Magic smoke...

[unnerv]

I just let the magic smoke out of one of waynes BB750's. Right after I got it all epoxied into a McModule. At least my V3U seems okay. It looked like there was enough clearance on the positive lead to the luxeon but when I went to focus, poof! /ubbthreads/images/graemlins/mad.gif I quickly backed it off and tried it on momentary and it still lit up. I pulled the bezel off and checked the voltage and was only getting 5.5 volts. I managed to pull the board out of the epoxy with a new xacto knife and some brute force. Ah well back to the drawing board.

 

[tvodrd]

unnerv,

S**T happens! I just managed to ruin a BB500 during final potting on an "unobtainium" LED. You have my sympathies!

Larry

 

[unnerv]

thanks, I guess if nothing else at least it's good for Wayne
/ubbthreads/images/graemlins/wink.gif

I think I just need to stock up on this stuff for emergencies like this. It kills me that the shoppe is only about 1/2 hour away, but since they don't have a retail front...so close yet so far away. Anyway good luck with yours. Hope you have a spare and didn't damage the LED

 

[kakster]

This may make you feel a little better : I was fiddling around with a minimag reflector, trying to make it fit into my 5 watt Mclux. Somehow or other i managed to knock a lead off the luxeon without noticing. Fired it up, nothing happens, scratch head in bafflement. I then noticed the broken lead and also the now-familiar rotten stench of fried Badboy.
Here's the best part; I potted the bezel cavity full of Arctic Alumina, and am now stuck with a useless McLux bezel until i can find someone to grind it all out for me. /ubbthreads/images/graemlins/banghead.gif /ubbthreads/images/graemlins/banghead.gif

 

[tvodrd]

unnerv,

I had the pleasure of meeting him at the CPF Get-to-gether. I'm currently working on an expedient fix, but need his input! It wouldn't hurt to PM/EM him and see it if a Will-Call is possible. (Speaking of EM, Wayne, pls check it /ubbthreads/images/graemlins/grin.gif )

Larry

 

[unnerv]

I check last time I had an emergency. Wayne and Cindy have been great. My last rush order Cindy got it out that afternoon and I received it the next day. I could probably use a cooling off period anyway, so I am just going to let the order take whatever time it takes Good luck with your project.

Patrick

 

[BentHeadTX]

I smoked some Badboys myself,
First one was a Badboy 500 first run a few months ago, I broke a lead on the LED and received some smokey treats. A month ago I blew up a BB500 board, not sure why. That was OK though, replaced it with a Madmax+ in my 2D M*g R2H. The blown BB500 sandwich was replaced with a BB400 Q3L and it is brighter, cooler and lasts longer. Sometimes there is a silver lining in those magic smoke clouds.

 

[dat2zip]

That's a bummer. Sorry to hear about this. Comes with the frontier we are forging. /ubbthreads/images/graemlins/smile.gif

tvodrd,

I mailed a special package to you. Did you get it? Also, the wiring diagram images are now up and posted.

 

[tvodrd]

Hi Wayne,

Haven't gotten it yet. I seem to have saved my assembly by "sandwiching" another BB500 on the bottom of the potted dead one. That microconvertor in my order you delivered at the get-to-gether is in the process of becoming a module this afternoon. /ubbthreads/images/graemlins/smile.gif

Larry

 

[tvodrd]

Wayne,

Got it! And thank you very much! I hope you put a choice R2H on the extra board and got rid of the "greenie." /ubbthreads/images/graemlins/grin.gif

On topic- I lost the magic smoke from a Micro Convertor today. Do not accidently ground the +LS connection with it powered! /ubbthreads/images/graemlins/mecry.gif /ubbthreads/images/graemlins/banghead.gif

Larry

 

[CM]

I modified a BB for open circuit protection since I'm constantly using it as a LED tester and the likelihood that I'll leave it open is pretty high. It's rather easy if you have parts and the ability to work with 0603 SMT resistors. The circuit in the BB's (also DB's and wizard I think) operate in current sense mode. There is also an unused mode, voltage sense, for voltage regulation that is not utilized. Pin 1 is the voltage feedback that I isolated by cutting the pad from the ground connection. Next I soldered a 2M resistor between pins 1 and 2, then a 260K resistor from pin 1 to ground. The circuit servos to maintain a voltage of 1.263V at pin 1 and the two resistive dividers above sets a voltage limit of 10.7V which should work for 1W or 5W LED's. Below 10.7V, the circuit operates in constant current mode. The data sheet has more details on how to set the voltage limit. I did this mod after frying another tester circuit by accidentally leaving it unloaded. This regulator is definitely my favorite for LED circuits since it's so versatile. Drawback is the MSOP package which requires a *real* PCB for utilisation.

CM

 

[georges80]

[ QUOTE ]
CM said:
I modified a BB for open circuit protection since ...The circuit servos to maintain a voltage of 1.263V at pin 1 and the two resistive dividers above sets a voltage limit of 10.7V which should work for 1W or 5W LED's.
CM

[/ QUOTE ]

You need to be careful, 10.7V 'might' be safe. I know that at 12V on my HP powersupply, with current limit set low that the very short spike that occurs due to output caps being fully charged to 12V is enough to 'blow' a 1W Luxeon. i.e. the 12V at the output cap has enough 'punch' that the 1W Luxeon get's damaged before the powersupply can drop the voltage down to meet the current limit.

Just a note to not assume that just because you have a current limit that the higher voltage is 'safe' to the load...

george.

 

[tvodrd]

I went to salvage the LS from the microconvertor pill a couple of posts back, and discovered the wire had broken, not shorted/killed output transistor. I reconnected it and it lives. /ubbthreads/images/graemlins/grin.gif That microconvertor survived a few dozen turn-ons with its output open-circuited! Different critter from a BB!

Larry

 

[CM]

georges80,

I'm well aware of the limitations of this setup. The limit is strictly a stop gap measure to keep the regulator from blowing in open circuit cases. It is meant to protect the ckt, not the LED. However respectfully take some issue with some of your comments.

First, the output cap of the switcher is probably considerably smaller than the output cap of your HP supply. I bet you have at least a couple hundred uF electrolytic in your power supply. Yes, that and the 12V will probably blow a 1W. I don't think the ceramic cap in the BB has the oomph to pop a 1W LED since it's probably no more than 10uF, probably more like 1uF due to the switching frequency of the regulator.

Second, switchers are a different beast than linear supplies in that their transient response is considerably faster. This one has a response time on the order of tens of microseconds (due in part to smaller inductors and caps from the 1MHz+ switching frequency). This is probably enough to prevent the LED from being destroyed. We've routinely seen these LED's overdriven well in excess of an amp so I think that statement has some element of truth to it.

Finally, that's my story and I'm sticking to it /ubbthreads/images/graemlins/grin.gif

CM

 

[georges80]

[ QUOTE ]
CM said:

First, the output cap of the switcher is probably considerably smaller than the output cap of your HP supply. I bet you have at least a couple hundred uF electrolytic in your power supply. Yes, that and the 12V will probably blow a 1W. I don't think the ceramic cap in the BB has the oomph to pop a 1W LED since it's probably no more than 10uF, probably more like 1uF due to the switching frequency of the regulator.

Second, switchers are a different beast than linear supplies in that their transient response is considerably faster. This one has a response time on the order of tens of microseconds (due in part to smaller inductors and caps from the 1MHz+ switching frequency). This is probably enough to prevent the LED from being destroyed. We've routinely seen these LED's overdriven well in excess of an amp so I think that statement has some element of truth to it.

Finally, that's my story and I'm sticking to it /ubbthreads/images/graemlins/grin.gif

CM

[/ QUOTE ]

Well, it was just a 'warning'. Give it a go on some Luxeons and see what happens - if they survive, no harm done...

I assumed it was a stopgap measure - to save buying a powersupply with current limit /ubbthreads/images/graemlins/tongue.gif

Driving them at over 1A means nothing - it's the transient high voltage (and we're only talked 12V) that will destroy the LED.

I'll happily let you do the 'test' with your Luxeons /ubbthreads/images/graemlins/grin.gif I don't plan to destroy any more of mine...

Note, the HP supply has current limit on the output - so it will not provide a huge current pulse from the output electrolytics - it will current limit. Again, the damage is done by the high voltage transient.

I'll have to hook up my current probe to my scope and do a transient test to see what the HP supply's output looks like when the load is first applied (I'll use a load resistor...).

george.

 

[CM]

George,

Here's a snippet from DS23 from Lumileds:

"All products are not sensitive to ESD damage (+/- 16,000V by HBM condition"

(HBM is human body model--CM)

If they can survive 16kV using the HBM, then 12V is a yawner for these babies. Something's definitely wrong if 12V smoked one of your Luxeon's. I'll do a similar test on an o-scope with the BB on Monday at work. I suspect I'll find a waveform similar to the transient response test on the LT data sheet.

Cris

 

[georges80]

CM - ESD is tested with many kv but only a few pF of capacitance (something like that) - i.e. uJ of energy.

I hooked up the scope & current probe. With a 10 ohm resistor I saw about 2msec from 12.8V down to 3.5V (350mA current limit). Of course for that small amount of time the current was actually up at 1.28Amp - makes sense, 12.8V/10 ohm - exactly what the current probe showed.

So, it does take the HP a couple of msec for the current limit to respond to a major load transient - probably as you say due to the output caps and how ever HP decided to implement the current limit - no schematics of course...

I guess if you're bored and want to blow up some Luxeons, pick a bunch of caps from 1uF to 100uF, charge them to 12V and dump it across a Luxeon and see when the Luxeon bites the big one /ubbthreads/images/graemlins/grin.gif

All, I know is that if I set the HP to 6V and 'drop' that across that across a 1W it survives. Of course I'm now more cautious and set the voltage to about 4V when 'testing' Luxeons for Vf etc. Fortunately the ones I blew up were some 'future lottery' Mxx bin - no loss /ubbthreads/images/graemlins/grin.gif

george.

 

[CM]

(CV^2)/2 = 57 uJ from 1uF and 10.7V
(CV^2)/2 = 25,600 uJ from 100pF (HBM) and 16kV

It's the squared voltage that drives energy. Using ESD as a baseline, the energy is almost 450 times that compared to the BB with 10.7V limit.

I think even 12V on the BB would be safe. I humored myself and subjected a batwing to 50 turn ons from 10.7V on the BB. No visible damage seen except for tired finger /ubbthreads/images/graemlins/grin.gif

2ms response from a bench supply is pretty good. I'm gonna have to get one of those and dump my ratshack supply.

Cris

 

[georges80]

Cris, sounds like the BB is safe then. The HP supply I have is a hp e3611a - pretty nice model - just under $300 - some on ebay occasionally.

I know I took out a few 1watters when set at 12.8V and 350mA current limit. They then just 'glow' a little. I blew one up and couldn't believe it died, so tried a 2nd and when it died I figured one more for good luck and it died too /ubbthreads/images/graemlins/grin.gif after that I determined there was some pretty good correlation between the supply setting and 1watters expiring...

I presume it's a combination of CV^2 and discharge time. The bigger C will give a longer time constant for a given load - more time to cook/zap things.

george.

 

[CM]

$300? Holy cow. Better hold on to that $50 rat shack supply. Naaaah. Quality comes at a price I guess. Thanks for the info. Another flashaholism-related expenditure.

Yeah, the C is for capacity /ubbthreads/images/graemlins/grin.gif The bigger the capacity, the bigger the damage.

Cris