3rd Annual BLF Contest - Hand Made - Nickel's Colorizer---------- (Sept. 30, DQ'd Not Finished In Time)


* To make a flashlight that uses both analog and digital means to mix XM-L RGBW color led's.

* Use 7 segment displays to view 30 bit color mixing (i.e. 10 bit PWM/channel) in real time.

* Use modular 7 segment display pcb's if possible.

* Provide cooling to 36W max led array using PWM'd laptop fan.

* Use a bunch of cheap laptop pulls for power, and display voltage with 5mm rgb led.

* And lastly, try to use aluminum shapes, screws, wood, and sheet goods to make an interesting form factor.


Order coming from Industrial Metals Sales, plus what I have on hand to put this together.

Here is a thought on how the AL faceplate is going to look over the lexan protection lens. Some 1-1/2" sched 40 pipe sleeves to hide the Ledil RGBW optics. Those optics are for the MCE, but initial reaction is they seem to blend the colors well for the XM-L. For those that are curious, the Ledil color optics include a frosted TIR, with a very fine textured pmma? lens cover.

Time is slipping by, less than three weeks left, now.

When ordering up the aluminum shapes and sheet materials, I ran into a circular reasoning problem. Order bare minimum? But shipping is kinda expensive.....shapes are cheap so order more to justify shipping....shipping getting more expensive so add more cheap shapes....Well anyway, I ended up on the stocked inventory side of things .

Out comes the metal cutting blade and a line of 3in1 oil along the cut.

Gonna get some samsung blue tops out if I can.

I got to twisting the pack around trying to get the batteries prized free from the glue on the back. Too much bending and twisting, I think I prematurely cut some of the taps and BZZZT, small smoke event, and into a bucket and out the door (three steps away). Thankfully when things straightened out, no short, BE CAREFUL!!!

Don't bend and twist a lot getting them out, or this could happen!!!


Got a couple of those craptastic 4X18650 wrapper ripper holders. Going to build charge plugs so as to avoid taking them out of the craptastic holders all the time. One is going to be 1S4P for the leds, and the other is 4S1P to run the fan, and a 5V (1.5A) regulator for the electronics. Going to use a 0.33Ω 3W resistor to knock the fan voltage down from the 4S pack to approximate 12V.

Below is a pic on how the CAD design is transferred to the frame of the lantern style flashlight. Screw/glue blocks will hold the four sides together, and provide standoffs for the internal AL sheet used for the battery, fan, and cooling compartments. Lots of drilling and tapping to be done for the #8 brass screws, pots, etc.

Since my favorite pen style center punch finally broke ( it was made of aluminum and the tip was bent), I went a looking, and came up with an automatic one. Never heard of such a thing, but it works great.

Time to recharge the battery.

Really had to hog out the holes for the pots, they had a 3/8" bushing and plenty short at that. To top it off, no panel nuts, thankfully they are same as tv ant panel nuts, so found some of those.

More drilling and tapping. I spent way to much time just finding a way to clamp things up in position and not interfere with the tap handle.

Here is how the screw blocks are attached, and side panel attachment. The blocks also provide support for the AL sheet compartments for the battery's, fan, and cooling chamber.

Counter sunk holes for the brass screws were hit and miss. Some good, most O.K. and some a bit sloppy. Such are ways of hand drilling, and keeping things perpendicular with the tap .

Cut sheet for fan and battery compartment. Batteries will be on the bottom and underneath the horizontal shelf. More cuts required to fit in the cooling chamber partition; which attaches to the 1" standoffs in the front behind the emitters.

The box dimensions are 5-1/4"W x 7"L x 4"H.

This is where the project ends prior to the contest ending. Pretty happy thus far on how it's turning out, and anxious to get on with it. An after contest section will follow on thru project completion.

In hindsight the project, from a mechanical perspective, could have been simplified to a great degree by using lighter gauge panels, excepting the front face where the emitters will be. Also, using thru bolts would have reduced most all the hole tapping action which was a real pain. Some things you just don't realize till you start digging in, and start facing the challenges presented to you.

Of course the electronics were a little over the top, and chewed up a good amount of time. A certain satisfaction is derived by placing a challenge for ones self, and continue on thru the journey no matter where it ends. I am sure the other contestants feel the same way with their creative and very unique builds. Out for now.


Here is a pic of the dusted off prototype for a single channel color control. Using a PIC18f4331 for the brains. Will be employing shift registers (MC74HC595) and pnp fets to multiplex the 4 digit CA 7 segment display(s). Need to put some lower value resistors to let some more current thru, as you can see the rather dim display in the pic. A 5 turn pot on an a-d input is used for extra precision that will be used set the channel PWM levels 0-1023.

A 24 detent encoder is setup, but haven't developed the code for that yet. The idea is being able to scroll thru a menu of interesting? colors just by twisting the knob.

Figured out the encoder operation today, yay. Usual coding and setup snafu's, way too many edit code and compile operations. Fair to say not a code guru, haha. One of the reasons for selecting the 18f4331 is because it has a built in quadrature encoder interface, or QEI module. The internals keep track of the index count (1 click on the encoder), which way it is going (forward or backward), and how many counts before it will cause an interrupt and reset the counter.

Here is a vid with the encoder lighting up some leds in order, and in whichever way the knob is twisted. When the build is complete, I hope to have 24 or so preset colors (like a color wheel) based on the knob position.

4 Digit 7 Segment PCB's, Or?

So worked up some 4 digit 7 segment multiplexed boards with the HC595, and a couple of dual pfet's in the SC70-6 package size. The '595 RST (reset) pin is tied to VCC and the OE (output enable) pin is tied to ground. Got the displays, HC595's, Fairchild FDG6316P's, and some other stuff from DipMicro. DipMicro has been a good place for NOS parts at a reasonable price.

The board dimension is 1.2" x0.8" which makes for less than a square inch (good for OSHPark). I may go ahead and make it bigger to lessen the approx. 0.24" between (horizontally stacked) board edges (i.e.gap), due to the right angle pin connectors. Also would like to add some mounting tabs. If that happens, than I will order from one of the budget China suppliers. A second possibility is to go ahead and incorporate the display boards onto a single large board with the micro and everything else......hmmm. I'll have to sleep on that.

A Single Large Board it is!

Way too much times spent on this pcb board nonsense. Heh, I guess I'm procrastinating on getting the old reliable hacksaw and drill out and start making some aluminum fairy dust, because the materials came in last week. My shoulder is getting sore just thinking about it.

Have to get the pcb design out though, cause it's going to take close to three weeks for the turnaround. So no excuses from now on?

Utter nonsense...

The Colorizer main boards are in. I have to say this is the most complicated board to date for me. Hope it works out, there is no real backup plan at the moment if something is routed wrong, or if I missed something.

Top side of board.

Bottom side of board.

And here is a rendition of the 2 channel slave (AMC7135) boards that have been sent off to OSHPark.

Have been hit and miss on the software lately. I found a pushwheel binary counter switch (0-6) in the switch drawer. If I remember correctly, this came from an old scrounged HP computer bay sized cd player. The binary switches are too expensive to buy from one of the big box companies. Have that prototype'd now to change the three different groups of modes that I anticipate using:

(1) 4 color with two or three brightness levels each.

(2) 4 color mixing by using the four different pots, with display on the 7 segment displays.

(3) Encoder switch to cycle thru 24 or so colors like a color wheel.

So the Colorizer board has been loaded up. The doohickey hanging off the back is my pfet hack to fix the bum pfet traces on the pcb as noted in an earlier post. Still having problems keeping track of the multi (4X) multiplexed displays (4 digit) shift registers. I have the first two 'HC595's turning the pfets on and off as each of the digits are scrolled thru. So more debugging required as I try to get my head straight; bailout would be use four spare pins of the mcu to do this directly.

Slapped a switch, charge port, 4S balancing plug, and an atc 5A fuse on the nominal 14.8V battery bank. Loaded the AMC7135 2 Channel slave boards https://budgetlightforum.com/t/-/35492 with a conservative 1A to begin testing.

Mmmm, cooking up something good?

The contest has come to a close, and it's a DNF for me. Here are the last bits of the build prior to contest close. The build will continue on in a clearly defined Colorizer After Hours section .

Some time is being spent on debugging the first go at the Colorizer pcb. There are at least a half dozen and counting flaws in the design like pfet track mixup, missing pullup and pulldown resistors, routing a track for temp sensor, and so forth.

Got the displays on track which consisted of a having one of the common anode wires to wrong pin(s), and a couple of bad solder joints during reflow, and a solder whisker between two pins. This was due in part to the hand placed solder paste (with syringe) being a little bit sloppy.

Encoder code is not working that was developed on the breadboard, so more debugging of pcb required. What is left is the four color with 3 level UI group, and the analog (potentiometer) color mixing group modes.

Near complete electronics setup, including a Dallas 1-wire temp sensor (ran out of a-d pins), Panasonic encoder/switch, for color wheel settings, binary pushwheel switch to change group settings, 2 channel slave AMC7135 board(s) for up to 4+ amps per channel (assuming triple XM-L color emitter configuration), Displays with color channel indication, laptop fan, etc.

Couple of analog color settings. So they are supposed to approximate tangerine, chartreuse, and Blue Violet respectively. Display readings are prior to using the actual PWM values based off a very large lookup table.

A crappy video from my 2mp first gen camera showing the 4 Color group modes. The UI uses short clicks to change power levels, and a long click to advance to the next color at the present power level.

That's all for now, till the project moves forward in an after contest mode.

subscribed :slight_smile:

+1 I like colours

More, More!

Sounds Great and very unique. Looking forward to the build!

I am but a minion serving the Masters, and observers, of the BLF flashlight universe :slight_smile: .

This is sounding complicated!

Good luck with the build :beer:

It’ll be one piece or concept at a time, hopefully it will add up to a flashlight.

I dusted off my breadboard prototype for one channel that I have had kicking around. Needless to say it will take more than an ATTiny 13A to run the works. Bonus points if I get the encoder working for a menu of 24 colors, one for each detent.

Yep, I'm lost as usual. The electronics are to blame.

I’ll try and cover them up ASAP, so as not to confuse.

Lacking the mechanical genius of the MRsDNF’s of the world, I try and deflect attention away from the physical build when possible. :smiley:

Fooled about with the encoder switch today and got that working the way I want to. Put up a quick video of the encoder in action. Triple xm-l color faceplate idea put to paper.

4 Digit 7 Segment artwork complete. Put up a pic of the pcb in the Electronics section.

Damn, that looks like some serious circuitry you developed there. Looking forward to seeing more.

Time is slipping by! Colorizer boards came in today, and they look pretty good. Sent off 2 channel AMC7135 slave boards to OSHPark. Start a couple of cuts on the 4” flat bar. Updates in OP.

Shaving things kinda awfully close to the deadline. Hope I can make it?

Did anyone else have the same problem as I did reading this. Google translate wont work here? It looks like black magic to me. I should add that l love watching black magic.

MRsDNF, Ha! No, cutting threads, knurling, chopping up reflectors, making bezels, and such is the mystery! :slight_smile:

You work hard to finish it and I’ll work hard on a prize to honor it.

RBD, That’s mighty generous, but not the reason I entered the BLF contest. I’ll do my best to finish.

For me the contest is way to explore new ideas, or act on something that has been on the back of my mind for a while. It’s definitely the journey, or path traveled that interests me most. Best of all, is to see all the crazy creative builds the others are doing.

Don’t know who will get it, just offering encouragement. Any who do this know it’s not about the prizes anyway. Those can’t compare with the sense of accomplishment upon finishing a complex build.