Smoking P4 in Mag Drop-in (literally!)

SnowplowTortoise

Newly Enlightened
Joined
Feb 9, 2007
Messages
106
Hi folks. This is my first post so I'll try to keep it simple to minimize my chances of writing something stupid. I've had the mo' lumens virus since childhood, but I have zero experience in electronics tinkering and this was my very first attempt at soldering. Thus, there is strong chance that I've done something wrong that would be obvious to most of you.

Here's the situation: I have a Mag 4D with alkaline cells. I've been using the Mag 3W drop-in for some time now with no issues. Unable to leave well-enough alone, I pulled off the Luxeon on the drop-in and replaced it with Seoul P4 U-bin from the Sandwich Shoppe. I used Arctic Alumina to adhere the LED to the base. Because the prongs on the LED were not long enough, I used some strips of metal to extend down into the base (cut up a penny – did you know they're not solid copper anymore?). Then I globbed a bunch of rosin-core solder all over the place, installed it back into my 4D, and fired it up. Amazingly, it lights up.

But here's the issue: after a minute or so, the light begins to dim and turn very blue. Smoke then ensues. At first I thought it was the reflector melting (which it was), but the thing continued to smoke after the reflector was removed. Why is this thing getting so much hotter than the Luxeon? Did I do something wrong in the execution or is the whole concept flawed?

Thanks very much for your help. This site will be my undoing!
 

Gnufsh

Enlightened
Joined
Sep 12, 2006
Messages
242
Location
Shasta Lake, California
SnowplowTortoise said:
(cut up a penny – did you know they're not solid copper anymore?).
Pre-1982 pennies are ~95% Cu, the more recent ones have a zinc core. I sort out and save some of the pre-1982 ones to use for heatsinking, etc.
 

Sable

Enlightened
Joined
Sep 27, 2006
Messages
454
Location
Arizona
I can't give you a really excellent answer, but one thing that has been noticed by more than a few people is that the MagLED modules have really, seriously lousy heatsinking ability. I don't know that this would be a huge deal considering the Luxeon part works on it - perhaps there's an airgap or something unintentional between your Seoul's slug and the LED post?
 

tebore

Flashlight Enthusiast
Joined
May 10, 2006
Messages
2,141
Location
Toronto, Ontario. CAN.
Sounds like you had a short. The P4 runs much cooler than the Luxs do because it's more efficient. Remember the SSC die is the anode if the AA epoxy wasn't completely set and ran thin just enough it will short.
 

chimo

Flashlight Enthusiast
Joined
Sep 16, 2004
Messages
1,905
Location
Ottawa, Canada
Since the slug of the P4 is connected to the anode, have you checked for a ground fault between the slug and the heatsink?
 

Gunner12

Flashaholic
Joined
Dec 18, 2006
Messages
10,063
Location
Bay Area, CA
I remember Newbie posting something that said Seoul P4 LEDs turn blue when they are either overdriven or when they get too hot. So you might want to check for heat.
I might also be a short.

Good luck with your mod
 

c0t0d0s0

Enlightened
Joined
Jun 19, 2006
Messages
265
If there was a short, it wouldn't light up at all. Something is very wrong with heat management here... dimming and spectrum shift towards blue is the first sign of the slug being overheated with these SSC P4s. How thick was the AA layer? Also, did you measure the current flowing through the LED?
 

Illum

Flashaholic
Joined
Apr 29, 2006
Messages
13,053
Location
Central Florida, USA
SnowplowTortoise said:
Because the prongs on the LED were not long enough, I used some strips of metal to extend down into the base


something tells me the contact extension you used to lengthen the Emitter's leads is shorting to the heatsink somewhere below the adhesive...

[/font said:
SnowplowTortoise] after a minute or so, the light begins to dim and turn very blue. Smoke then ensues....


tis may suggest that the magleds driver output more voltage or current that the P4s operational specs and you've essentially cooked it
 

Long John

Flashlight Enthusiast
Joined
Mar 16, 2006
Messages
2,307
Location
Spain, near Cadiz
Same thoughts like Ilum_the_nation, I think the vf of the SSC-P4 is lower than the vf of the Luxeon.

Best regards

____
Tom
 

AndyTiedye

Flashlight Enthusiast
Joined
Oct 28, 2006
Messages
2,033
Location
Santa Cruz Mountains
c0t0d0s0 said:
Something is very wrong with heat management here... dimming and spectrum shift towards blue is the first sign of the slug being overheated with these SSC P4s.

The smoke is a pretty good hint in that direction as well.
 

CM

Flashlight Enthusiast
Joined
Sep 11, 2002
Messages
3,454
Location
Mesa, AZ
...Then I globbed a bunch of rosin-core solder all over the place, installed it back into my 4D, and fired it up. Amazingly, it lights up.

I bet you somehow bypassed the regulation circuit in your zeal to mod the light and are now direct driving the LED. Since the Mag LED module has virtually no heatsinking, the die is heating up to the point where you are experiencing severe drop in output. The shift toward blue in tint is another symptom of an overdrive condition, enough to possibly cause smoke. Do you have an ammeter that you can use to measure current draw from the 4D cells? I bet (again) that you are pumping a few amperes of current through the poor LED.
 

SnowplowTortoise

Newly Enlightened
Joined
Feb 9, 2007
Messages
106
Thanks for all of your help folks. Nice to know about the pre-1982 pennies! Upon closer inspection, there was indeed a connection being made between the housing of the drop-in and one of my lousy soldering jobs. I filed away the offending segment and it seems to be running cooler now. I had no idea that a short circuit could cause that kind of problem while still allowing the device to produce so much light.

More soldering practice for me! Thanks again,
Craig

Edit per CM's post: I think you nailed it. I was expecting it to be brighter once I elimated the short, but I think it might actually have gotten less bright. In my limited understanding of electronics, I believe that would support your idea, no? Well, so now everyone knows what a P4 does in direct drive from 4 D-cells!
 
Last edited:

Illum

Flashaholic
Joined
Apr 29, 2006
Messages
13,053
Location
Central Florida, USA
doing this type of soldering you generally want to have a fine tipped iron and thin rosin-core solder [I use around .022-62/36/2, though they make even smaller]

I use .062" for normal occaisons, which makes a mess at times:ohgeez: but I use what I have
 

CM

Flashlight Enthusiast
Joined
Sep 11, 2002
Messages
3,454
Location
Mesa, AZ
SnowplowTortoise said:
Edit per CM's post: I think you nailed it. I was expecting it to be brighter once I elimated the short, but I think it might actually have gotten less bright. In my limited understanding of electronics, I believe that would support your idea, no? Well, so now everyone knows what a P4 does in direct drive from 4 D-cells!

LED's can survive some pretty harsh treatment. Have you run it for very long under those conditions? If you kept it down to say, a few seconds, I would guess that it should have survived with minimal damage. If you ran it for much longer, then most likely irreversible damage has occured. Remember also that the LED module, when it is working correctly, will diminish drive current to the LED as it heats up. The effect can be quite pronounced so take your reading (or make your judgement) when it is cool.

Modding is fun, isn't it? Hold on to your wallet around here :D
 

nanotech17

Flashlight Enthusiast
Joined
Jul 31, 2006
Messages
2,150
Location
N 03°13.884° / E 101°29.329°
You should be using only 2D cells = 3v .
And not 6D cells = 6v !
So far the SSC P4 highest Vf in the market is only USWOJ which is J bin = 3.5v - 3.75v.
As you mentioned earlier that the colour suddenly changed to blue it shows that the LED is overdriven.
Now try use it with only 2D cells + 2D dummy cells,i'm sure if the LED is still ok than you will be surprise how bright it's gonna be :)
 

chesterqw

Flashlight Enthusiast
Joined
May 9, 2005
Messages
1,968
Location
singapore,jurong
maybe the "driver"(if there is one or a resistor)

is adjusted for driving the luxeon vf? and maybe the seoul p4 has a much lower vf rating, causing smoking ham?
 

CM

Flashlight Enthusiast
Joined
Sep 11, 2002
Messages
3,454
Location
Mesa, AZ
When discussing Lux I's or III's or Cree or Seoul P4, some are not understanding when they talk "Luxeon Vf" and "SSC-P4 or Cree Vf" when discussing he ability of constant current drivers to drive one manufacturer's LED or the other. The driver could care less except it will get longer runtime with lower Vf LEDs. Constant current drivers are "not adjusted for driving (filll in the blank for your single die LED) Vf". They are adjusted for driving constant current to the load regardless of *small* process related variations in Vf (I can feel some of you wanting to pipe in and complicate the discussion with buck, boost and buck/boost but I want to keep things simple here)
 
Last edited:

Long John

Flashlight Enthusiast
Joined
Mar 16, 2006
Messages
2,307
Location
Spain, near Cadiz
CM said:
The driver could care less except it will get longer runtime with lower Vf LEDs. Constant current drivers are "not adjusted for driving (filll in the blank for your single die LED) Vf". They are adjusted for driving constant current to the load regardless of *small* process related variations in Vf (I can feel some of you wanting to pipe in and complicate the discussion with buck, boost and buck/boost but I want to keep things simple here)

The vf is important for a setup, also using a driver. When the stock replacement uses i.e. a LuxIII TxxL (vf about 4V, in bad cases more) ermitter, driven at XXmA, a TxxH ( or another Led with lower vf) will get much more mA.
Not to forget we are speaking about a drop-in with nearly no heatsink.

Best regards

____
Tom
 
Top