Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Driver for Cree CXA1304
77 views
Skip to first unread message
Al Grant
Jul 7, 2021, 5:08:15 AM7/7/21
to
I had a go at creating a DC-DC driver for a pair of Cree CXE1304's with a high/low switch.
This is for a lighting project on a boat, so supply is 12.5-14.8 DC.
I tried one with a LM2940-T 10v but I could not get the heat under control on either the LED's or to some extent the LM2940 - I tried to add resistor on the input to the 2940 which reduced overall current to 0.8A but it just made the resistor the new heat problem.
Is there a circuit I can create to drive two of these 1304's? Preferably one where heat is not such an issue.
Thanks
Al
This is for a lighting project on a boat, so supply is 12.5-14.8 DC.
I tried one with a LM2940-T 10v but I could not get the heat under control on either the LED's or to some extent the LM2940 - I tried to add resistor on the input to the 2940 which reduced overall current to 0.8A but it just made the resistor the new heat problem.
Is there a circuit I can create to drive two of these 1304's? Preferably one where heat is not such an issue.
Thanks
Al
Al Grant
Jul 7, 2021, 5:15:28 AM7/7/21
to
Photo of schematic here : https://photos.app.goo.gl/ZKeafiZwcSF87XUv5
Rick C
Jul 7, 2021, 8:29:07 AM7/7/21
to
I don't have time to find a part for you and I don't have the specs on the LED, but you should be able to find switching LED drivers if you look around for them. Using a switching driver will solve the heat problem in the regulator. The only thing you can do about the heat in the LED is to reduce the current along with the light output.
--
Rick C.
- Get 1,000 miles of free Supercharging
- Tesla referral code - https://ts.la/richard11209
jla...@highlandsniptechnology.com
Jul 7, 2021, 9:56:17 AM7/7/21
to
On Wed, 7 Jul 2021 02:08:11 -0700 (PDT), Al Grant <biga...@gmail.com>
wrote:
Is that a 9 volt part? How much current do you want?
A switcher would be efficient, but a bit more trouble than a linear
reg. The LEDs will need to be kept cool no matter what regulator is
used.
If you're only dropping a few volts, and need to heat sink the LEDs
anyhow, a linear reg should be OK.
You could just buy some 12 volt LED bulbs with built-in regs, but no
design fun. The car side/turn light replacement bulbs are cool.
--
John Larkin Highland Technology, Inc
The best designs are necessarily accidental.
wrote:
A switcher would be efficient, but a bit more trouble than a linear
reg. The LEDs will need to be kept cool no matter what regulator is
used.
If you're only dropping a few volts, and need to heat sink the LEDs
anyhow, a linear reg should be OK.
You could just buy some 12 volt LED bulbs with built-in regs, but no
design fun. The car side/turn light replacement bulbs are cool.
--
John Larkin Highland Technology, Inc
The best designs are necessarily accidental.
Al Grant
Jul 7, 2021, 10:17:46 AM7/7/21
to
Yes there would be no fun if I just went and brought a part ready to go, although it is getting a little frustrating now.
Are you saying that for a given current, the LED's will reach a certain temperature T which will be the same, regardless of whether driven by linear or PWM?
So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
I have a heat sink on the LM2940 which seems to be coping at about 50 deg C.
jla...@highlandsniptechnology.com
Jul 7, 2021, 10:32:06 AM7/7/21
to
On Wed, 7 Jul 2021 07:17:43 -0700 (PDT), Al Grant <biga...@gmail.com>
switcher. Most of the power applied to an LED gets converted to heat.
>
>So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
Very wild guess for that little piece of aluminum, 30 K/W unless it's
glued to something bigger. Each LED is 9 volts * 0.8 amps = 7.2 watts.
2 of them is 14 watts. Say 10 watts of heat, since some of the power
comes out as light.
30 K/W * 10W is a temp rise of 300C.
The numbers are rough, but suggest the problem.
Worse, LEDs get less efficient when they are hot. Spritzing an LED
with freeze spray can be dramatic.
>
>I have a heat sink on the LM2940 which seems to be coping at about 50 deg C.
Higher supply voltage will make the reg dissipate a lot more power.
wrote:
>On Thursday, July 8, 2021 at 1:56:17 AM UTC+12, jla...@highlandsniptechnology.com wrote:
>> On Wed, 7 Jul 2021 02:08:11 -0700 (PDT), Al Grant <biga...@gmail.com>
>> wrote:
>> >I had a go at creating a DC-DC driver for a pair of Cree CXE1304's with a high/low switch.
>> >
>> >This is for a lighting project on a boat, so supply is 12.5-14.8 DC.
>> >
>> >I tried one with a LM2940-T 10v but I could not get the heat under control on either the LED's or to some extent the LM2940 - I tried to add resistor on the input to the 2940 which reduced overall current to 0.8A but it just made the resistor the new heat problem.
>> >
>> >Is there a circuit I can create to drive two of these 1304's? Preferably one where heat is not such an issue.
>> >
>> >Thanks
>> >
>> >Al
>> Is that a 9 volt part? How much current do you want?
>>
>> A switcher would be efficient, but a bit more trouble than a linear
>> reg. The LEDs will need to be kept cool no matter what regulator is
>> used.
>>
>> If you're only dropping a few volts, and need to heat sink the LEDs
>> anyhow, a linear reg should be OK.
>>
>> You could just buy some 12 volt LED bulbs with built-in regs, but no
>> design fun. The car side/turn light replacement bulbs are cool.
>>
>>
>>
>>
>>
>> --
>>
>> John Larkin Highland Technology, Inc
>>
>> The best designs are necessarily accidental.
>
>The LEDs are : https://cree-led.com/media/documents/ds-CXA1304.pdf
>
>Yes there would be no fun if I just went and brought a part ready to go, although it is getting a little frustrating now.
>
>Are you saying that for a given current, the LED's will reach a certain temperature T which will be the same, regardless of whether driven by linear or PWM?
Unfiltered PWM will be less efficient than linear or a filtered
>On Thursday, July 8, 2021 at 1:56:17 AM UTC+12, jla...@highlandsniptechnology.com wrote:
>> On Wed, 7 Jul 2021 02:08:11 -0700 (PDT), Al Grant <biga...@gmail.com>
>> wrote:
>> >I had a go at creating a DC-DC driver for a pair of Cree CXE1304's with a high/low switch.
>> >
>> >This is for a lighting project on a boat, so supply is 12.5-14.8 DC.
>> >
>> >I tried one with a LM2940-T 10v but I could not get the heat under control on either the LED's or to some extent the LM2940 - I tried to add resistor on the input to the 2940 which reduced overall current to 0.8A but it just made the resistor the new heat problem.
>> >
>> >Is there a circuit I can create to drive two of these 1304's? Preferably one where heat is not such an issue.
>> >
>> >Thanks
>> >
>> >Al
>> Is that a 9 volt part? How much current do you want?
>>
>> A switcher would be efficient, but a bit more trouble than a linear
>> reg. The LEDs will need to be kept cool no matter what regulator is
>> used.
>>
>> If you're only dropping a few volts, and need to heat sink the LEDs
>> anyhow, a linear reg should be OK.
>>
>> You could just buy some 12 volt LED bulbs with built-in regs, but no
>> design fun. The car side/turn light replacement bulbs are cool.
>>
>>
>>
>>
>>
>> --
>>
>> John Larkin Highland Technology, Inc
>>
>> The best designs are necessarily accidental.
>
>The LEDs are : https://cree-led.com/media/documents/ds-CXA1304.pdf
>
>Yes there would be no fun if I just went and brought a part ready to go, although it is getting a little frustrating now.
>
>Are you saying that for a given current, the LED's will reach a certain temperature T which will be the same, regardless of whether driven by linear or PWM?
switcher. Most of the power applied to an LED gets converted to heat.
>
>So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
glued to something bigger. Each LED is 9 volts * 0.8 amps = 7.2 watts.
2 of them is 14 watts. Say 10 watts of heat, since some of the power
comes out as light.
30 K/W * 10W is a temp rise of 300C.
The numbers are rough, but suggest the problem.
Worse, LEDs get less efficient when they are hot. Spritzing an LED
with freeze spray can be dramatic.
>
>I have a heat sink on the LM2940 which seems to be coping at about 50 deg C.
Al Grant
Jul 7, 2021, 10:39:51 AM7/7/21
to
> >
> >So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
> Very wild guess for that little piece of aluminum, 30 K/W unless it's
> glued to something bigger. Each LED is 9 volts * 0.8 amps = 7.2 watts.
> 2 of them is 14 watts. Say 10 watts of heat, since some of the power
> comes out as light.
>
> 30 K/W * 10W is a temp rise of 300C.
>
> The numbers are rough, but suggest the problem.
>
> Worse, LEDs get less efficient when they are hot. Spritzing an LED
> with freeze spray can be dramatic.
> >
> >I have a heat sink on the LM2940 which seems to be coping at about 50 deg C.
> Higher supply voltage will make the reg dissipate a lot more power.
Well I can go bigger and bigger on the heatsink but no matter how big I go, eventually it will hit thermal equilibrium, so whats the solution if bigger doesnt fix it?
> >So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
> Very wild guess for that little piece of aluminum, 30 K/W unless it's
> glued to something bigger. Each LED is 9 volts * 0.8 amps = 7.2 watts.
> 2 of them is 14 watts. Say 10 watts of heat, since some of the power
> comes out as light.
>
> 30 K/W * 10W is a temp rise of 300C.
>
> The numbers are rough, but suggest the problem.
>
> Worse, LEDs get less efficient when they are hot. Spritzing an LED
> with freeze spray can be dramatic.
> >
> >I have a heat sink on the LM2940 which seems to be coping at about 50 deg C.
> Higher supply voltage will make the reg dissipate a lot more power.
On the reg side - how about something like :
https://www.aliexpress.com/item/32615558869.html?albpd=en32615558869&acnt=494-037-6276&aff_platform=aaf&albpg=296904914040&netw=u&albcp=9444695485&sk=UneMJZVf&trgt=296904914040&terminal_id=7828cb7241b441b4bda16b3de5984be3&tmLog=new_Detail&needSmbHouyi=false&albbt=Google_7_shopping&src=google&crea=en32615558869&aff_fcid=e16db09691cd446e9dae3f9bf80937b7-1625668585027-08314-UneMJZVf&gclid=CjwKCAjwoZWHBhBgEiwAiMN66f2DTxdzY9o5BSuGD8RXMlPZ1miCcKv9iMQLVTc-gPVeS6XCDeuWaBoCnSIQAvD_BwE&albag=99457316601&aff_fsk=UneMJZVf&albch=shopping&albagn=888888&isSmbAutoCall=false&aff_trace_key=e16db09691cd446e9dae3f9bf80937b7-1625668585027-08314-UneMJZVf&device=c&gclsrc=aw.ds
jla...@highlandsniptechnology.com
Jul 7, 2021, 11:18:15 AM7/7/21
to
On Wed, 7 Jul 2021 07:39:47 -0700 (PDT), Al Grant <biga...@gmail.com>
wrote:
>> >
Cree wouldn't sell them if they didn't work. They need a big enough
heat sink.
I guess that's a switcher. The MeanWell stuff is good, if that's not a
fake.
wrote:
>> >
>> >So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
>> Very wild guess for that little piece of aluminum, 30 K/W unless it's
>> glued to something bigger. Each LED is 9 volts * 0.8 amps = 7.2 watts.
>> 2 of them is 14 watts. Say 10 watts of heat, since some of the power
>> comes out as light.
>>
>> 30 K/W * 10W is a temp rise of 300C.
>>
>> The numbers are rough, but suggest the problem.
>>
>> Worse, LEDs get less efficient when they are hot. Spritzing an LED
>> with freeze spray can be dramatic.
>> >
>> >I have a heat sink on the LM2940 which seems to be coping at about 50 deg C.
>> Higher supply voltage will make the reg dissipate a lot more power.
>
>Well I can go bigger and bigger on the heatsink but no matter how big I go, eventually it will hit thermal equilibrium, so whats the solution if bigger doesnt fix it?
If "thermal equibrium" is 10C above ambient, bigger works.
>> Very wild guess for that little piece of aluminum, 30 K/W unless it's
>> glued to something bigger. Each LED is 9 volts * 0.8 amps = 7.2 watts.
>> 2 of them is 14 watts. Say 10 watts of heat, since some of the power
>> comes out as light.
>>
>> 30 K/W * 10W is a temp rise of 300C.
>>
>> The numbers are rough, but suggest the problem.
>>
>> Worse, LEDs get less efficient when they are hot. Spritzing an LED
>> with freeze spray can be dramatic.
>> >
>> >I have a heat sink on the LM2940 which seems to be coping at about 50 deg C.
>> Higher supply voltage will make the reg dissipate a lot more power.
>
>Well I can go bigger and bigger on the heatsink but no matter how big I go, eventually it will hit thermal equilibrium, so whats the solution if bigger doesnt fix it?
Cree wouldn't sell them if they didn't work. They need a big enough
heat sink.
I guess that's a switcher. The MeanWell stuff is good, if that's not a
fake.
Clive Arthur
Jul 7, 2021, 11:54:33 AM7/7/21
to
On 07/07/2021 15:17, Al Grant wrote:
> So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
There's not much in the way of surface area on that compared to an LED
> So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
heatsink. The LEDs you have look very much like ones used in domestic
luminaires, maybe 10W or 5W and they have lots of fins.
https://internationallamps.co.uk/product/integral-ildlfr70b001-6w-led-downlight-fire-rated-ip65-dimmable-3000k/
They do get warm, you can't avoid that. Your PSU could be made more
efficient using a switcher, but with 9V LEDs running from 12.5V the
game's hardly worth the candle. Or candlepower.
I don't know what your application is, but multiple smaller LEDs spread
out over a larger area might be an option. Or an LED panel.
--
Cheers
Clive
Jasen Betts
Jul 7, 2021, 5:30:55 PM7/7/21
to
On 2021-07-07, Al Grant <biga...@gmail.com> wrote:
> On Thursday, July 8, 2021 at 1:56:17 AM UTC+12, jla...@highlandsniptechnology.com wrote:
>> On Wed, 7 Jul 2021 02:08:11 -0700 (PDT), Al Grant <biga...@gmail.com>
>> wrote:
>> >I had a go at creating a DC-DC driver for a pair of Cree CXE1304's with a high/low switch.
>> >
>> >This is for a lighting project on a boat, so supply is 12.5-14.8 DC.
>> >
>> >I tried one with a LM2940-T 10v but I could not get the heat under control on either the LED's or to some extent the LM2940 - I tried to add resistor on the input to the 2940 which reduced overall current to 0.8A but it just made the resistor the new heat problem.
>> >
>> >Is there a circuit I can create to drive two of these 1304's? Preferably one where heat is not such an issue.
>> >
>> >Thanks
>> >
>> >Al
>> Is that a 9 volt part? How much current do you want?
>>
>> A switcher would be efficient, but a bit more trouble than a linear
>> reg. The LEDs will need to be kept cool no matter what regulator is
>> used.
>>
>> If you're only dropping a few volts, and need to heat sink the LEDs
>> anyhow, a linear reg should be OK.
>>
>> You could just buy some 12 volt LED bulbs with built-in regs, but no
>> design fun. The car side/turn light replacement bulbs are cool.
>>
>>
>>
>>
>>
>> --
>>
>> John Larkin Highland Technology, Inc
>>
>> The best designs are necessarily accidental.
>
> The LEDs are : https://cree-led.com/media/documents/ds-CXA1304.pdf
>
> Yes there would be no fun if I just went and brought a part ready to go, although it is getting a little frustrating now.
>
> Are you saying that for a given current, the LED's will reach a certain temperature T which will be the same, regardless of whether driven by linear or PWM?
P=VI so that's approximately true, because V is only approximately the same.
> On Thursday, July 8, 2021 at 1:56:17 AM UTC+12, jla...@highlandsniptechnology.com wrote:
>> On Wed, 7 Jul 2021 02:08:11 -0700 (PDT), Al Grant <biga...@gmail.com>
>> wrote:
>> >I had a go at creating a DC-DC driver for a pair of Cree CXE1304's with a high/low switch.
>> >
>> >This is for a lighting project on a boat, so supply is 12.5-14.8 DC.
>> >
>> >I tried one with a LM2940-T 10v but I could not get the heat under control on either the LED's or to some extent the LM2940 - I tried to add resistor on the input to the 2940 which reduced overall current to 0.8A but it just made the resistor the new heat problem.
>> >
>> >Is there a circuit I can create to drive two of these 1304's? Preferably one where heat is not such an issue.
>> >
>> >Thanks
>> >
>> >Al
>> Is that a 9 volt part? How much current do you want?
>>
>> A switcher would be efficient, but a bit more trouble than a linear
>> reg. The LEDs will need to be kept cool no matter what regulator is
>> used.
>>
>> If you're only dropping a few volts, and need to heat sink the LEDs
>> anyhow, a linear reg should be OK.
>>
>> You could just buy some 12 volt LED bulbs with built-in regs, but no
>> design fun. The car side/turn light replacement bulbs are cool.
>>
>>
>>
>>
>>
>> --
>>
>> John Larkin Highland Technology, Inc
>>
>> The best designs are necessarily accidental.
>
> The LEDs are : https://cree-led.com/media/documents/ds-CXA1304.pdf
>
> Yes there would be no fun if I just went and brought a part ready to go, although it is getting a little frustrating now.
>
> Are you saying that for a given current, the LED's will reach a certain temperature T which will be the same, regardless of whether driven by linear or PWM?
The extra "efficiency" of PWM translates to more heat in the LEDs and
less heat in the regulator.
--
Jasen.
LM
Jul 11, 2021, 5:51:50 PM7/11/21
to
On Wed, 7 Jul 2021 02:15:24 -0700 (PDT), Al Grant <biga...@gmail.com>
wrote:
My browser is asking my password when I try to to look at that file.
wrote:
My browser is asking my password when I try to to look at that file.
LM
Jul 11, 2021, 5:53:23 PM7/11/21
to
On Wed, 7 Jul 2021 07:17:43 -0700 (PDT), Al Grant <biga...@gmail.com>
Some leds are available in a Star package. Which is easier to cool.
Martin Brown
Jul 12, 2021, 6:35:42 AM7/12/21
to
On 07/07/2021 15:17, Al Grant wrote:
> On Thursday, July 8, 2021 at 1:56:17 AM UTC+12, jla...@highlandsniptechnology.com wrote:
>> On Wed, 7 Jul 2021 02:08:11 -0700 (PDT), Al Grant <biga...@gmail.com>
>> wrote:
>>> I had a go at creating a DC-DC driver for a pair of Cree CXE1304's with a high/low switch.
>>>
>>> This is for a lighting project on a boat, so supply is 12.5-14.8 DC.
>>>
>>> I tried one with a LM2940-T 10v but I could not get the heat under control on either the LED's or to some extent the LM2940 - I tried to add resistor on the input to the 2940 which reduced overall current to 0.8A but it just made the resistor the new heat problem.
>>>
>>> Is there a circuit I can create to drive two of these 1304's? Preferably one where heat is not such an issue.
>>>
>>> Thanks
>>>
>>> Al
>> Is that a 9 volt part? How much current do you want?
>>
>> A switcher would be efficient, but a bit more trouble than a linear
>> reg. The LEDs will need to be kept cool no matter what regulator is
>> used.
>>
>> If you're only dropping a few volts, and need to heat sink the LEDs
>> anyhow, a linear reg should be OK.
>>
>> You could just buy some 12 volt LED bulbs with built-in regs, but no
>> design fun. The car side/turn light replacement bulbs are cool.
>
>> On Wed, 7 Jul 2021 02:08:11 -0700 (PDT), Al Grant <biga...@gmail.com>
>> wrote:
>>> I had a go at creating a DC-DC driver for a pair of Cree CXE1304's with a high/low switch.
>>>
>>> This is for a lighting project on a boat, so supply is 12.5-14.8 DC.
>>>
>>> I tried one with a LM2940-T 10v but I could not get the heat under control on either the LED's or to some extent the LM2940 - I tried to add resistor on the input to the 2940 which reduced overall current to 0.8A but it just made the resistor the new heat problem.
>>>
>>> Is there a circuit I can create to drive two of these 1304's? Preferably one where heat is not such an issue.
>>>
>>> Thanks
>>>
>>> Al
>> Is that a 9 volt part? How much current do you want?
>>
>> A switcher would be efficient, but a bit more trouble than a linear
>> reg. The LEDs will need to be kept cool no matter what regulator is
>> used.
>>
>> If you're only dropping a few volts, and need to heat sink the LEDs
>> anyhow, a linear reg should be OK.
>>
>> You could just buy some 12 volt LED bulbs with built-in regs, but no
>> design fun. The car side/turn light replacement bulbs are cool.
>
> The LEDs are : https://cree-led.com/media/documents/ds-CXA1304.pdf
>
> Yes there would be no fun if I just went and brought a part ready to go, although it is getting a little frustrating now.
>
> Are you saying that for a given current, the LED's will reach a certain temperature T which will be the same, regardless of whether driven by linear or PWM?
Pretty much. LEDs are roughly constant voltage devices so that power
>
> Yes there would be no fun if I just went and brought a part ready to go, although it is getting a little frustrating now.
>
> Are you saying that for a given current, the LED's will reach a certain temperature T which will be the same, regardless of whether driven by linear or PWM?
dissipation is directly proportional to average current. You need
heatsinking or they will get hot. Datasheet should include details.
I generally just bolt them onto a rectangular sheet of 3mm aluminium.
> So far this heat sink has been remarkably ineffective : https://photos.app.goo.gl/p6mts5TiobFy87wS7 when the total circuit current is 0.8A.
>
> I have a heat sink on the LM2940 which seems to be coping at about 50 deg C.
is about 12v nominal). They lend themselves to being bolted onto a
simple metal plate heatsink and then a modest dropper resistor or better
still a constant current source in series with the chain.
The way you have it now a fairly high proportion of waste heat is in the
linear regulator. A better match of LED device would cut that a lot.
For a one off and starting from where you are now this switching part
might be the path of least resistance if you can find it.
OnSemis NCP3065
https://www.onsemi.com/pdf/datasheet/ncp3065.pdf
(other brands are available)
--
Regards,
Martin Brown
0 new messages