Q About Heatsink Isolation

Im a little confused about heatsink isolation on luxeon stars.

I have heard before that it is possible to share one Heatsink with multiple stars or with other components.

I have been having some trouble on my ATV project because not all my heatsinks are isolated. For example...

1)
I have 2 Red 1 watt Star-O LED's still connected at the PCB. Is it safe for me to run them still connected? (the heatsinks are also epoxied to these 2 that are still attached and the Heatsinks also touch. Is this safe?

2)
Now what If I connect the LED's to the frame on my ATV (ground) using an aluminum bracket that is connected to the heatsinks. I have done this on accident because I made my brackets from aluminum (only workable material I could find). Is this bad? (HS is +)

3)
Now I also have some 3 watt blue LED's that are heatsunk, and the aluminum bracket connects to the heatsink, then the aluminum bracket connects to the frame (ground). is this bad? ( HS is -)

4)
Can I run just the blues connected to the frame (ground) or just the Reds successfully if not together?

I feel like im loosing my mind and sleep over this. I need to know which of these thigns are safe to do and which are not.

Thanks for your help and time,
JI

If the Luxeons are on the original star heat sink, they are isolated. The heatsinks may be connected to ground, each other, it does not matter.

Blue, Royal Blue, Green, Cyan and white bare emitters may be used on a grounded heatsink without problems. Red, Amber and Red/Orange emitters need total isolation.

But he is testing his red stars and such and postive is going through them to the heatsink. So I guess his aren't being isolated by the star board...

He is also using artic silver which is semi-conductive..

EDIT: See my response in your other thread about the spacer stand offs. I also have another idea that would be easy...

Thanks man,

Im not using any emitters but I am using 3 watt blues, so it is safe for the blues to ground the heatsinks to the frame (ground)?

And it it safe for the Red (stars) to share the same heatsink as long as the heatsinks are not grounded? or even the the red Star_Os need compleat seperation (ie. not the same heatsinks and aluminum board must be seperated)

I found this info on the Star3 PDF

"3. Electrical insulation between
neighboring Stars is required –
aluminum board is not electrically
neutral."

So the heatsinks connected the the aluminum backboard on the PCB board can be connected to the frame?

The red stars can be connected to the same heatsink, or to the same grounded heatsink as long as they are not bare emitters?

I notced that when I checked the frame for voltage I was finding some when the Red's Heatsinks were attached to the frame...But the frame(gound) is used to power other things on the ATV also so I was concerned.

[ QUOTE ]
IsaacHayes said:
But he is testing his red stars and such and postive is going through them to the heatsink. So I guess his aren't being isolated by the star board...

He is also using artic silver which is semi-conductive..

EDIT: See my response in your other thread about the spacer stand offs. I also have another idea that would be easy...

[/ QUOTE ]

yea im working with the spacers idea, looking for someplace to by some nylon spacers...Only troule is Im not sure this would protect the inside of the hole in teh bracket from teh bolt being used to mount it.

I also noticed voltage from the Cyan LEDs (star-O's)seem to flow into the heatsink, and it registers as Positive also... Whats your other idea?

"3. Electrical insulation between
neighboring Stars is required –
aluminum board is not electrically
neutral."

The alluminum board they are refering to is the Star they are mounted on. I don't know why lambda said they are isolated on the board, unless that's just his experience. It looks to me like they would conduct since the thin pcb board type thing looks like copper where the luxeon is mounted to.

Definatly isolate them from the ground at least on the Red ones, if you are testing conductivity through them to the heatsink then it doesn't matter what anything says, you obviously need to isolate them.

The blues should be ok but I don't know if them grounding will mess up your current regulation going to them or not. I'd isolate them just in case if I were you...

Sepearating the 2 Reds from each other would be a good idea, since this will most likely affect the current limiting, as one could have a different voltage foward than the other one and draw more/less current or cause the regulator to dump more power into both of them by being confused...

Not sure how your regulator powers each luxeon or is hooked up to them, you might need a seperate one for the reds and the blues since the blues are connected negativly together on the braket, and same for the reds only they are pos, so that could cause problems. I'd have to visualise it more


Oh yeah, check out my idea about gromets in the other thread. That is the way to go. That way you have less work and it also will insulate around the bolt to where the bracket will be fully insulated!

Basicly isolate blues/reds from the frame and from each other would be the best way to cover all bases without trying to figure out what is going on. That should let the regulator work without having to figure out what's going on. That's what I'd do so you don't have to do any more trial and error checking. It may not be nessasry to isolate everything, but it covers all bases.

The grommets can be found at any hardware store. The will insulate the hole in the braket and around the bolt. It won't touch at all and is only one part per hole to mess with. just drill install, and bolt up. You might need to put a washer on the end of the bolt to press firmly on the gromet so the bolt head doesn't go through the gromment, depending on the size of the grommets and the size of the bolt head.

I think you should be set after this if your current regulator is doing it's job (which it seemed to be doing with the cyans in the begining) /ubbthreads/images/graemlins/thumbsup.gif good luck! /ubbthreads/images/graemlins/grin.gif

/ubbthreads/images/graemlins/frown.gif I just toasted another star-O


I tried compleating the circuit with a resistor but it got hot and started to smoke before I could get a reading, I had 2 Cyan stars heatsunk so I figured I would give them a shot. (one of them) So I hooke dit up to my regulator and it registered .3amps (I think it's on amps I thoguht it would read lower than jsut .3..then my Computer PSU shut off (its an old broken one so i figured wth and I turned it back on.) Then I tested the Cyan star again this time it was saying .8, Suddenly the LED fried /ubbthreads/images/graemlins/frown.gif

Damn Im frying these things left and right..Over the past few months I have toasted 6 of them! errr actually only 5, one of them I broke.

/ubbthreads/images/graemlins/frown.gif No idea whats going on with the circuit so now it looks like I will have to build another one, and Im out 1 more star... GRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRR

The problem with seperating them (the reds) is the the bracket I had made fits them while they are connected, if i seperate them i will have to get another bracket made which will take me up to 3 months. /ubbthreads/images/graemlins/frown.gif which is the end of riding season...The gromits sound like the way to go if I can find some the right size..

Man this really sucks, this stuff is really hard to figure out..I knwo the circuit is wired correctly, it appears it is a bad component /ubbthreads/images/graemlins/frown.gif

Unless your confident about the circuit I'd go with some form of aftermarket one... the NiMH battery back (which you could make easily) seems like a good idea to smooth out spikes too. Might be worth it to save the headaches! I feel your pain! /ubbthreads/images/graemlins/frown.gif

hey thanks for stickin with me on all this. I really appreciate it! I gotta find out about those blues, ether or not they need to be isolated. if not I have nothing to worry about as I know a way to isolate the reds (compleatly. I know they did not fry when I tested it out so I guess I should check it out first right?

The circuit I am confident about, I made one for my friend and it works great...just not efficient, but that does not matter in this application...Unless i go with teh battery packs too. Looks like I jsut got a bad part, the LM317HVT must be toast..I bought it out of the package, so it figures...Guess ill stick with the sealed package radio shack ones /ubbthreads/images/graemlins/smile.gif

With the badboy circuit, how much runtime could I get running 2 3 watt blues and 2 1 watt reds? thats roughly 3 amps so I would need a 1500mah battery pack to last about 20-30 minutes right? That might be a good idea so I could flip it on when I have the bike off (good for saftey on a dark night at the top of a dune)...Vs how long the battery pack with a circuit like mine + a battery pack; mine is not very efficient. Then again mine might be efficient enough...

My circuit costs me about $10 vs the 40+ shipping for the badboys

On the other hand I kinda like it how the lights get brighter when the bike revs up..

What gets me on the battery pack, is how to charge it. Would I not have to make some sorta special charging circuit that shuts off when the pack reaches full? otherwise I would think it would get overloaded quite fast...


WHAT about the flashlight mods where I see 3 stars all mounted to the SAME heatsink.........

[ QUOTE ]
Justintoxicated said:
WHAT about the flashlight mods where I see 3 stars all mounted to the SAME heatsink.........

[/ QUOTE ]

For the whites, no isolation is necessary when wired in parallel. For red family luxeons & whites wired in series, you must isolate the die slugs.

[ QUOTE ]
Slick said:
[ QUOTE ]
Justintoxicated said:
WHAT about the flashlight mods where I see 3 stars all mounted to the SAME heatsink.........

[/ QUOTE ]

For the whites, no isolation is necessary when wired in parallel. For red family luxeons & whites wired in series, you must isolate the die slugs.

[/ QUOTE ]
thanks anyone else know about the other colors, my cyans appeard to be + on the HS acording to my multimeter?

All mine are wired in series...But for my other project where Id rather leave the slugs connected, im wiring the Cyans in parallel they are similar to whites right?

[ QUOTE ]
IsaacHayes said:

I don't know why lambda said they are isolated on the board, unless that's just his experience.

[/ QUOTE ]

Ok, for 1 watt stars, this is true. They are isolated. The 1 watt star uses a copper heat spreader insulated by a layer of the PCB from the heatsink.

3 watt and 5 watt stars may not be insulated as the emitter is mounted on the aluminum heatsink directly.

To simplify testing, use a battery to test the Luxeons and when testing your power regulation circuit, use a 1 ohm resistor to simulate the Luxeon in the circuit. Once eveything is working, then try the two together.

cyans are the same as blues, just diff color.

I for one always isolate the LS from the sink.
Even in the situations where its really ok, I just dont like the idea that it can coneect through the sink and body like that.

Plus, if you always isolate it, you know you are safe.
I usually do this by applying alittle nail polish to the bottom of the LS, letting it dray, then epoxying it in place.

You could also use two layers of epoxy. Letting the first cure well. I find this not to be the best bond though, AA epoxy should not be added to cured AA epoxy.

jtice, what type of epoxy are you using? I would be concerned about heat transfer if it isn't thermally (not electrically) conductive heatsink compound. I have some experience with building electronic circuits, using mica insulators for electrical insulation (they are highly conductive of heat). But I've been told that using them with Luxeon emitters will fry the emitter. The stars have some heatsinking on them, which helps, but eventually they will get hot at higher power levels if the conductivity to a large heatsink isn't high enough. -RussH

JIT, note that the White/blue/green LEDs operate at a different voltage than the Red/orange/yellow-amber LEDs. This means that no matter what you should have a different LM317 feeding them.
In a current limited circuit, the 317 will raise the voltage until the set amount of current flows. But the required voltage (Vf) is a bit different for White vs. Red LEDs. This means that they will get the required current at different voltages, say for example 2.8v for red, and 3.4v for white luxeon 1w. One LM317 can't do both at once....
As far as always using isolation as jtice said, I like it, but it can be difficult to get enough heat transfer. Often you can isolate the entire heatsink electricaly as isaachayes said. Grommets should be easy to find, 1/2 in. & up at any hardware or electrical place, but smaller sizes like 3/8" are more of an electronic item.

yes listen to RussH. right on about the regulator. Also having more than one luxeon of the same type (blue or red), say 2 blues or one blue one cyan, they can vary from luxeon to luxeon on Vf. Allthough not as much as red vs blue. So one might be overdriven a little compared to the other. One coudl be Vf of 3.4 and the other Vf 3.7... you could solve that by measuring the curent going to them and using a resitor to keep both the same brightness...
But first things firts, isolate your mounting brakets, and use seperate current/voltage regulators for the blue/cyan/green and red/red-orange/amber.

Also for the grommets you could check at a car stero shop, as they use these for installing power wire through car firewalls. They might have small ones used for 12 or 10gauge wire that would work..

[ QUOTE ]
IsaacHayes said:
yes listen to RussH. right on about the regulator. Also having more than one luxeon of the same type (blue or red), say 2 blues or one blue one cyan, they can vary from luxeon to luxeon on Vf. Allthough not as much as red vs blue. So one might be overdriven a little compared to the other. One coudl be Vf of 3.4 and the other Vf 3.7... you could solve that by measuring the curent going to them and using a resitor to keep both the same brightness...
But first things firts, isolate your mounting brakets, and use seperate current/voltage regulators for the blue/cyan/green and red/red-orange/amber.

Also for the grommets you could check at a car stero shop, as they use these for installing power wire through car firewalls. They might have small ones used for 12 or 10gauge wire that would work..

[/ QUOTE ]

LOL
Yea I think the reds would be a little overpowered driving them at 1 amp /ubbthreads/images/graemlins/ooo.gif especialy since my cyan fried at 800ma (on a small heatsink)..I did not however realize that I needed a different regulator for 1 watt red vs 1 watt blue, I thought if I kept the current at a steady 350ma, that each LED would produce its proper Vf...The advantage to regulating curent instead of voltage... Is this false then? So far it seems to allow the LEDs to run at different Vf when I measure them (very slightly different Vf of the same color but both run at 350ma...er a little less than 350ma really) I would think I could plug as many 1 watt stars into the same line that is regulated to be 350ma all the way through...Someone correct me if I am wrong, but I have heard of others using LM317T's in this manner before, I have a link if you would like to see it.

1) My main concern is wether or not the Blue's heatsinks need to be isolated from ground.

As for 1 watt stars being isolated, here is my observation.
:
connecting the neg to - and pos to + the heatsink picks up a positive voltage wether cyan or red. However connecting neg to - and pos the the heatsink does not allow the LED's to light up at all.

2) So this would mean the LED is isolated from the HS in the case of both Red and Cyan. Does this mean that I could share heatsinks with any 1 watt star of the same type? since they are isolated from the heatsink? I don't understand how the HS can recieve a voltage but not allow it to flow into the star? perhaps the voltage the HS generates is the voltage that flows across the star?

3) Would this indicate that they should be allowed to share a HS (same color same watt)? Otherwise I may have to make a new bracket, my mounting assumed they were atached together...

Ill check out a sterio shop or 2 when I get the chance. It does not sound like the lights will be ready for this weekends ride. Im going to build a new circuit tonight. The gromots I need would need to fit a bolt smaller than 3/8th...probably a size smaller than that (they are metric bolts...probably 6mm or so)

4) what kinda wire is the best for connecting components together on the back of a circuit board? Last tiem I used stranded because it was all I had and it sucks really bad /ubbthreads/images/graemlins/smile.gif

Another thing, I just read that annodizing aluminum prevents it from conducting electricity (in the clear annodizing post)????

Sounds like a perfect way to isolate the heatsinks...I lapped all mine with sand paper to help the transfer of heat, I did not realize I was making them electricaly conductive..

Is this true? if so I would have a simple solution of purchasing annodized heatsinks inthe future nad using artic alumina epoxy(could not find any locally so I went with artic silver epoxy), but I have my doubts...Can Someone verify this?