DIY SST-50 focusable U/W light

[hanachan]

Hi,
ive just made up a prototype.

This light can be an ideal light for me.

lighting part:

left: for this light (SST-50) right: for R5 light already reported.

Spec:
size: 18cmx9cmx9cm (not include handle)
weight: not measured
LED: Cree SST-50 (DX says:shrug:, IMO by luminous)
mode: 5 :low, mid1, mid2, high, very high(direct drive)
battery: 4x18650 2500mAh 3.7V
regulator: 16xAMC7135
run time: not measured
waterproof: 30m depth (i hope)

i will use this on land for a while and make it completely waterproof.

thank you.

 

[hanachan]

making reprt

***LED***

i got SST-50 LED by purchasing the flashlight (SKU 32592) form DX. it was cheapest when i bought.
FYI: this light has NO driver or regulator! nothing on the PCB. but really bright one.

the LED was very tightly glued to the heat sink. so i got it by drilling the heat sink.

***LENS***

this is a frontal lens got from a junk telephoto lens of SLR. focus length is about 85mm. effective diameter 48mm.

moving lens mechanism is the same as previously reported ie using two magnet rings.
i tried to restrict the magnet filed with iron wires but no significant difference was observed.

***HEAT SINK & BRIGHTNESS CONTROL***

heat sink is 3 aluminum plate 1mm thickness.
regulators are 4x 1400mA AMC7135 board (SKU 1886).
left pict shows first wiring. i checked the current and found thin wiring was not enough
for much current and exchanged the wire for thicker one.

mode change mechanism is very simple.
low: 1x1400mA 7135 board
mid1:2x1400mA 7135 board
mid2:3x1400mA 7135 board
high:4x1400mA 7135 board are respectively effective.
very high: direct drive

i tested about its current not a little. ill report the results:

battery: 4x18650(2500mAh) parallel initial voltage 4.1V.
current measured by cheap DMM & usual cord

1) 4x7135board ->4.24A (expected current 5.6A)
2) direct drive--->5.24A over
3) 1x7135board-->1.41--1.45A (expect 1.4A)
4) 2x7135-------->2.9A (2.8A)
5) 3x7135------->4.1-4.15A(4.2A)
6) 4x7135------->4.41A(5.6A)

i made hall IC switch & tested it (FET:30V. 30A, 40mOhm)

the maximal current was 3.26A with 4x7135 boards.
double the FET 3.67A.
but using simply a toggle switch the current was 4.8A with the same wiring.

*no switch
only 2 FETs parallel:
1)1x7135board-->1.41-1.45A
2)2x7135board-->2.76-2.79A
3)3x7135board-->3.05A
4)4x7135board-->3.07A

1x FET:
1)1x7135board-->1.41-1.45A
2)2x7135board-->2.54-2.81A
3)3x7135board-->2.74A
4)4x7135board-->2.81A
final voltage is 4.07V.

i exchanged the thin wire fore thicker wire as shown at above pic.
1x7135board, 2x7135board same as with thin wire.
3x7135board-->3.68A
4x7135board-->4.07A

Conclusion:
thick wire is necessary to keep much current.
this FET can be used for 2x7135board (2.8A) around this voltage (4.1-4.07V).
a cheap toggle switch (125V, 3A) is very good.
a hall IC switch (i made) is interfered with wiring around 2.5A current.

so i decided to use a toggle switch for this light control.

***WHOLE ASSEMBLY***
connection of lighting, switch and battery parts was not so easy.
it is necessary to keep strength and to be disaasembled easily.

connecting plate were all hand made of aluminum.
explanation of assembly is difficult for me. sorry only show the picts.

after assembly i checked the current with full charged batteries (4.2V) again.
1) 1x7135board-->1.42-1.44A (expected 1.4A)
2) 2x7135board-->2.88-2.90A (2.8A)
3) 3x7135board-->4.35-4.40A (4.2A)
4) 4x7135board-->5.79-5.88A (5.6A)....over drive!
5) direct drive--->10A over!! .......umh dangerous!!!

thank you.

 

[hanachan]

ive just uploaded the beam shot on land of this light & DIY R5 focusable light.
please see the video at YOUTUBE.
content:
C2 MCE->R5 focusable->SST-50 focusable-> MCE (another background)
->both R5 & SST-50 max wide->both focusing->both max wide 1.4A driving
->both same area 1.4A driving

SST-50 looked more brighter than R5 directly seeing by naked eyes.

im satisfied with the result.

thank you.

 

[hanachan]

reserve: use under water

 

[hanachan]

i posted a making report. thank you.

 

[odessa.dk]

Nice work

 

[Packhorse]

Good stuff.
I love the way it looks like a SLR camera. So cool.

I also found similar problems using MOSFETs at low voltage. As yet I have no solution.

One problem I do see in your light is the focal length. At 85mm the set up is going to be very inefficient.
at 85mm with a 48mm lens you are probably losing half the lumen inside the barrel. Of course at wide angle the efficiency would be some what better.

 

[DM51]

Good stuff.
I love the way it looks like a SLR camera.

It looks just like a Hasselblad!

Nice work - I'll follow this thread with interest!

 

[wquiles]

Nice work :twothumbs

 

[hanachan]

thank you for the comments.

I also found similar problems using MOSFETs at low voltage. As yet I have no solution.

i think main issue is low voltage. i want a MOSFET working at low voltage (for example 3.3V)
with low resistance (some mohm) which may resolve the issue.
but it is very difficult to get such a MOSFET.
in practical view point i will test the MOSFET with very low resistance
which i can get activated with high gate voltage (about 5V) by another battery.

One problem I do see in your light is the focal length. At 85mm the set up is going to be very inefficient.
at 85mm with a 48mm lens you are probably losing half the lumen inside the barrel. Of course at wide angle the efficiency would be some what better.

very good pointing out.

a lens with longer focal length has to be placed far from the emitter to focus,
resulting loss of light as you say. but its size (diameter) is usually large
so the light amount the lens can get is not so little. and it can concentrate much light
than the lens with shorter focal length (the lens with longer focal length can create a smaller real image).

i will discuss about this issue when i post the beam shots on land.

 

[DIWdiver]

thank you for the comments.

i think main issue is low voltage. i want a MOSFET working at low voltage (for example 3.3V)
with low resistance (some mohm) which may resolve the issue.
but it is very difficult to get such a MOSFET.
in practical view point i will test the MOSFET with very low resistance
which i can get activated with high gate voltage (about 5V) by another battery.

QUOTE]

Having worked with them for years, I don't find it hard to select FETs with less than 10 mOhm resistance at 3.5V (or even 3.0V) gate voltage. Here are some examples in stock at www.digikey.com:

PSMN1R2-25YL
IPD031NO3L
PSMN3R5-30YL

It's not obvious from the description how they will act at low gate voltage, but there is usually a graph in the data sheet. Maybe if you post (or PM me) some specifics about what you are looking for (package, voltage, resistance, etc), I can help you select a good FET?

D

 

[hanachan]

@DIWdiver
thank you for a kind message.

i saw the pdf about PSMN1R2-25YL the fig.7 shows over 90A current flows
at Vds=3.5V & Vgs=3V. this MOSFET seems to work well in my low voltage condition.

do you know how & where i can buy small number of this stuff as cheap as possible?
i can find out some MOSFETs which may be suitable for my purpose
but they are sold only a lot 100-1000 etc. i need only two or three.

 

[DIWdiver]

do you know how & where i can buy small number of this stuff as cheap as possible?

If you are really in Japan, I'm not going to be much help, being hopelessly American myself. The link I suggested in my previous post sells them for $3.02USD each, but shipping would be expensive! But there must be similar places in your area.

D

 

[Packhorse]

Is there a better way? When direct driving ( or resistor/ AMC 7135) a SST even 1/10 of a volt drop makes a huge difference to the output of the LED. I thought about a relay driven by a reed switch but I cant find one that has a low enough operating voltage and high current capacity. Would a transistor set up be better?


I have also been working on a adjustable focus system but I keep coming up with problems. The problem I have with a fixed LED and a moving lens is that for a wide angle beam the lens has dropped down the barrel and further away from the front lens. What happens here is the barrel will now shadow the beam reducing the wideness of it. If the aspheric was fixed at the front just behind the front sealing lens and the LED moves instead this problem is greatly reduced. Although it does raise other issues. I guess its not that much of a problem if you dont want a super wide angle but for a video light I think it is a problem.

 

[Packhorse]

I tried IRFLZ34NPBF's.
Although they work they still drop a few tenths of a volt.

 

[gcbryan]

...

I have also been working on a adjustable focus system but I keep coming up with problems. The problem I have with a fixed LED and a moving lens is that for a wide angle beam the lens has dropped down the barrel and further away from the front lens. What happens here is the barrel will now shadow the beam reducing the wideness of it. If the aspheric was fixed at the front just behind the front sealing lens and the LED moves instead this problem is greatly reduced. Although it does raise other issues. I guess its not that much of a problem if you dont want a super wide angle but for a video light I think it is a problem.

I don't have any solutions really other than to point out that the flood to throw lights on DX move the entire head so the barrel effect that you are describing doesn't apply so much.

It would bring new problems however in sealing the sliding head as the DX lights using this method aren't dive lights.

Also, regarding reed switches ... aren't hall effect sensors another solution to the same problem? I don't fully understand them so I really am asking.

 

[Codiak]

Also, regarding reed switches ... aren't hall effect sensors another solution to the same problem? I don't fully understand them so I really am asking.

nice find though the voltage range is good, does it have the same issue with current?

I may call digikey tomorrow

 

[Packhorse]

I have not used hall effect sensors before. From what I understand they will only switch a certain voltage out. I think if you then feen this to a MOSFET the MOSFET will only put out the same voltage so you will need some more circuitry to fix this.
I could be totally wrong though, it wouldnt be the 1st time.

 

[Packhorse]

Yes, sliding the whole head would solve that problem but introduce several others.

1st is moving parts that need to seal.

2nd is that as the head moves forward the pressure inside reduces and as it moves back the pressure increases.
When you dive the deeper you get the more external pressure is exerted on the light This will act on the head and push it down on the body changing the focus. You could have some sort of locking system but its not ideal. Or you could have it threaded but it may be very hard to turn while under pressure.

 

[gcbryan]

Yes, sliding the whole head would solve that problem but introduce several others.

1st is moving parts that need to seal.

2nd is that as the head moves forward the pressure inside reduces and as it moves back the pressure increases.
When you dive the deeper you get the more external pressure is exerted on the light This will act on the head and push it down on the body changing the focus. You could have some sort of locking system but its not ideal. Or you could have it threaded but it may be very hard to turn while under pressure.

I wonder if you could vent it. Seal the emitter but let water into the moving head. This would eliminate the pressure issue but I guess it would distort the beam angle?

I also thought Hall effect sensors could pick up rate of change. I'm totally talking out of the side of my mouth here but the ugly light switches to strobe mode only when you rotate the ring in a certain direction and at a speed that is great than normal. I thought this was possible with Hall effect sensors but maybe not.