Retrofitting my Command Electronics (yay Schoolcraft Michigan) fixtures with some custom made LED lighting.I chose the Philips Lumileds LUXEON Rebel, in a 3500K bright-white color temperature. At one watt each LED puts out about 100 lumens, and each fixture receive either three or six of these.First I mount my LEDs to a carrier board that interfaces them with electrical connections and also helps transfer heat away.The LEDs are bonded by solder paste, which I melt with a paint stripper gun. The brown residue is a byproduct of a rosin flux, which I used to thin out my solder paste some. The paste is a little out of date and needs help.Next the carrier board is attached to a hefty slab of aluminum, which I cut down to fit inside the fixture. I'm going to upgrade to a wider piece of aluminum, cut to fit the entire inside of the fixture. A silicone based metal filled compound is applied between the carrier and the slab, to aid in transferring heat, you can see some of it oozing out on the edges and middle hole.The LED does not connect directly to the 12V system, it connects to a driver board. The driver is a DC to DC converter, that takes in whatever voltage the coach DC system is running at and limits it to a safe value that is exactly what the LED wants (in my case, about 9v, 350mA.)With the fixture's cover back in place, I hooked it up to a little 12v battery and it's time to test things out!I'll add some more pictures when I install this in the coach...

Hey Gordon,From my experience with heat dissipation, I don't think you're going to get much better dissipation by increasing the dimensions of the aluminum plate. Heat from devices like this transfer from conductive to convective in a fairly small locality. You may get better reflectivity, though. If better reflectivity is a goal, polishing the plates may bring good results. This is a larger heatsink polished for better thermal conductivity, it was accomplished with ever decreasing grades of sand paper and polishing compounds:

Nice project!One watt for 100 Lumens is a very good ratio.

Have you done any real testing for RFI NOISE from your controller boards. My experience with LED VOLTAGE REGULATORs is they are prone to produce high Radio Frequency Interference and in my case interferred with my OTA HDTV reception, VHF WX RADIO Interference, and really big time interference to my ham Radio VHF operations...My way out was to use LED BOARDS that did not include VOLTAGE REGULATORs like the $4.95 LED BOARDS from EBAY CHINA...Just asking....Roy Ken

westend wrote:Hey Gordon,From my experience with heat dissipation, I don't think you're going to get much better dissipation by increasing the dimensions of the aluminum plate. Heat from devices like this transfer from conductive to convective in a fairly small locality. You may get better reflectivity, though. Nice polish job on that block!You are correct, increasing the size of the plate alone does not help much with convective cooling. My goal is to increase the thermal mass. If the LEDs are heaters, running on a fixed power, the more mass they have to cool, the less heat will be "trapped" inside them. Entropy of the mass is always working to cool it... at least, that's the theory I'm working off of. As long as I can keep the system under 200F with high ambient temps, I'll be happy. I don't have anything fancy like a FLIR camera, but I am going to get one of those laser pointer temperature probes.If the big slab of aluminum can't keep things cool, I'll ditch the plastic fixtures entirely and go with a recessed heat sink in the ceiling. There's a new piezo-electric LED cooler I've been wanting to try. :)

RoyB wrote:Have you done any real testing for RFI NOISE from your controller boards. My experience with LED VOLTAGE REGULATORs is they are prone to produce high Radio Frequency Interference and in my case interferred with my OTA HDTV reception, VHF WX RADIO Interference, and really big time interference to my ham Radio VHF operations...My way out was to use LED BOARDS that did not include VOLTAGE REGULATORs like the $4.95 LED BOARDS from EBAY CHINA...Just asking....Roy KenI don't have much experience in that area. I did follow the manufacturer's reference design as best I could which they claim minimizes RFI/EMI, including good quality high frequency bypass capacitors on the input and output. The high frequency "switch node" on my circuit board is wrapped by a ground plane, and I'm using a RF shielded inductor.As sloppy as some of the LED circuit boards I've seen are, I believe your claim of interference. I understand they want to keep costs down, but doesn't the engineer that designed the board have any pride in his craft, I mean, come on!

LED Retrofit, not off the shelf! (lots of pics)

16 Replies

SCVJeff
Explorer
Feb 22, 2014
The one thing you can do to help RFI coming off the boards is to twist the +&- together both in and out of the stitcher board. Start the twist as close as you can all the way to the LED board. Likewise to the power switch on the input side. When I was working on reducing RFI on some LED's I have, this has the single biggest effect on radiation coupling out the antenna.

And since you're obviously a decent solder person, some ultra short lead .01uf Ceramic caps across the in and out of the boards helps as well. But that might already be what C2 & C3 are for on the board.

Question: Why didn't you just buy 3000k 3W boards rather than building them?
GordonThree
Explorer
Feb 22, 2014
RoyB wrote:
Have you done any real testing for RFI NOISE from your controller boards. My experience with LED VOLTAGE REGULATORs is they are prone to produce high Radio Frequency Interference and in my case interferred with my OTA HDTV reception, VHF WX RADIO Interference, and really big time interference to my ham Radio VHF operations...

My way out was to use LED BOARDS that did not include VOLTAGE REGULATORs like the $4.95 LED BOARDS from EBAY CHINA...

Just asking....

Roy Ken

I don't have much experience in that area. I did follow the manufacturer's reference design as best I could which they claim minimizes RFI/EMI, including good quality high frequency bypass capacitors on the input and output. The high frequency "switch node" on my circuit board is wrapped by a ground plane, and I'm using a RF shielded inductor.

As sloppy as some of the LED circuit boards I've seen are, I believe your claim of interference. I understand they want to keep costs down, but doesn't the engineer that designed the board have any pride in his craft, I mean, come on!
GordonThree
Explorer
Feb 22, 2014
westend wrote:
Hey Gordon,
From my experience with heat dissipation, I don't think you're going to get much better dissipation by increasing the dimensions of the aluminum plate. Heat from devices like this transfer from conductive to convective in a fairly small locality. You may get better reflectivity, though.

Nice polish job on that block!

You are correct, increasing the size of the plate alone does not help much with convective cooling. My goal is to increase the thermal mass. If the LEDs are heaters, running on a fixed power, the more mass they have to cool, the less heat will be "trapped" inside them. Entropy of the mass is always working to cool it... at least, that's the theory I'm working off of.

As long as I can keep the system under 200F with high ambient temps, I'll be happy. I don't have anything fancy like a FLIR camera, but I am going to get one of those laser pointer temperature probes.

If the big slab of aluminum can't keep things cool, I'll ditch the plastic fixtures entirely and go with a recessed heat sink in the ceiling. There's a new piezo-electric LED cooler I've been wanting to try. :)
RoyB
Explorer III
Feb 22, 2014
Have you done any real testing for RFI NOISE from your controller boards. My experience with LED VOLTAGE REGULATORs is they are prone to produce high Radio Frequency Interference and in my case interferred with my OTA HDTV reception, VHF WX RADIO Interference, and really big time interference to my ham Radio VHF operations...

My way out was to use LED BOARDS that did not include VOLTAGE REGULATORs like the $4.95 LED BOARDS from EBAY CHINA...

Just asking....

Roy Ken
ScottG
Nomad
Feb 22, 2014
Nice project!
One watt for 100 Lumens is a very good ratio.
westend
Explorer
Feb 22, 2014
Hey Gordon,
From my experience with heat dissipation, I don't think you're going to get much better dissipation by increasing the dimensions of the aluminum plate. Heat from devices like this transfer from conductive to convective in a fairly small locality. You may get better reflectivity, though.

If better reflectivity is a goal, polishing the plates may bring good results.

This is a larger heatsink polished for better thermal conductivity, it was accomplished with ever decreasing grades of sand paper and polishing compounds:

Forum Discussion

LED Retrofit, not off the shelf! (lots of pics)

16 Replies

About DIY Maintenance

Related Content

Generator shelf fabrication

Dealers lots are filling up

WFCO Retrofit?

Cassette toilet Retrofit

Reroute Lot Sewer Pipes