Designing a focusable LED dive light.

I wouldnt hold out for the light described in this thread as nothing has really been decided on yet. Also there has been no clear intent to make it commercially available. Although it may be in small numbers it probably will not be cheap.
 
Packhorse said:
I wouldnt hold out for the light described in this thread as nothing has really been decided on yet. Also there has been no clear intent to make it commercially available. Although it may be in small numbers it probably will not be cheap.
Click to expand...

Booooo :thumbsdow And there was me getting all excited 😉

I had spoken to Will begining of September, and he did say that he was interested in doing a project like this, but not for a few months. When this thread started up, I thought this might turn into a collaborative project, with a small number of lights produced. Maybe I need to contact Will over email to discuss just doing this as a one-off if this isn't going anywhere?
 
Last edited:
Gees Whitenoise, you didnt think I was serious did you?

Progress so far. Light head is based on a W200 and will use a Ahorton aspheric. LED will be a SST-90.
Body is POM although I may swap to aluminium later.
It will be a 2 mode ( plus off) light running 6 -7 amps on high and about 1.5 amps low.
Driver will be AMC 7135 based ( yes that is alot of AMC 7135s!).
Heat will be a huge issue. It will have a thermal switch in the head and if it gets to hot the light will switch to low mode. But I am hoping that when used it water it will run cool enough.

Battery pack will have to be either a single string of Li Ion cells at 4.2v max or NiCd NiMh 3.6 or 4.8v pack.

It will be focusable. Although I have a feeling the SST-90 will not focus down tight enough so I may switch to a SST-50 and run it at 5 amps.

It will not be a hand held light. I have no interest in such a light and it would be harder to build.

Switching will be done on the cannister.

minimax2.jpg


MiniMax1.jpg
 
*Download reported issues with 12 amc 7135's. Have you confirmed 6amp is an option? 18-20 is a big sandwich. Unless you do parrallel drivers: switching 'off' to 'low' powers up the first set (~5 amc 7135's), switching to 'high' adds the rest of them. It could be a completely dumb design, no mcu just two reed switches the first selectively powers some 7135's, the 2nd switch connects the remainded of the 7135's.

As I think about it, DarkZero has build a few sandwiches, perhaps he can confirm you're inside the parameters?

I don't believe the sst-90 has the peak lumens / mm2 that we're looking for. Imho it's a mis-application. The -50 is really nice under an aspheric, but I haven't had the set up to try it head to head with xre r2 or xpg (I need more hands).
Maybe someone with enough aspherics could test definitavely the performance of sst-50 vs r5

Even an r2 at 2amps is impressive (with heatsinking). But the intensitity is higher then a 6amp sst-90
 
Packhorse, you're finished aren't you?
Looking at the pic you posted:
If the tower from inside the head is taken out and instead attached firmly to the protruding battery post. This forms a new longer 'tower' sticking up from the battery tube stock. The new tower has the emitter sits at the top. (the battery and the driver has been moved to the canister because that's where you said you want the swtich, but this means a custom canister to disipate the intense heat of the 7135's?)
Lengthen threading inside the battery tube stock and increase the corresponding run of the threading up the bezel.
Put aspheric in the bezel. Screw it onto the stock, using the longer threading run to permit focusing the light while retaining waterproof integrity.
That is an incredibly simple design. To think the original post spawned a forum thread that contains mention of servo's and multiple lenses....
1>Take the insides from the head and attach them to the battery tube.
2>Use the original threading to become focus threading.

Simplifying it that much is remarkable.
 
Last edited:
The way I was thinking of doing the switching was to use a 2 mode quad AMC7135 that would drive 3 other quad AMC 7135 boards. This will give me 5.6 amps on full and 0.56 amp (10%) on low + I would then use another quad AMC7135 board that bypasses the 2 mode control chip.

This would give me 7 amp on high and 1.96 amp on low.
The thermal switch would be hooked up to the AMC's controlled by the 2 mode chip. If it goes into thermal shut down then the output would be 1.4 amp.

I wouldnt drive a R2 at 2 amps in such a critical application. The last thing I would want is failure in a deep dark wreck or cave.
But I get what you are saying about the intensity they produce. I know the SST-90 will not beat it but what I am trying to do is get close to it with a larger beam angle.

I have tried a R4 XP-G aspheric and its beam was less intense than a XR-E due to the dispersion angle of the LED. IMHO they are not suited to aspherics. SST's suffer the same problem but with the sheer grunt of lumen they produce I dont think it will mater too much.

My choice not to go with the SST-50 was based on my impression that it would not put out as much light than my current set up without over driving it.
 
Linger said:
Packhorse, you're finished aren't you?.................
Click to expand...
No. So far I have built the "handle" for the W200 head and the "tower" that the LED will sit on top of.

This tower is threaded at the base with the same 25.4mm x 1.5mm thread the W200 has.
So the tower threads down into the handle, the head threads down on top of it and as you said does the focusing. When screwed all the way down the beam will be at its widest. Unscrew it about 2 revolutions and it will be at its tightest. It will take another 6 or so revolutions to unscrew the heap totally so chances of unscrewing it right off by accident are minimal.
Drivers will be mounted inside the base of the tower as will the thermal switch.
minimax3.jpg
 
Last edited:
Nice size power lamp .....

But for driver ...... I 'm not recommanding multiple 71350

( as electrical engineer it is the same as if you want to run 12Amps machine with 12 1 Amps fuses. If one fuse fail becomes ''domino effect '' and all fuses fails one by one . Problem it is that maybe comes to some damage when the current it is not cut-off immediately )

Why not use some like :

http://www.candlepowerforums.com/vb/showthread.php?t=246371 . You can take a little overdimensioned FET and you are on safe.

If you tell me dimensions I can try to draw this circuit.

Order of some PCB it is simple here is the very simple program to draw it and than it is supported complete pruduction of few samples if you are not
familiar with fabrication of PCB.
http://www.expresspcb.com/ExpressPCBHtm/Download.htm

( but it is very simple special with laser printer and foil laminator)
just for info :
http://www.therecordingart.com/0925...B3-CB11-4D3E-9F61-D44449A434D3/DIY/DIYPCB.pdfhttp://www.therecordingart.com/0925...B3-CB11-4D3E-9F61-D44449A434D3/DIY/DIYPCB.pdf
 
lucca brassi said:
( as electrical engineer it is the same as if you want to run 12Amps machine with 12 1 Amps fuses. If one fuse fail becomes ''domino effect '' and all fuses fails one by one . Problem it is that maybe comes to some damage when the current it is not cut-off immediately )
Click to expand...
You may be right about not a good idea with multiple 7135's but your analogy is incorrect.

If a AMC failed it the total current being passed thru would reduce by 350ma.

If a 1amp fuse failed the current being passed thru would still be the same but the remaining fuses would carry more current each.

As for the other suggestions they are good by I dont want to go building a driver from scratch. Im looking for a simple solution. A resistor is the other option.
 
whitenoise said:
Yeah I did! 😳 😉
Click to expand...

Actuly I had put the idea on the back burner for several reasons but really wanted to build a small headed can light. Half way thru I though "If I just add this and do that it will be focusable!"

Its a long way from the original brief. eg focus will not be all that tight, no switching on the head and of course only a can light.

I think having a focuasble hand held will dictate it will be either very large, low run time or low output ( or maybe all 3).

So no, I will not be making a hand held version.
 
I have read this thread and see no reports of issues with multiple AMC's except heat. Since the Vin is relatively close to Vf I dont see this as a problem. Using 4 NiCd's or NiMh cells would be an issue though but they would allow the light to stay in regulation longer. I dont mind the light dropping out of regulation after say 2 hours. It will mean the run time will be increased as it draws less current.

So lets say VF is 3.7v and current draw is 7 amp. Battery pack is 20ah Li Ion. I woiuld guess it would stay in regulation for about 1.5 hours and then be close for another hour, after that it would be noticeablely dimmer but may run for another hour or 2 before the battery goes into protection.

This leads me into another question. I have 8 unprotected 18650's and I have a PCB that hap 2 of the protection IC's on it. Will this flow enough current? Can I parallel a second one up if needed?
 
Packhorse said:
This leads me into another question. I have 8 unprotected 18650's and I have a PCB that hap 2 of the protection IC's on it. Will this flow enough current? Can I parallel a second one up if needed?
Click to expand...

When I used some of these LiIon protection chips, they themselves had a max current rating since after all they are protecting the cells from over-discharge. Without knowing what circuit you have, it would be hard to tell the max. current before they trip. When they do trip, you have to remove power before they will re-set (at least the ones I used). This is one of the ones I used (if memory serves me right):
link at Battery Junction ...

They have other circuits with higher current ratings as well.

Will
 
Well Darkzero gave it a go-ahead in that thread

reading one post leads to another leads too ... the key point seems to be the controller to 7135 path, which concerns VDD and the enable pin. The controller uses pwm to turn the 7135's on / off with this and set the output mode.
It seems to run at a very low power level.

Additional learning - the Amtel controller chip needs 3.2 to work (2.6 but there's usually a diode with a .6v drop in from of it) and up to 7.0 (7.6 with the diode still present).

The 7135's are most efficient near led vf, and decrease with increasing voltages, while the waste heat increases at higher voltages. sellers list 6.0v as a max but Ktronik reports accidentally inserting a 3x3aa holder into a 3D light (over 10v) and the chips desoldered themselves with the heat but survived. Overvoltage condition is blinking but blinks are random (desyncronized) across multilpe chips.

The p (power, total voltage * ma to led) does not influence the controller chip. So controller doesn't know if it's 400ma out or 2.8a out. With reguardes to the current application, it means controller should be ok with 7A as well.

>>If you follow Packhorses link, there is schematic of these drivers. Can anyone test ma to a single 7135 through the vdd pin? Or if current increases connecting multiples? (I would make a mess of it)
 
I had a go at an ssr-50 + aspheric lens last night. I've obviously got the wrong lens for it (too small) but it does put out a pleasing amount of lumens. I don't have a Packhorse W200 to compare it too, but definately bigger / brighter / better then the Packhorse IST I've got.

In reguards to emitter domes dispersing too many lumens, have a look at saabluster's thread here:
http://www.candlepowerforums.com/vb/showthread.php?t=246857
redoes the emitter domes, to varing degrees of effectiveness.
 
it's an easy mod from Packhorse's can light to a hand light.
-You need the increased threading
-tower 'base' threaded so it secures into the stock instead of press-fit to base of head, spring used between lens / tower to ensure good contact pressure
-cut down the tower to proper height
-add a boost board and run it off 2nihm (or 2x 14500 :devil🙂

The shear brilliance of Packhorse's design idea, as it's evolved in this thread, is that the dead simple modifications from the w200. KISS, an excellent example of CPF innovation and ingenuity. The one hard step is if you can get the threading done. The rest is simple DIY stuff (simple being anything I can do)
 
Simple=good.

Which is good since my machining skills are simple!:crackup:

Yes it wouldnt be that hard to adapt my set up into a hand held but.... What LED to use? XR-E? With a single 18650 is will not provide alot of run time at full output. 2 14500's would be even less.
And then you need some way to switch it on and off.
And it would require even more work than my setup so cost would be more. (if you were to buy one).
 
You must log in or register to reply here.

Similar threads

D
APLOS AP70C dive light
Replies
4
Views
4K
ABTOMAT
A
S
Advice for refurbing a dive light
Replies
1
Views
4K
donn_
D
lucca brassi
some new build at divelight forum ?
Replies
0
Views
2K
lucca brassi
lucca brassi
D
CREE XML U4 diving light
Replies
0
Views
21K
DimitrisV
D
W
Soup can light, high amps with CC, is it possible?
Replies
11
Views
4K
DIWdiver
DIWdiver

Latest posts

Back
Top