LED Backup Light - 3 * XPG

[betti154]

Hi,

After one of my backup lights flooded the other week, I looked at buying a replacement. As I couldn't find one at the price, function and reliability point I set about making myself one.

Due to the caves dive I do, I wanted 5 hours of burn time as a backup solution so went for the following config:

• Light: 20mm triple XPG R5 board, individually addressable 1S3P
• Battery: Single 18650 3100mah panasonic
• Driver: lFlex, configured with duo mode (3500ma high, 600ma low)
• APEM piezo control

The good:

• form factor is small yet nice in the hand
• sealed design should never flood, charging is through port plugs with magnetic HallSW by taskled
• on high, the output is solid at approx 1100 lumens
• on low, burn time is 5 hours

The bad:

• optic is not very tight, beam is ~30 degrees
• on high, burn time limited to ~40min before fade begins and the odd flicker
• on low, output isn't very good (I think I need to increase this and sacrifice burn time)

 

[lucca brassi]

Nice work Damien ! You have charging pins on both sides ? any venting during charging ?

 

[betti154]

The charging sockets are on both sides (pos and neg), but the port plug never unscrew so there is never any venting. It's my understanding that venting is not required for lithium batteries, though I'm open to comments here.

To charge you just plug 4mm banana plugs on each side and use a magnet to make the sockets live (they're inert by default). I haven't done it yet, but was considering making some kind of cradle.

 

[Packhorse]

Nice.

Did you consider using a MAX 1555 Li Ion charging chip? Simply supply power to the terminals and away you go. No need for internal magnets or external charger.

 

[betti154]

Did you consider using a MAX 1555 Li Ion charging chip? Simply supply power to the terminals and away you go. No need for internal magnets or external charger.

Answer is no, as I've no idea what one is. I'm only using the Hallsw as that was all I could come up with. Works well, but if there is a no magnet solution please point me to it.

 

[dexter9]

Hello betti, can you make a drawing/schematic of the body of the light?

it looks very nice.

 

[betti154]

Hello betti, can you make a drawing/schematic of the body of the light?

it looks very nice.

Sorry but I dont really have he time right now, as am flat out at work or on video projects. Basically the body is one piece with hole each end, one tapped m16 for piezo and another is open 25mm ID (OD is 35mm) with flush walls along the body. Alloy top is bored out in front to fit optic, then stepped out to take 5mm thick acrylic (glued in with jb weld because I'm cheap and lazy, glass and oring would be nicer and not as committing).

The the back of the alloy piece steps down to 25mm to fit the body and has a barrel type oring seal, held in with a grub screw on the wet side of the oring (this is just a safety measure as the top is tight, almost too tight). A bad feature is the hole for the grub screw wants to pinch the oring when I inserted the alloy piece, but I eventually got this in cleanly. A threaded locking solution would obviously be better but again I'm too lazy to do threads on lathe.

Electronics are relatively simple. LEDs wired in parallel to lflex, which goes to battery and piezo via capacitor 100uf from memory. Port plugs as charging sockets go to hallsw and the to battery.


Regards,
Damien Siviero
http://damiensiviero.com

 

[Packhorse]

http://www.maxim-ic.com/datasheet/index.mvp/id/4002

Please not I have not tried these with "wet connection charging contacts".

But I cant see any reason they would cause issues.

 

[b-bassett]

id be conncernd with using the maxim chips directly on wet connectors, if there is any voltage potential between the 2 inputs there would be some current flow when submerged, posibly causing galvanic corrosion.
a simple defence againt this should be the use of a diode on the pins, probably not required, but id use them anyway to make sure.

only diss-advantage of the diodes would be a voltage drop. but with a 5v supply ( usb?)and the low dropout of that chip, this shouldn't be a problem.

 

[Packhorse]

I think you can input up to 7 volt into the Maxim. Put a bridge rectifier inside the light and you could charger it off AC or any polarity DC connection.
I had intended to do this to my PPO HUD project but never got around to it.

 

[betti154]

I guess the 7v limit would preclude using this on anything but single cell battery solutions, unless you can get higher voltage ones.

I've used the same sealed charging approach on my video lights (battery in head) for about 6months now, quite like it. On those I built the magnet into a charging plug so don't even notice is (they use 4S packs). Was planning the same on next video light build with a 5s2p pack in canister, but put charging sockets in head.