Deepoutdoors Cold Fusion - Flooded Head

[undrwater]

Be sure the original LED and driver can't be cleaned up and made to work.

I don't hold much hope for this as the whole head was flooded with saltwater, and never rinsed out (the previous owner was unaware that the light head had flooded). Even if it works for now, I wouldn't hold my breath for too long. If any of the LED's work, I may use them in another project.

 

[mpenders]

I picked up one of these lights awhile back, also inoperative, and have tried to revive it. The head in my light is fine, but the problem lies in the electronics located in the battery can. From what I can tell, one of the inductors was damaged, and a fet was fried. With no support from Niterider, I tried having a local shop replace the bad components, but didn't have much success. I could get the batteries to charge properly, but it would not light/switch correctly.

From what I understand, this light needs both the pcb and switch in the head, and another board in the can to both work with each other for the light to work properly. The board in the can both regulates the charging of the batteries, and driving the output of the light (on/100%/50%/off). There is the magnetic switch in the head along with the leds (3x SSC P4's, I think).

I've also toyed with the idea of using it with an xml with ahorton aspheric, but not sure there is enough space without machining down where the P4's were heatsinked. Assuming the innards of the head are toast due to the flood, and that you're moving on to trying a different led/driver combination - you'll likely not need the driver board in the can. I'd be very interested in it, if you plan on pitching it out.

Either way, I'm interested on how you progress on this. Best of luck.

Mike

 

[undrwater]

OK. Here is my (poor) sketch of the Fusion Head:

Here is the top-down on the LEDs:

Here's a pic that shows what I THINK is the switch:

 

[undrwater]

This shows the back of the LEDs. Where they sit on top of the heatsink.

Here's the top-down view of the head.

 

[lucca brassi]

On your place I'll put that head on grinder or lathe and clean out complete head from those inserts and get clean cylindrical hole . That mean holder for leds and centre thin barrier between cable inlet and reed witch room. You should leave upper raw metal alone because behind is lead for magnet )

Next thing is to made on lathe nice H shape heatsink with tight fitting (4-5mm metal thickness ). Keep in mind to remember best position to put reed swich on right place .

Put in first half H shape heat sink driver and in second led (opposite side ) . Heatsink should trapped with screw drilled through H perpendicular to the enclosure




Then I think you have place to choose driver , optics and led by your own taste ( probably XML , ) or SST50 .

I see no reason to save existing state with those triple P4 , because I think you get troubles with temperature dissapation.


I suggest a ''clean install'' , should save many headaches :thumbsup:

 

[wquiles]

I suggest a ''clean install'' , should save many headaches :thumbsup:

+1

Excellent advice, plus the P4's are so outdated by now, that even a "simple" 3x XP-G board to replace the existing 3x P4's would be a HUGE improvement:
http://www.ledsupply.com/indusstar-3up.php

if you go with that 3x board I linked above, you can easily pick an optic (made for those boards) that makes the whole thing plug-n-play

Will

 

[b-bassett]

definatly recomend getting that pedastal removed, gives you a flat clean base to work off of, plus there would then be enough room for the led + optics

a single xml with ahorton's lens will work perfectly in there, or as also sugested a triple board with a cute narow optic should be a major improvment over the old.

i think the driver will need to fit into one of those recess's though, that could be a little challenging, but not impossible. (there are a few boards around with the driver and LED on a single board)

the reed switch arrangement should be easy enough to replicate, but depends on where your putting the driver.

 

[undrwater]

On your place I'll put that head on grinder or lathe and clean out complete head from those inserts and get clean cylindrical hole . That mean holder for leds and centre thin barrier between cable inlet and reed witch room. You should leave upper raw metal alone because behind is lead for magnet )

Next thing is to made on lathe nice H shape heatsink with tight fitting (4-5mm metal thickness ). Keep in mind to remember best position to put reed swich on right place .

Put in first half H shape heat sink driver and in second led (opposite side ) . Heatsink should trapped with screw drilled through H perpendicular to the enclosure

http://www.scubaboard.com/gallery/showphoto.php/photo/173685/title/1030/cat/5565


Then I think you have place to choose driver , optics and led by your own taste ( probably XML , ) or SST50 .

I see no reason to save existing state with those triple P4 , because I think you get troubles with temperature dissapation.


I suggest a ''clean install'' , should save many headaches :thumbsup:

I like this idea just for being more "homemade". I also like the xml with the aspheric lens (been looking about the site and like the beams). About the heat sink, I have a hard time picturing what you are talking about. I imagine the "H" fitting into the head where the "legs" are touching the walls leaving the open sections such that the driver and switch can fit into the current "wells" that are already there. Is this correct? Hardest part to understand was:

Heatsink should trapped with screw drilled through H perpendicular to the enclosure

By this, do you mean screw down into the center of the "H" then into the body of the existing heat sink (what is left after machining)?

Finally, you mention using a reed switch. Do you think that is what is appropriate for the existing magnet? It is spring loaded, so it does not stay in the position, but springs back. I thought the hall toggle was more appropriate based on this. Thoughts?

Thanks all for your help!

 

[DIWdiver]

The Hall Toggle may be better, if you can get the sensor into the correct position, and the magnet moves far enough to turn the hall on and off. It could actually be too sensitive to place directly over the magnet. You'd have to play with it to see what works.

If you need lathe work to clean out the head, I may be able to help. PM if interested.

Don

 

[lucca brassi]

About the heat sink, I have a hard time picturing what you are talking about. I imagine the "H" fitting into the head where the "legs" are touching the walls leaving the open sections such that the driver and switch can fit into the current "wells" that are already there. Is this correct?

yes

By this, do you mean screw down into the center of the "H" then into the body of the existing heat sink (what is left after machining)?

Not centre - just through side of one ''legs''. I used ''recessed ''allen screw M4 for easier work . I also put some arctic silver 5 between housing walls and heatsink.

If you imagine cross section of upper picture is symetrical through centrehorizontal and vertical .

No rocket science it is simple and it works ;-) . Sorry for language it is hard to express sometimes specially if you want to talk multi- dimensional ;-).
and google also don't help too much.

 

[b-bassett]

maby a quick sketch up woud help

this is how id try to do it

i dont think youll have enough room for the driver (blue bit) though so it may have to be just the front 'cup' with the driver tucked into one of those recesses.

the green bit is where the reed switch/ toggle board would go.

im not sure quite how much room is in those recesses, but there are a few drivers that take a momentary switch input, (i.e your sprung loaded magnet), perhaps one of these will fit? would certainly lower the parts count. thus fitting in the head easier.

the DW driver iv got here would fit the bill nicley i suspect, its 25mm diameter and 11 high ( with inductor)

whatever you go for is going to be a squeeze, but is certainloy do-able

just had a thought . . i think most tripple boards and optics would take up less space then a single led and aspheric, in that case there would be more room for the driver.


N.B. dimensions are rough.

 

[DIWdiver]

It looks like the aspheric is not going to fit. It would only leave a few mm for heatsink thickness and driver. I think Lucca's idea is probably best, with a triple board and optics. It won't get you the beam quality of an aspheric, but it will fit!

If you were really determined to put the aspheric in, you could cut a hole in the lens and glue on another piece in front of it. The second piece would be smaller diameter, to fit within the bezel diameter. That would get you additional length equal to the thickness of the original lens, which might be just enough. I would definitely use polycarbonate for this, as it machines well, glues well, and is tough enough to survive being stuck out in front of the bezel.

 

[undrwater]

It looks like the aspheric is not going to fit. It would only leave a few mm for heatsink thickness and driver. I think Lucca's idea is probably best, with a triple board and optics. It won't get you the beam quality of an aspheric, but it will fit!

If you were really determined to put the aspheric in, you could cut a hole in the lens and glue on another piece in front of it. The second piece would be smaller diameter, to fit within the bezel diameter. That would get you additional length equal to the thickness of the original lens, which might be just enough. I would definitely use polycarbonate for this, as it machines well, glues well, and is tough enough to survive being stuck out in front of the bezel.

What I think I'm hearing you say is to cut a hole in the FLAT lens? Then glue another flat lens on top of it just larger than the diameter of the hole? The existing lens is glass, so I suppose I should get two of the polycarbonate type?

 

[DIWdiver]

I just had another idea. If you ran multi-core cable, you could locate the driver in the canister, and not need it in the head. You'd have to make sure you got a driver that could handle the LED being remote. That could be a problem for some drivers.

BTW, I just confirmed that the dimension that b-basset shows as 26-30mm in his sketch is in fact very close to 28mm for tightest focus (about 8 degrees Full Angle). If you want wider beam, make that distance less. 20mm gets you about 40 degrees FA.

 

[DIWdiver]

What I think I'm hearing you say is to cut a hole in the FLAT lens? Then glue another flat lens on top of it just larger than the diameter of the hole? The existing lens is glass, so I suppose I should get two of the polycarbonate type?

Yes, exactly.

This would be a rather extreme design choice, but I'm sure it could be made to work. If this were my light, I would only do this if I REALLY wanted the beam that only the aspheric can provide. Or maybe if I wanted to prove I could do it.

 

[undrwater]

I just had another idea. If you ran multi-core cable, you could locate the driver in the canister, and not need it in the head. You'd have to make sure you got a driver that could handle the LED being remote. That could be a problem for some drivers.

BTW, I just confirmed that the dimension that b-basset shows as 26-30mm in his sketch is in fact very close to 28mm for tightest focus (about 8 degrees Full Angle). If you want wider beam, make that distance less. 20mm gets you about 40 degrees FA.

Interesting. I'm guessing multi core would include more than two wires? I'm beginning to think it's time to bite the bullet and go the 3 up way as was mentioned before. Perhaps use the aspheric lens in a mag light head?

 

[b-bassett]

id double check all measurments befor you go spending any money, we know the xml + aspheric is very tight, but i wouldnt want to splash out on ta tripple + optic, then still not have enough room.

i think thers a good chance of getting a driver in one of those recesses, could even attach the reed to the driver and then slot it in place.

i know the DW driver iv got is a bit high, but i think the inductor can be moved off the board, might help things. a 7135 based driver could be built to fit nicly but your battery pack is too high a voltage for a single led, so thats not very practical.

you definatly need to work out what will fit though. dont spend out on things that may not fit. it all adds up very quikly.

 

[lucca brassi]

time to http://www.bhlens.com/product_detail.aspx?pid=1 ?

 

[b-bassett]

shame that fresnel has a focal length of 30mm, would have been near perfect

 

[undrwater]

mpenders mentioned the board on the battery pack. Do any of you think it would be useful to see that? I'm beginning to wonder if it's time to find another host for a head?