[troutie-mtb]

Just seen this build

7 XPe/g s in a tiny can

it is a tidy build but I can see it being useable anywhere but arctic zones

 

[znomit]

Theres no rule saying you need to run them at 1A.

 

[jmitchell13]

I would say the guy building that light isn't worrying enough about heat.

 

[troutie-mtb]

Theres no rule saying you need to run them at 1A.

OH YES THERE IS :devil:

 

[yetibetty]

I have looked at that light numerous times and every time I do I wonder if the light still works. Maybe it is only run on full for very short bursts, but I bet those LEDs are not happy little LEDs

I am quite obsessed with the getting rid of heat thing though.

 

[lidarman]

That is an LN2 Dewer mount...didn't you see the fill port on the side?

What is the problem?

 

[ifor]

There is the following on the thermals on the web site.

"Thermal protection is set to 60 degrees Celsius. On slow speed trails at 0 degrees it can be run full a couple of minutes, after that thermal protection seems to work fine."

So it is close to the limit themaly but if you let the maxflex do it's job then everything is OK. Flat and downhill I expect it to be cool enough climbing it may drop back but even at L3 and hot your looking at 700+ lumens from the leds pleanty for climbing... If you were in a hot climate and night riding at say 25+ celcius then things may be not so good, but global warming will have to go a long way before that worries me in the UK....

Most of the lights here will not have the thermal stuff in the flex driver come in unless you realy do stop for a few minutes we are probobly being too conservative.....

 

[odtexas]

Sometimes I think we go a bit overboard on heat sinking and finning our lights.:madman:
Then there is the other extreme like we see here.
Regardless he seems like an innovative builder.
Hopefully he will consider joining our little dysfunctional community

 

[OverTheHill]

In my limited experience with building lights from aluminium section with no real account of thermal management either from the casing or the driver I would have to “Yes “ to troutie’s question. The main caveat to that being that I have not yet attempted to run multiple LEDs at anything approaching 1 amp.

We are all indebted here to the likes of troutie, deesta, yetibetty et al for sharing their ground-breaking designs and knowledge that push the envelope of what is possible when building our own lights. I liken it to Formula 1 where the technology of ultimate automotive performance filters down and finds its way on to family cars. Having said that not everyone needs a £100K Ferrari and the majority of us are content to beetle away in sheds and garages to satisfy our own ends.

In normal use while riding my lights hardly get warm so with a degree of vigilance on my part if stationary for any length of time, I have managed so far to avoid a burn-out. Added to that the cost of the whole of one of my lights is less than a Flex driver so it wouldn’t be the end of the world if it did go up in smoke.

At the end of the day though, especially where round-bottomed fins are concerned, who can deny their aesthetic appeal?

 

[tamen00]

I do not think that light would work very long for me, even set at 350ma... the ambient temp for several months of the year where I ride is 35-40c, and I use the lights 5 days a week during that time (it is so hotyou can only ride at night!). A lot of the lights that I build are very similar to the Ktronics light or other round housings with 3xpe/xpg with a bunch of cooling fins. Running 700ma, the thermal protection kicks in even while rolling 20mph on the road bike (set at 60c). If the ambient temp is 25C, all of my lights will run stationary on the bench without tripping the thermal protection - what a huge difference outside temp makes!

I have a couple of summer lights that run at 500ma... and a couple of winter lights that I have set at 800 and all seems good! I think that the environment we use the lights in dictates the amount of cooling that is needed... in the UK, you can get away with a lot more than in the AZ desert in the summer!

(FYI... a magicshine only lasted about 30 minutes on my road bike this summer before completly cooking itself )

 

[HEY HEY ITS HENDO]

nice find troutie! .. :thumbsup:

in the UK, you can get away with a lot more than in the AZ desert in the summer

ha, dont we know it ...... -7 most of last week, all the trails are snow covered
............... and i`ve got 10 days holiday :madman:

 

[AlisterG]

Perhaps a RC motor heatsink or two could be wrapped around the can for a bit more cooling?

 

[tamen00]

nice find troutie! .. :thumbsup:

ha, dont we know it ...... -7 most of last week, all the trails are snow covered
............... and i`ve got 10 days holiday :madman:

SNOW?!??!!! What is that? We have it rough here this time of the year... 75 degrees F during the day, sunny, no wind... but it does get down to 50 or so at night so we have to break out the knee and arm warmers for the chilly nights

 

[vroom9]

Even if you don't get them hot enough to fry the LEDs the lifetime and output can be compromised. Remember the super long life time of LEDs is speced at 350ma and 25C. I think even the XP-G's are that way. Now who runs them like that? Not too many in this place that's for sure. Run them at max current and high temps and the lifetime is cut by a large factor. Hard to say with any kind of certainty, but depending on exactly how hot it could be a low as a few hundred hours. Now for me that is fine. By the time I log a few hundred hours on a light I will likely be building the next version. It might even be long enough for two or three generations of lights.

Don't forget that cooler temperatures yields more light! Now that's a reason for good cooling if I ever heard one.



Finally, while not politically correct... Merry Christmas everyone.

 

[yetibetty]

My short answer to trouties question is probably, but there is no harm in it and it’s a good habit to get into.

Long answer: I remember the old days when electronics didn’t work properly until they had “warmed up” and people would switch the TV on half an hour before the programme started so that by the time they sat down to watch it the thing had a decent picture as it was now nice and warm.

Those days are gone and most modern electronics are full of heatsinks and/or fans and the last thing they want to do is warm up. The laptop that I am typing on has its internal fan whirring away at the moment.

LEDs are just like most other modern electronics that work best and last longer when kept cool. They also produce more light when cool.

This is easy to achieve on a bike light when on the move due to the air flow and no problem what so ever if the rider is a racer type who rarely stops or never rides slowly.

If on the other hand you ride slowly, stop a lot perhaps to take in the night wildlife or have to ride slow due to the two feet deep mud out there, then yes cooling fins do help.

The nice thing about cooling fins is that they are free and a great way to save a bit of weight and give something else to do to the light whilst waiting for the LEDs or driver to make that seemingly never ending journey from the other side of the world.

 

[Goldigger]

Just seen this build

7 XPe/g s in a tiny can

it is a tidy build but I can see it being useable anywhere but arctic zones

Looks like the lumicycle halogen housing to me..

 

[OldMTBfreak]

I live in Sunny NW Florida. I feel that temp management is a necessary part of a light. I have been using Taskled drivers BECAUSE of the thermal management. I have had my triple MCE throttle back to level 3 many times during slow climbs. Yes, I'm slow most of the rest of the time also. Without more finning, active cooling (aka. cosmo) or LN2, cutting back the power in a overheat condition is the only way to save the LED's. As Vroom posted, excessive heat WILL shorten the led's life, and it will shorten the controllers life. Most importantly though, above 25C, the amount of light emitted falls off. According to the MCE spec sheet, output flux falls to less than 80% at 100C junction temp. Also, the junctions thermal resistance is 3 deg C/watt (from led's heatsink to star). Then we got 2 deg C/watt from the star to the lamp housing. This is a long torturous path that the heat has to flow from the led to the heatsink. To say the above in a different way, we need to move the heat from the led to the heatsink. We need to hold the junction temp as near to 25C as we can. Every watt put into the led raises the junction temp by 5 deg C. The bottom line is this; for heatsinks, the more the better. Now (for all you arty-farty types) the job is to make all that heatsink pretty. Think "round bottom Finns". James

 

[Kaitsu]

No, we don't

My bike was couple of minutes outside, temperature meter could only show -22deg, couple of minutes later it went blank.

And yes, we really have a heat transfer problem. The Lumis with 7xXPG only keeps at [email protected], even at -22. On the other hand my toes need some extra warming. I'm missing some heat transfer path here... How to keep toes and fingers warm without thermic-ic and extra batteries?

 

[Bobblehat]

Can anyone enlighten a real noob? I have a heat type question.

I know it's not the same as running normal current through the led(s) for hours, but what effect does heating xpg's up on a hot plate to a temperature hot enough to remove them from a board to swap them out? Are the removed leds still useable and has their lifespan or output been altered significantly? The tail end of Troutie's video in the "It's not square ... " thread seems to apply a good amount of heat for a good while?

Just trying to put this in perspective with running them "normally", but maybe on the hot side of "normal".

 

[OldMTBfreak]

The spec sheets give you the max temp you're supposed to heat the led up to.