[PSUBS-MAILIST] LED Light test

[Alan via Personal_Submersibles]

Thanks Steve,
that's great to know about the heat on the acrylic lens & your testing.
At the moment I am driving my lights with a buck / boost constant current driver 
that has a 60V input for my 48V system. It took a lot of work tracking down a 
unit that was suitable for 48V. 
A friend put me on to those linear regulators & is trying to convince me to go
with them. I will keep your information in mind should I need to argue my way
out of them.  They would be good for a Sub on a 36 V system, as most of the
high power LEDs are 36V.
Have a look at the flip chip LED I'm using if you haven't already seen it.
Smaller than the usual cob LEDs. 
http://www.tyf-led.com/downloadRepository/82befdc2-15b3-4ff0-a553-6f53a4d8e404.pdf
Got mine with pigtails on for $6:50- each.
Cheers Alan



Sent from my iPad

> On 9/03/2017, at 9:54 PM, Stephen Fordyce via Personal_Submersibles <personal_submersibles at psubs.org> wrote:
> 
> Hi Alan,
> I'm glad your lights are going well!  There are a couple of things I may be able to contribute after several iterations of designing the circuit for my TFM caving/diving helmet lights.
> 
> 1. Linear vs switching LED drivers. 
> I looked at the datasheet for your LED driver chip and it appears to be a linear regulator - this in general terms means it creates a variable resistance to get the output current/voltage right, and as such can create a lot of heat which may be wasteful.  Ie. if you have a 12V supply and your LED needs 4V, 67% of the power will be lost in the driver!  But if you have a 12V supply and your LED chain needs 11.5V, then it's quite efficient.  They are simple and easy, and small as you point out, with no external components.
> 
> A switching driver uses very fast (MHz) PWM along with an external capacitor/inductor to drop and smooth out the voltage, and is ~90+% efficient over a big range of voltage drops.  This means it creates much less heat and saves power.
> 
> I changed to switching drivers for my LEDS (input 7-8V, output 3-4V) and noticed a huge difference in heat and efficiency.  Just something to be aware of if you run into heat or runtime problems.
> 
> 2. Acrylic windows
> I have used lasercut acrylic (6mm) in my caving lights (which I've had up to 80degC according to the internal temp sensor.  Lasercutting allows me to get an oblong rectangular shape with holts for screws.  The lights have a full high temperature burn test out of the water and then multiple dives to 100m in my pressure pot, with no acrylic failures to date.
> 
> Cheers,
> Steve Fordyce
> 
>> On Thu, Mar 9, 2017 at 6:12 PM, Alan via Personal_Submersibles <personal_submersibles at psubs.org> wrote:
>> Tested my 80W LED light in a housing, to see how hot it got in & out 
>> of the water.
>> The good news was it didn't get warm at all after running at 70W for
>> 25 minutes in the kitchen sink. Out of the water it hit 120F (49C) in 7
>> minutes. The LED has a maximum temperature rating of 60C so was keeping
>> back a bit from that. 
>> I dropped the current down & ran it out of the water at 33W & 25W but it climbed 
>> to 125F in 20 minutes for 33W & 15 minutes for the 25W.
>> I am using a cast acrylic 8mm thick lens, & this didn't get hot either in air
>> or water. Using acrylic instead of glass is saving me $90-.
>> Am making a few minor housing design adjustments for the final iteration.
>> Have also ordered a dozen of these 50 V 350 mA constant current regulator &
>> LED drivers. 
>> 
>> http://www.onsemi.com/PowerSolutions/product.do?id=NSI50350AD
>> They are TINY; about 1/4" square & you run them in series to get the desired
>> amperage. They have a wide input voltage & would be great for a 36V system.
>> Alan
>> 
>> 
>> Sent from my iPad
>> 
>> _______________________________________________
>> Personal_Submersibles mailing list
>> Personal_Submersibles at psubs.org
>> http://www.psubs.org/mailman/listinfo.cgi/personal_submersibles
> 
> _______________________________________________
> Personal_Submersibles mailing list
> Personal_Submersibles at psubs.org
> http://www.psubs.org/mailman/listinfo.cgi/personal_submersibles
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.whoweb.com/pipermail/personal_submersibles/attachments/20170309/eaaa3f9f/attachment.html>