A Solar Rechargeable FlashLite - Done! (2014 BLF Handmade Contest)

Here is the final Trim of the Solar Rechargeable FlashLite (2014 BLF Handbuild Contest)

In the Beginning:

Here is the pile of materials. It’ going to be a solar rechargeable light (Free Energy) :slight_smile: . A couple of 18650 K2 LifePo4 batteries, FET direct drive, Hi CRI XML-2 led, 7segment voltage indicator, two button switch thingy. At least that is the hope.

Now as OL says, GET GOING!!!

First off, is the AL bar frame that was handsawed, so nice pretty mating edges, ha, ha . At first I was going to use the sawzall on it ( because it was going to be allowed this year?), but didn’t have the right blade. No straight holes, or tapping because, they are hand-drilled. This project is more about function for me, call it industrial design, or whatever.

Next pic is a mockup of the 1” x 1/4” AL bar frame. I originally was thinking of just face mounting the mini solar panels to a back-plate, that in turn was also face mounted to the AL frame, and then picture framed with some 1/8” abs sheet to hide the edge. Now I’m looking at doing an inset job, if it is feasible, for a couple of reasons:

1. I couldn’t stand the thought of getting down to the last drill and tap of holes in the AL bars for the faceplates, and have an accident, like breaking off the tap :_( . That would just about mean a complete redo.

2. It is only going to take a little sanding, or filing, the two mating mini panel edges to get a nice snug fit. Still have to have a back-plate, but precludes the need for the abs trim. And, if something goes wrong with the faceplate mounting, I won’t have to redo the frame, because I’ll be screwing into some inset 1/4” tube, or angle.

The electronics are going to use a multiplexed 7 segments that I scrapped from a satellite receiver that was found in the local recycle/scrap metal pile. I may still buy a proper 3 digit common cathode one, because it would actually be cheaper than to buy some 2V Vgs (because of LifePo4 batteries) pfet’s to source the common anode displays.

About the only pfet part number I’m coming up with, that’s in a TO92 pkg. (no SOT23’s thank you), is something like a Zetex ZVP4105A http://www.digikey.com/scripts/DkSearch/dksus.dll?Detail&itemSeq=147381264&uq=635325857287589608 Unless someone has a lead on another pfet, low 2V Vgs, TO92 part they like to use? I have some BS350P’s but their Vgs threshold is too high at 3.5V.

After trying to solder together last years copper build, getting frustrated, and with some very uneven results, it’s time for a new train of thought. This year is the year for mechanical stuff, like lots of screws!

So let’s cut a panel for the voltage indicator side. Low and behold the Hack(saw)-A-Line jig, patent pending, ha, ha.

An ultrafine Sharpie and pocket rule is my friend.

Drill, drill, and then drill some more. Tap, tap, tap and tap for the 4-40 screws. I tried the recommended drill size of a #43 (0.089”) but had a hard time getting the tap started. So went with a 3/32” (0.094”), it’s a little loose, but the screws seem to snug up just fine.

I can’t believe there isn’t a hole poking out the side or something here. Kinda looks more like an explosion proof enclosure as opposed to a solar flashlite. Broke a couple of 1/16” drill bits, but no tap yet, whew! I thinking I will change out the socket head screws on the side panel to button heads once the project is complete.

Time to make a window in the case, somehow I always leave myself plenty of filing to do afterwards. Lots of holes, and some nipping between them with some diagonal cutting pliers. The half round and round bastard files seemly in constant use on this build.

Using some 1/8” lexan sheet that I had on hand for the 7 segment window, then secured with some 2-56 screws. This also creates a standoff for the pcb board.

Mockup, with the pcb board and battery, everything is a snug fit so far, and looks like everything still works. Using point to point wiring with some 30 ga. Kynar (wire wrap) makes for a real nice birds nest. Didn’t I tell myself, that I wasn’t going to do this pt. to pt. stuff again?…oh well, rules are made to be broken I guess :slight_smile: .

Time to start work on the pill. Added a couple of copper discs to fill in the recess of the C8 pill, where the star would normally be placed. This should get good heat transfer from the Noctigon mcpcb, thru to the aluminum frame. One thing I noticed was that when applying pressure with the thumb and forefinger to the pill, then moving it across the bastard file, made for a high in the middle (rocking horse) surface. By using a piece of 1/2” pipe to apply pressure in the middle, I was able to create a much flatter surface.

Got the momentary switch parts going. A stack up was created to keep the 14mm x 6mm switch boot in place with some aluminum plate and then the 12mm switch (robbed from a remote switch) is soldered onto a piece of proto board. I expect to glue a piece of nylon bushing to the push button, so as to make up for the slop created by the spacer plate. A jst plug was used for a quick disconnect from the main board.

The solar pv panel side of the lite is going now. Cooked up some homemade AL 1/4” angle rails, then mounted them to the main frame with some more 4-40 screws. Also more .060” AL sheet laid in on top of the rails to create a back plate for the mini panels. A small bit of filing was required to fit the the panels length wise (EDIT: don’t do this…damages conductor leads). The panels will be silconed in when the lite becomes finished.

Here the panels are set in silicone, and ready to wire.

Things are looking a little plain on the 7 segment display side, so set about adding a decorative face plate. A CAD drawing was used, so I could try and keep the center punch inline, and overall layout uniform. The original faceplate made a good template for the cutout and attachment screw holes.

Cutting out the window for the inset display panel.

Finished decorative faceplate. Lets drill for the programming and UART ports, and paint the back faceplate black for contrast.

Big visions of doing the upside down C8 pill and a TIR, and it just didn’t happen (at least for now). So that and about 5 other iterations with a reflector and so on, still nothing. Finally, it is just screw the noctigon direct to the frame with a little Artic Silver Ceramique 2. Used a piece of the 1/4” x 1” AL frame material and a little Lexan for the cover. This accounts for the pure flood in the beam shots in post #2.

Here is the final trim for the Solar Rechargeable FlashLite. The blank grommets are covering the programming port on left and the UART port on the right.

The brains of the Solar Rechargable Lite is starting to come together. Decided to go with with some M74HC595 shift registers, instead of multiplexing, for the 7 segment display. The micro is a PIC16f1827, needed a few extra pins over say a PIC12f1822. Using a temp programming setup for now. I look to have an access port setup behind a blank rubber grommet down the road.

Put together the (2P) K2 18650 LifePo4 pack. The tabs made for easy soldering. A bit of copper plate and a mini clamp made sure that no extra heat was heading to the battery. A Deans micro plug was used, just in case the lite needs to be cracked open for a boost charge.

PV panels hooked up and tested. Had a big whoops, as one panel was not producing any current. Located which panel, and had to remove and replace. Note to self, “Do Not File Ends of These Solar Panels!”. I had filed through one of the foil conductor tapes that they wrap around from the cell to the back of the pasteboard? that they use as a backer.

The parallel mini PV panels had open circuit voltage (Voc) that was 6.39V. Maximum short circuit current (Isc) was 322ma.

Tested the PV charging capability yesterday (6/25/14). The weather was not cooperating, it was completely dull gray overcast, no sun breaks at all. Today is worse, overcast and it’s been raining off and on all day.

Test ran from 11:30am to 5pm. Battery was discharged on the hobby charger before start of test. After test, the battery was once again discharged and the discharge capacity was 353mah. Nothing earth shattering as expected, but you could use low beam for 8 to 10 hrs. Wish I could have gotten a full sun day, but not expected to be sunny again till the 29th? I could follow up with more data post contest.

A pic of the Vishay 70N02 direct drive NFet. When I first started using the momentary switch, it was randomly skipping ahead a couple of modes. So put a 0.1ufd cap in parallel with the switch, and that was enough to handle the debouncing perfectly.

P.S. when in doubt, through a plug on it.

I can now program the lite with an adapter that fits the 2x6 female plug. Right next door, a conversation with the UART port is available.

6/27/14 update:
Overvoltage protection was brought up for the LifePo4 battery pack and I think I have created a sofware solution to handle this. I was going throw up a pic of the setup but hopefully the explanation in post #47 is good enough. Here is code to turn on the display as “888” which will burn 240ma beginning at 3.65V (sorry for no indents/tabs):

RefVolts = AD16f1827(FVR) ’internal a-d of FVR
BatVolts = 52224 / RefVolts
If BatVolts > 365 Then
Call DisplayNum 888
If mode > 0 Then
Call DisplayNum BatVolts
BlankDisplay = True
Call DisplayNum BlankDisplay
end if
end if
wait 1 s
Goto Main

6/29/14 Update - Solar charge test #2
So same test as previous one which was 5 1/2 hrs from same time frame of 11:15am to 4:45pm. This time the weather was cooperating much better, with the test starting with partly cloudy conditions early on, to mostly clear later in the day. The battery pack was once again discharged by the Thunder AC6, then solar charged, then discharged at 0.5A (approx. medium mode) once again to get the mah packed in. This time the solar charge attained 835mAh, hurray! So three days of similar conditions would put a full charge in the 2P 18650 LifePo4 pack (1250mAh ea.).

Do not have a very good place at the house for beam shots. So you get to look at the hop vines :slight_smile: . For reference, I have thrown in a UltraFire 501B with a XPG R5. The solar recharge lite has the Cree high cri XM-L2 that I got with last years contest/certificate from Illumination Supply. As you can see the beam is total flood. Will be useful when the power goes out at the house, or as a lantern? when out camping.

501B XPG R5

Solar Recharge Lite (Low)

Solar Recharge Lite (Med)

Solar Recharge Lite (High)

Missed April fools by 43minutes. :stuck_out_tongue:

Not in my time zone :slight_smile: .

In tribute to Magic Eight Ball, in lieu of Reserved etc.

It’s kinda related, and does involve free electrons, more info later today.


I invented 5000 lumen OTF 1x18650 flashlight.
Uses the prototype Cree XO-P emitter.

I’ll sell it for $800
Shipping takes 45 days.

I already have lots of free energy, i just don’t have a way to convert it to my preferred form or store it for later use


Pic of materials up.

And lost the spammy, april foolery subject headline before I really do get deleted…Yay…:slight_smile: . “Free Energy, Send Me Your $$$, and I will send a pamphlet on how to make your Free Energy! (2014 BLF Handmade Contest)”

That’s a deep subject…

New pic is up.

This is a mockup of the AL bar frame, and gives a general sense of the project direction.

You can dissolve a tap out of aluminum using nitric acid. It can’t be too strong or it will passivate the steel and not dissolve it. Or it can’t be too weak.

I did not know that. No chemistry experiments for me on this project heh, heh, but as the infamous Walter White says “respect the chemistry”.

As someone who broke a hand tap in a bar of aluminum in last year’s contest, I’ll note that there’s always the other side for a do-over.

Ha, no broken tap, or do-over necessary………yet. More pics and verbiage.

Nice effort. You can drill and tap holes for me anyday. Where did you get your Ultrafire Sharpie from. I I need one to reanodize my Ultrafire flashlights.

Que passa, bwana?