Author Topic: 10W spot light hack/constant current  (Read 4423 times)

Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 994
  • Kudos: 13
10W spot light hack/constant current
« on: November 11, 2013, 06:34:36 PM »
I just bought some of these http://www.holidaycoro.com/product-p/842.htm and plan to mod my spotlight http://www.ebay.com/itm/New-LED-Underwater-Spot-Light-10W-12V-900-1000LM-Pure-White-for-Aquarium-Pool-/130896510781 to rgb and run from a DSC.  It has been said the DSC and MR16 are "constant current" and I infer that I can run a raw led from them?  Thoughts?  THANKS! 
Sam
 Last year's video: https://vimeo.com/150560653

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,303
  • Kudos: 127
    • Granbury Christmas Lights
Re: 10W spot light hack/constant current
« Reply #1 on: November 11, 2013, 08:31:40 PM »
http://diylightanimation.com/index.php?topic=3356.msg64034#msg64034

refers you to the DLA Wiki for the MR16, it does tell you how to build or modify for constant current

http://www.diylightanimation.com/wiki/index.php?title=Manual_MR16


For the 350ma used by HolidayCoro's mistake LED, the resistor to be used is 112 ohms. (using the datasheet formula in Table1 of the TI page)

Jon
« Last Edit: November 11, 2013, 08:43:29 PM by JonB256 »

Offline smeighan

  • Developer
  • ******
  • Join Date: Mar 2013
  • Location: Highlands Ranch, Colorado
  • Posts: 1,035
  • Kudos: 11
    • Nutcracker123
Re: 10W spot light hack/constant current
« Reply #2 on: November 11, 2013, 09:39:36 PM »
Dirknerkle sent me this web  page
http://led.linear1.org/1led.wiz

I have some 1W leds i bought for $0.42 each.
http://www.aliexpress.com/store/product/promotion-1W-high-power-led-100-110lm-35mil-led-chip-taiwan-epistar-chip-with-heatsink/701799_545853431.html

I am going to use them as strobes on my 8' holdman star

here were tehir specs

Power: 1W
Beam angle: 120 degrees
Forward current: 350mA
Forward voltage: 3 to 3.6V


i am using 47 ohm 2W resistors to give me 200ma current. i tried 50ma to 300ma by varying the resistance. i am driving them with the 27ch 12v dmx controller that ray sells.

man, are they bright.

i ordered cases from macebobo that you can find in this thread
http://www.diychristmas.org/vb1/showthread.php?1087-1W-Star-LEDS-and-cases-for-Dirk-Cheap-Strobes-and-other-uses

he sells the cases for $0.30 each


thanks
sean
Sean
Littleton, CO Latest releases http://nutcracker123.com/nutcracker/releases xLights/Nutcracker Forum http://nutcracker123.com/forum/index.php Facebook [url=https://www.facebook.com/groups

Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 994
  • Kudos: 13
Re: 10W spot light hack/constant current
« Reply #3 on: November 13, 2013, 07:11:05 AM »
Thanks guys.  Jon, do you think the DSC could be modded to constant current?  Looks like same topology.  Remove caps/resistors/mosfet jumpered and replace 4.7k resistor to 112 ohms.  Can I ask how you calculated the 112 ohms?  Still learning. 

Sean, thanks for the links.  I tried to get another member over at DIYLA to do some strobe boards.  I love down in the dirty DIY over at DIYC.org just wish I could follow the posts...  Way over my head.  I like Komby's work on a board that runs a strobe routine.  I would prefer not to sequence strobes and just turn a controller on to do it, plus I think it would be more random and not limited by sequencer playback. 

OK, now I could care less about RGB spot lights and want 32 1W led strobes!

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,303
  • Kudos: 127
    • Granbury Christmas Lights
Re: 10W spot light hack/constant current
« Reply #4 on: November 13, 2013, 07:20:28 AM »
Can I ask how you calculated the 112 ohms?  Still learning. 

using the datasheet formula in Table1 of the TI page -

http://www.ti.com/lit/ds/symlink/tlc5940.pdf   (scroll down to Table 1, the "Iout(ideal)" formula)

solving for RIREF: you put in 350ma (0.350) for IOUT, divide by 31.5, divide by 1.24, then take the inverse to solve for RIREF

ps - did I ever mention that I was a math major and math teacher many years ago? :)

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,303
  • Kudos: 127
    • Granbury Christmas Lights
Re: 10W spot light hack/constant current
« Reply #5 on: November 13, 2013, 07:39:40 AM »
Jon, do you think the DSC could be modded to constant current? 

I'm doubtful. The reason the MR-16 can be modified to Constant Current is because of the TI LED control chip (TLC5940) referenced above.
The DSC does not have that chip. There is a "lone" resistor on the DSC right next to the Lynx Logo. That might be a reference resistor in use by the PIC chip to set a current value.

But, since that would be a mod to RJ's board, you can't ask it there. And I only own one DSC and am not currently willing to try a mod.



Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 994
  • Kudos: 13
Re: 10W spot light hack/constant current
« Reply #6 on: November 13, 2013, 07:47:37 AM »
Thanks Jon.  I assume now that if I try and run the 10w led with a DSC I will fry it?  Not sure how I got off track, but somewhere I read the mosfet is what made the MR16 and DSC constant current.  Oh well...

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,303
  • Kudos: 127
    • Granbury Christmas Lights
Re: 10W spot light hack/constant current
« Reply #7 on: November 13, 2013, 09:12:39 AM »
The "deal" about using a Constant Current supply is that your attached LED doesn't need a current limiting resistor - that is handled by the circuit.

So, with constant current, you control brightness by switching the current on and off.  Always on is max brightness. 50% bright would be getting current pulses 50% of the time. But it is always the same peak current, just on less time for dimming.

With constant voltage supplies, you attach a device that is pre-built / designed to run at a specific voltage. 12vdc or 5vdc etc.
Then the Voltage is switched on and off in what looks like a square wave. But it is always the same peak voltage, just on less time for dimming.

About the MOSFETs. In the MR-16 and the DSCs, normal builds, the MOSFETs have 12vdc applied on them all the time. There are three legs to a MOSFET - A Source, a Drain and a Gate. The 12vdc is applied across the Source and the Drain, but it will only conduct when a proper Gate signal is applied. In these circuits, the Gate signal is always large enough to cause the MOSFET to conduct 100%. But you can control how long the Gate signal is applied to control brightness of the attached LEDs.

So, the MOSFET is always full ON or OFF. The reason for that is primarily heat. A Full ON MOSFET has minimal resistance to current flow, so less heat. A MOSFET that is OFF has virtually no current, so no heat. You'll notice that MOSFETs rarely have heat sinks. Not needed.
« Last Edit: November 13, 2013, 09:44:28 AM by JonB256 »

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,303
  • Kudos: 127
    • Granbury Christmas Lights
Re: 10W spot light hack/constant current
« Reply #8 on: November 14, 2013, 08:57:26 AM »
zwiller - curious if my last post was helpful at all. It got a little technical but I thought it might help you decide which way to go.

JonB

Offline arw01

  • Hero Member
  • *****
  • Join Date: Oct 2013
  • Location:
  • Posts: 891
  • Kudos: 0
Re: 10W spot light hack/constant current
« Reply #9 on: November 14, 2013, 02:24:10 PM »
The "deal" about using a Constant Current supply is that your attached LED doesn't need a current limiting resistor - that is handled by the circuit.

Ok, so in I have no idea what I am talking about, one does not NEED constant current, just something to limit the upper end of the current so the led does not burn itself out?

At some dropped amount of current the led will not light if I recall anything I ever read on electronics.

e.g. a DSC doesn't using timing to dim, it drops the voltage, and by theory if we (my wag) drop voltage too much the led will just be off and if we just give it a bit of voltage it may be dim and likely not linear between the colors.

Or any I completely and utterly off base here?

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,303
  • Kudos: 127
    • Granbury Christmas Lights
Re: 10W spot light hack/constant current
« Reply #10 on: November 14, 2013, 03:33:21 PM »

Ok, so in I have no idea what I am talking about, one does not NEED constant current, just something to limit the upper end of the current so the led does not burn itself out?

At some dropped amount of current the led will not light if I recall anything I ever read on electronics.

e.g. a DSC doesn't using timing to dim, it drops the voltage, and by theory if we (my wag) drop voltage too much the led will just be off and if we just give it a bit of voltage it may be dim and likely not linear between the colors.

Or any I completely and utterly off base here?

LEDs are "current" devices, as are most semi-conductor devices. For the LED in question, 100% bright is 350ma. You could push more through it, but it would shorten the life and not be noticeably brighter. For any LED, the best thing to do is play with the Brightness control to find the lowest value where it still looks like 100% bright. For some, that may be as low as 70% (70% of 255 for DMX would be 178).

http://www.granburychristmaslights.com/DCx16.html

These pictures are from a DC controller using MOSFETs. They are running at 12vdc, but notice that the "half-power" picture, the LED is only ON half the time, so it looks less bright. This is a Constant Voltage controller like the standard MR16 or the DSC.

A Constant Current output would look just the same. The dimming comes, not from reducing the actual current, but by pulsing it with shorter pulses. The frequency output is usually around 400hz, so you can't see the pulses.

Your first comment about an LED turning OFF as the current gets low is correct. There is a threshold value. Rather than worry about where that low value is, LED dimming circuits generally choose to pulse the lights rather than reduce the current. 

Have you seen newer cars whose tailights seem to be flashing, especially when you turn your head while looking? They are pulsing the lights to control brightness, but at something much lower than 400hz, and that's why you can see it.

Finally - 120vac LED strings. They actually are dimmed by reducing the amount of current through them. The AC controller actually drops the voltage, which drops the current, until it gets low enough that the LED goes out. It is why AC powered LED strings are nowhere near as linear as Incandescent lights. The brightness of an LED is just not linear as current drops.
« Last Edit: November 14, 2013, 03:40:57 PM by JonB256 »

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,303
  • Kudos: 127
    • Granbury Christmas Lights
Re: 10W spot light hack/constant current
« Reply #11 on: November 14, 2013, 03:59:03 PM »
And another thing -

If you use a Constant Current driver for these 10W LEDs, you can power more than one of them per output channel by connecting in series.

If your DC supply was 12volts, you could probably power 3 of them and get full brightness from all 3. Because they are in series, if one of them has 350ma passing though it, then they all have 350ma thru them. But, if you tried 4 of them in series, they would never get more than 3 volts and that may not be high enough to get 350ma through the circuit.

If you do a search for Constant Current LED Drivers, you can find some for 350ma output. One I looked at was powered from 120VAC and it says the output is a maximum of 48volts DC.  That means you could probably drive 12 LEDs in series before maxing out the circuit.

Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 994
  • Kudos: 13
Re: 10W spot light hack/constant current
« Reply #12 on: November 15, 2013, 12:43:15 AM »
Jon,
Thanks for taking the time to explain.  Very informative and helpful.  I really enjoy learning the finer details of our blinky hobby.  You raise an interesting point about running multiples.  That's good to know, but I think most of the folks would want these discrete.  Seems like the MR16 setup for CC is the best tool for the job.  What is your opinion of adding current limiting resistor to a DSC to hack?  I think I am following this for the most part, but now lost on the fact that rgb has 3 different forward voltages and that the DSC is 12V common. 

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,303
  • Kudos: 127
    • Granbury Christmas Lights
Re: 10W spot light hack/constant current
« Reply #13 on: November 15, 2013, 06:39:33 AM »
What is your opinion of adding current limiting resistor to a DSC to hack?

I don't think the DSC can be modified to be Constant Current like the MR-16. The only active component on that board is the PIC chip and it does not have the circuitry like the TI chip on the MR-16.

Quote
lost on the fact that rgb has 3 different forward voltages and that the DSC is 12V common.

Yes, an RGB strip does have different forward voltages. Red has a different value than Green and Blue. That is why, if you look very closely, the current limiting resistors on a 12vdc or 5vdc RGB device have different values. The designer has pre-calculated the correct value so that all you have to do is apply the correct voltage and then not worry about the current. As long as you don't exceed the design voltage by much (13.2vdc is usually fine, but 24vdc is too high) then the RGB works and lasts a long time.

Offline zwiller

  • Falcon Beta Team
  • **
  • Join Date: Mar 2013
  • Location: Sandusky, OH
  • Posts: 994
  • Kudos: 13
Re: 10W spot light hack/constant current
« Reply #14 on: November 15, 2013, 08:44:33 AM »
I was thinking I could throw a current limiting resistor on the 12V feed to the led and that would allow me to use the DSC?  That said, I am warming up to using the MR16 and am starting to get crafty about a wiring scheme...

 

Back to top