago and I never quite got it to work right. The RBBB is a VERY minimalist version of the Arduino which is easy to put on a solderless breadboard. The kit cost $7 but you have to solder it together, including some very small components, so it can be chalenging if you are new to soldering, and damn near impossible if you have a crappy Radio Shack soldering iron. I had purchased a $150 Weller iron prior to the build, so I had nothing but inexperiance and lack of skill to blame for my troubles.

I actually managed to get the more difficult components soldered on without a problem but I somehow managed to overheat one of the pads on the pcb and I lost a connection to ground because of that. The board would work it was positioned just right such that the pad made contact, but it failed more often than not.

Tonight while working on the Xbees I decided that I should finally get around to fixing this so I came up with a rather inelligant, yet effective solution. While it is difficult to see in the photo, The pin that has the bad connection is the ground pin that I soldered the Pin 13 LED to. I must have spent too much time heating that pad up because I was trying to mount the LED to it. Regardless, my rather hoaky fix was to solder a wire onto the to of that pin and stuff the other end of that wire into the DIP socket of the corresponding AVR pin. I checked the continuity with my multimeter and lo and behold it worked without needing me to wiggle the pin until it could chance upon a connection. I proceeded to check the rest of the leads and found them all to be working, so perhaps I'll actually get to use the RBBB for a project sometime in the future.

Xbee Communication Part 1: Getting Started

I have spent my free time in the past few days playing with Xbee wireless modules. I move at a snail's pace because I'm terrible at managing my time effectively. I did manage to get the two modules to communicate. Not bad for like three intermittent hours of work. I spent a few celebratory minutes typing messages between the two before I decided to start making dinner.

The first step to working with the Xbees, aside from ordering the parts from Sparkfun.com, was to solder the breakout boards and wire the circuits up on breadboards. I managed to miss the recommendation that you purchase the 2mm sockets with the breakout board so that you can change the Xbee modules at a later date so I ended up soldering them directly to the breakout boards.

I also purchase a logic level converter from Sparkfun in order to step down the voltage of the serial data lines from the arduino to the Xbee from 5V to 3.3V. I had to solder header pins on this as well, but the SMD chips were already on the PCB so I didn't have to do the painstaking work of soldering those tiny things on myself.

I ran into a few problems getting the Xbees to respond to my serial commands. The first problem was that I seemed to have the TX and RX lines reversed. I could have sworn that they we wired as the diagram in Making Things Talk showed, but my mind often wanders from the page to the breadboard, so I could have been wrong. The second problem was just the fact that I had multiple serial terminals open at the same time. This resulted in some error messages or the terminals simply not responding at all. Personally, I prefer the error message, even if I didn't bother reading it.

I chatted with Lee about the problem at work and he told me about the conflict with multiple serial terminals and I was able to fix the problem as soon as I knew to only do one at a time. Now that I have the two talking I need to figure out a project where they might be useful. I'm sure that I'll think of something eventually. At least I've taken the first baby steps though.

DIY Data Cable for Garmin eTrex

I have to start out by saying that I am terribly disappointed in Garmin for the exclusion of a data cable with there base eTrex GPS device, as well as the ridiculously overpriced cables that they sell to those who neglected to notice that none was included. I was almost one of those people. Not to mention that it seems as if there should be a way to program waypoints by entering coordinates. Even if doing so involved a tedious menu where you had to use the up and down buttons to increment/decrement the numbers until you got the value you wanted. Then at least you wouldn't absolutely need a $40 data cable. Okay, end of rant.
Upon doing some research I found that other people were a bit peeved at this as well and many had taken it upon themselves to fabricate there own data cables. One such man even made an injection mold and manufactures a few styles of connectors. He then sends up to two of these connectors to anyone that requests them for whatever price that they pledge to pay. They then send him that amount after they receive the connector, but only if they are satisfied with them. Seems like a pretty novel approach to commerce. Also on his site he other peoples' DIY eTrex plugs. A few of which were the inspiration for my own.
My plug is simply a piece of plastic shim stock cut to the width of the eTrex socket, with enough length such that I could hot glue the wires to it to secure them. Also, it is notched to fit the little boss protruding from the eTrex socket, and added notched where the pins need to be in order to contact the pins on the eTrex. To make the pins I soldered some leads that I had snipped off of some resistors and LEDs while making a light for my Jack-o-lantern to the appropriate wires of a serial cable. The pinout and another excellent cable build can be found here. I then bent the leads around the shim and put a mark behind the location where the eTrex pins were located. I used the marks to locate holes that I made by poking a shirt-pin through the shim. Then I bent the leads to poke up through the holes with an arch in them to ensure that they put pressure on the eTrex pins. After the leads were thus located I globbed some hot glue on the wires to hold them in place. I'll try to think of a more elegant way of fastening the wires later. For now the big glob of hot glue will have to suffice.

PIC Programming

I've been trying to figure out how to program Microchip's PICs in order to create peripheral devices for the BOE-Bot such as stepper motor drivers and such. This has proven to be much more difficult than I had hoped. Apparently I need to get a degree in computer science to be able to achieve this. That, or I need to spend around $250 on a PICbasic Pro compiler to match the ease of use of the Basic Stamp. I've been trying to learn C but I haven't found any helpful websites that teach C programming as it applies to PIC devices. Perhaps I'm just being impatient and want it all to be uber simple like the Basic Stamp. Hopefully sometime soon there will be an open source Basic or Pbasic compiler for PICs. Then I can accomplish what I want to do without having to pay college tuition to advance my hobby. I've got to surf some more forums to see if I've missed one where someone writes tutorials that apply to beginners with little or no experience programming... wish me luck.

Sparkfun's Triple Output LED RBG on the BOE-bot

I bought a few doo-dads from sparkfun.com a while back and finally got around to playing with the triple output LED. It works like any other LED only it has four leads instead of two. I always forget what is the anode and what is the cathode, but regardless it has one lead that goes to ground and the others go to the output leads from the BOE-bot, with a 470 Ohm resistor to drop the voltage down to a level that won't fry the LED. Please note that the red portion of the LED had a lower max voltage than the green or blue. I ignored this and used the same LED circuit I had used in a previous BOE-Bot LED excersize. It hasn't killed the LED yet so I think it will live. Anyway I wrote a little code using pulse width modulation to fade from one color to another. It works pretty well. I'll see if I can post the video I took of it working. See the code below:

'RGB LED Changer

' {$STAMP BS2}
' {$PBASIC 2.5}

counter VAR Word
DO
FOR counter = 1 TO 100
PULSOUT 9, 1000 - (10 * counter)
PULSOUT 11, 10 * counter
PAUSE 5
NEXT
FOR counter = 101 TO 200
PULSOUT 11, 1000 - (10 * (counter - 100))
PAUSE 5
PULSOUT 10, 10 * (counter - 100)
PAUSE 5
NEXT
FOR counter = 201 TO 300
PULSOUT 10, 1000 - (10 * (counter - 200))
PAUSE 5
PULSOUT 9, (10 * (counter -200))
PAUSE 5
NEXT

LOOP

DIY 5G Ipod USB Charger

I've seen several USB charger projects online and these work well for most devices that charge off of USB, but never quite worked for my fifth generation Ipod. Most of these chargers are based off of the LM7805 voltage regulator, in order to step a voltage down to the 5V required for USB. What these designs lack is the data in-out signal that tells the 5G Ipod that it is a high power USB outlet. I recently found a suggestion online to use 4 equal resistors to make a voltage divider to supply a 2.5V signal to each data line on the usb connector. Check out the thread where I found it here for more details. Anywho, I used the bread board on my Parallax BOE-Bot to prototype the circuit for testing. You may notice that there is no LM7805 on this circuit. That's because the BOE-Bot has one integrated into it, which sped up my testing. After verifying that I did indeed get 2.5V on the data lines I plugged in the Ipod. Previously the Ipod needed to be reset in order to charge on such a circuit, but now it charges as if it were plugged into a real USB outlet. Now I just need to figure out if the LM7805 will hold up to the rigors of being wired into my truck's 12V power so that I can make a charger to integrate into the Ipod car dock that I'm slowly designing.