My new high-power home-made "Focussing lights"

Joyful · Post by **Joyful** » Sun Jun 28, 2015 3:37 am

I have found that my old led strip light was not powerful enougt to give me a bright image in my viewfinder when going down to 20x magnification.

If Jansjo is available in Cape Town, I have not found a supplier, so I went to my favourite electronics shop and bought me 3 x 10W 12V Rectangular high Power LEDs. I managed to locate a 12v power supply from Eurolux. Regrettably, my first switch-on resulted in an intense light, and a blown LED 2 minutes later ! Sadly, the current was way too high. So I experimented with 10W resistors, until I got the current to 850ma - which is about 50ma below the recommended 900ma current.

The stands are from a strip of 38mm x 3mm Aluminium (Aluminum to you guys!) which doubles as a very efficient heat sink.

See the pics ! I can now take pics by these lights at about 10x at 1/60th sec, 100ISO and f8.

Here are the tech specs for those who are interested
Power Dissipation : 10W
Forward Voltage : 9~12V
Forward Current : 900mA
DominantWavelength (tc) : 6000K-7000K
Luminous Intensity : 900~1000 lumens
View Angle : 120~140°
Life Span over : 100,000 hours
Operating Temperature : -20~+80°C

ChrisR · Post by **ChrisR** » Sun Jun 28, 2015 4:42 am

Way to go

- it's almost alarming that these things are coming out of China in 100W 10,000Lumen versions for about $5. 36V power supplies will have to catch up.

A few questions arise:

Have you measured the actual on-chip voltage? I wonder if you had a power supply which was over 12V at the load you were using?
I couldn't help wondering also, about the battery pack in the picture

.

How do you find the colour reproduction across a spectrum - assuming you've custom-white-balanced your camera to the nominal? There are of course grades for CRI - but I don't now how critical it really is.

I'd be concerned about the heat dissipation and the temperature actually ON the chips. It may be possible to monitor the volt vs current figures while they're warming up, to work back to what the temperature is likely to be.

10x at 1/60th sec, 100ISO and f8.

Sounds encouraging. (..though "f8?? 10x at NA 0.25 would be f/20..) There's obviously scope to go shorter.

Joyful · Post by **Joyful** » Sun Jun 28, 2015 4:46 am

Hi Chris - Thanks for the reply !

The battery pack in the background was an oops - it belongs to my strip led lighs !

I have not measures on-chip voltages - just the current, will check and let you know !

Joyful

TheLostVertex · Post by **TheLostVertex** » Sun Jun 28, 2015 11:03 am

ChrisR wrote:Way to go - it's almost alarming that these things are coming out of China in 100W 10,000Lumen versions for about $5. 36V power supplies will have to catch up.

A few questions arise:

Have you measured the actual on-chip voltage? I wonder if you had a power supply which was over 12V at the load you were using?

I have watched several videos on youtube in the past of people testing high wattage LED assemblies out of china. Frequently they are shown to draw much less(about 60% of rated or so). Likely just being from them shipped with badly rated drivers. Still worth watching out for if you are actually expecting a certain power draw. With something crazy like a 100w LED, I imagine heat dissipation will be a serious issue.

Joyful, have you decided on how you are going to diffuse the LED's?

Joyful · Post by **Joyful** » Mon Jun 29, 2015 3:16 am

Hi Steven -

Re diffusers - I have cut off the top portion from a white milk bottle which has given me best results at high magnifications.

Then I have made a 4-sided cube of white perspex 100 x 100 x 100mm for bigger things, then another similar cube of 150 x 150 x 150 for larger items.

My real plan is to use the "focussing lights" to find my start and end points, and my composition, then to use the Canon Macro Twinlite and the 430 flash for actual lighting - placed as close as possible to where my focussing heads were.

Will post results when I get something good enough !

Joyful

mawyatt · Post by **mawyatt** » Mon Jun 29, 2015 6:38 am

Very nice setup.

Being an electrical engineer (IC designer) but not an LED expert, I would think the LED would operate better (safer and with less variation) when operated under constant current conditions rather than constant or somewhat constant voltage condition. An LED is basically a semiconductor junction, in Gallium Nitride (GaN) for these types, and are fundamentally current mode devices.

I am sure if you Goggle this you will find some recommended constant circuits for driving these LEDs.

A resistor in series with a voltage source is not a very good current source unless the voltage source is much higher than the On voltage of the LEDs. The LED current can be estimated as the supply voltage minus the LED On voltage divided by the resistance. If the supply voltage is not much higher than the LED On voltage the current becomes highly dependent on the LED On voltage (not good). If the supply voltage is much higher than the LED On voltage, then the current becomes dependent on the series resistor (better). However this resistor will dissipate the current squared times the resistance in heat...so very wasteful!!

A much better solution IMO is a current source which keeps the LED current independent of the LED On voltage, or the supply voltage (or even temperature), all within reason of course.

Anyway, hope this helps.

Mike

ChrisR · Post by **ChrisR** » Mon Jun 29, 2015 3:29 pm

Mike, am I right in thinking that as the temperature goes up the forward voltage drops, so they take more current, get hotter etc? ie "Runaway".

mawyatt · Post by **mawyatt** » Mon Jun 29, 2015 4:58 pm

Chris,

Bipolar transistors are noted for this, the ratio of collector to base current called beta increases with temperature, it's called thermal runaway and is destructive if not limited. Most semiconductor junctions have a negative temperature coefficient, which as you indicated causes the LED On voltage to decrease with temperature. If the series resistance is small, which will happen with a low supply voltage near the LED On voltage, then as the On voltage decreases the LED current increases. The power dissipation will increase if the ratio of current increases faster than the On voltage drop, since power is the product of voltage times current. Increases in power dissipation will cause higher temperatures and lower On voltage causing runaway.

For example if the supply voltage is 13v and the nominal LED On voltage is 12v at room temp and the desired LED current is 1 amp. The resistor will be 1 ohm. Now if the LED has a -20mv/C temperature coefficient and the LED is dissipating 12 watts which causes it's internal temperature to rise 50C, then the new On voltage is 12 -50*0.02, or 11 volts, the current will now be (13-11)/1 or 2 amps. So the new power dissipation is 11*2 or 22 watts. This will cause the LED temperature to rise even more and cause more current to flow and eventually possibly destroy or damage the LED.

A constant current supply doesn't have this problem, nor does using a much higher supply voltage and larger resistor. In the above example if the supply voltage was 24 volts, then the resistor would be 12 ohms for 1 amp LED current. The LED temperature rise will still be 50C and the ON voltage drops to 11 volts, and the current increases to (24-11)/12 or 1.08 amps with a power dissipation of 11.88 watts, so the dissipation actually dropped!! I should add that the resistor will dissipate 14 watts under these conditions, so simple but not the most efficient means to power the LED.

Anyway, hope this helps,

Mike

Chris S. · Post by **Chris S.** » Mon Jun 29, 2015 6:14 pm

We should probably move this interesting thread to the Technical Discussion forum.

--Chris

Joyful · Post by **Joyful** » Tue Jun 30, 2015 12:08 am

Thanks for all the technical info - I will keep it in mind for my next upgrade.

Just for fun, I call myself a BSc Electrical Engineering Second Year (Failed) - way back in 1966 when LASERS were a gimmick no-one knew what to do with !

Joyful

ChrisR · Post by **ChrisR** » Tue Jun 30, 2015 1:10 am

Mike - thanks for the analysis. I have vague recollections of that from the dim and distant past!

Joy - some resonance (;)) there.

mawyatt · Post by **mawyatt** » Tue Jun 30, 2015 6:44 am

Joyful wrote:Thanks for all the technical info - I will keep it in mind for my next upgrade.

Just for fun, I call myself a BSc Electrical Engineering Second Year (Failed) - way back in 1966 when LASERS were a gimmick no-one knew what to do with !

Joyful

I am not far behind, got out of HS in 66. We used slide rules & CRC books for math tables in college. I remember when a student I knew showed up with an hp35 scientific calculator...wow!!!

mawyatt · Post by **mawyatt** » Tue Jun 30, 2015 6:49 am

ChrisR wrote:Mike - thanks for the analysis. I have vague recollections of that from the dim and distant past!

Joy - some resonance (;)) there.
I've found I've volunteered at a local school to help a biology teacher take on 13 year old girls with an electronics course. The tech stuff I can handle, there are books if I need them. Not about the girls though.

I was a adjunct professor for while back in 2000-2003 creating and teaching a couple graduate level courses in IC design, a lot of work!! But I enjoyed it working with kids (well they are kids to me). I ended up hiring my best students, which was a plus!!

Charles Krebs · Post by **Charles Krebs** » Tue Jun 30, 2015 11:10 am

Yup... the LED offerings get more interesting ever day.

There are a great many power supplies available as well. Don't know about availability and supply where you are, but ones I really like to use are seen here:
http://www.ledsupply.com/led-drivers/bu ... ed-drivers
http://www.ledsupply.com/luxdrive-const ... ed-drivers

Not too expensive, easy to set up, and dimmable with a simple POT.

My new high-power home-made "Focussing lights"

My new high-power home-made "Focussing lights"

Re: My new high-power home-made "Focussing lights"