SuperBulb explosions-long-one pic

Icebreak

Flashlight Enthusiast
Joined
Aug 14, 2002
Messages
4,998
Location
by the river
Disclaimer: Always use eye protection when experimenting with SuperBulbs.

The Subject: WA01185 9.6V, 3.15A SuperBulb in Mag 3D running 2x5x123.

I ran this bulb the first time for close to two minutes and melted the standard Mag reflector to the point that it began to collapse. I was so pleased. This was on 1x5x123. I ran it again maybe 20 times at around 30 to 40 seconds per use with a new standard reflector.

In anticipation of OP PMR and because other folks (much more knowledgeable than I) let me know that double barreling would combat 3A induced voltage drop, I configured a 2x5x123 pack using electrician's tape and special Office Depot inking dowels.

To make the connection with the little positive Mag spring, I crudely yet tightly, twisted a length of stripped 10-gauge single strand house wiring into concentric circles to a size that would connect with the batt pack's anodes and barely brush the interior of the battery tube. The tightest circle dipped downward a bit to seat itself nicely into the positive spring reservoir when under pressure. I made a similar yet smaller, flat spiral for the connection between the stripped (thanks soloco) cathode end of the pack and the original compression spring.

I was sceered of the power so (Can you tell I don't know what I'm doing?) I hooked the pack up to an old rechargeable drill motor for a 10, a 20 and a 30 second run just to take the initial surge of amps out of the pack.

Dropped the parts into the barrel and fired it up. Outstanding! Killer photonic authority. Ran it over a period of several days around a dozen or so times in 30 to 60 second bursts with no problem. (except warping another new Mag reflector which, for some reason, gives me a case of the jollies)

A few weeks ago I was reading CPF and a small palmetto bug appeared on the wall above my monitor. Palmetto bugs are large roach-like insects that find their way into even the most fastidiously kept households here in the south. If you are busy, you can run them off and they will eventually return in hopes of inspecting an empty coffee can which, I believe, they think is transportation to the outside world. They are correct. This is not a bug you want to squash in your home. I guess that's another thread.

Anyways, I spot him with the 1185 from a foot out and unlike all my other lights, he don't like it. He moves away giving me the antennae as he scoots. I'm liking this control over the little beastie but then, POOMPH!...glass crash...dim, dim, dimmer, nothing.

I punched the switch to off, wait a moment and removed the bezel and was a little surprised with my inspection. My apologies here. I have very little experience with photography. I borrowed a co-worker's camera and failed miserably at my attempts. I wanted to show the standard lexan lens because it was a kind of photograph of the crash seen. Also, I wanted to show the standard reflector because there was very obvious evidence of hot glass impacts to its' interior. I will submit an awful, fuzzy photo of the remnants of the exploded bulb in a moment.

The surprise I had was that the filament was still intact, black, but still intact. How can that be? I would also like to know what caused the pastel chartreuse powder to dry paint itself to the inside of the remaining shard of the bulb.

From there I thought about what had gone wrong.

Was it some unnoticed bit of fingerprint oil? I didn't think so because the bulb had many successful runs. Was it something to do with those crude spirals of copper? I thought, "Maybe."

Then, a wonderful thing happened. A cardboard tube arrived at my office. In it were two Perfect Mag Reflectors with Orange Peel texture. Man, are those not beautifully superior examples of craftsmanship? I dropped one in a new Mag 3D. Took the 2nd to last Ginseng GB, Carley potted, Welch Allyn 01185 from a ziplock bag and carefully cleaned it with rubbing alcohol. I let it dry for a moment and installed it, making sure that at no time the glass was touched. I then installed the copper DB anode piece, the 2x5x123 batt pack and the copper cathode piece; finally and gently bring home the end cap with compression spring.

I fired up the rig. That was incredible. Best beam I've ever seen in real life. Until, 15 seconds later POOMPH!...glass crash...dim, dim, dimmer, nothing.

Karap! Now I'm down to my last 1185. I have a renewed suspicion of the positive copper spiral. I've wrapped the outside of this piece in electrician's tape thinking that maybe it shorted the first and second bulb because the high level of power blew right through the anodizing of the inside of the battery tube.

Surely, the possible solutions are to get more CAWA01185s and change the 2x5x123 configuration to rechargeables.

Still, I would like to know how a stable system became unstable and what that greenish yellow powder is.

The teenage palmetto bug was last seen while he was flung then flew away from the coffee can transport, landed safely, scampered toward the end of the concrete steps, hesitated, gave me the antennae and wriggled into the surrounding grass. I miss him. I'm sure we'll meet again. What a rebel.

This blown flashlight bulb, I submit for your forensic skills.
WA01185.sized.jpg
 
I'm not sure I follow your battery installation method. But if I'm understanding it correctly, shorting of the + to the ground would result in the batteries getting drained, but should not damage the lamp.

Your description seems to indicate that the + connection to your pack was touching the light body (battery tube), and this should not be done. Mag bodies by default are negative ground (negative connection is through the body). But because they are anodised it wouldn't short (unless there scratches or other damage to the anodizing).

If the pack was shorted to the battery tube (light body) the lamp would actually be recieving less power (electricity follows the path of least resistance). The batteries may explode in this scenario, but the lamp should not. A short to the light body should result in a dimmer lamp, and a very hot light body.

The only thing that could cause more power to the lamp would be a series battery connection between your 2 parallel-5 banks (that would result in 30 volts to the lamp).

What I suspect is that 15 volts is too much for the lamp with the parallel pack. With a single bank of 123 cells the voltage drop would be running the lamp closer to spec. With the parallel packs you are probably getting closer to the full 15 volts which may be too much for that lamp.

Just a guess, though...
 
I had read some time back that for every 1 amp draw from 123s, there's a 0.5 volts drop. For paralleling 2 packs of 5x123, you're talking maybe 1.67 amps to each stack of 123. That's 4.175 volts estimated voltage drop total from each stack of 123s, resulting in at least 11.35 volts (estimated) to the bulb which is overdriving it quite a lil, of cos since you've dropped the voltage a bit, it should be around 10 volts, which is a modest overdrive.

I feel for you man. /ubbthreads/images/graemlins/frown.gif It shouldn't have burst like that. That's one PMR down the drain... /ubbthreads/images/graemlins/frown.gif

That pastel coloured thing in the remaining shard of glass, is it the same one inside the PR base? Is it possible that during potting, they have somehow cracked the bulb during the drying of the compound (maybe expansion during drying or something)

My condolences man.... /ubbthreads/images/graemlins/frown.gif
 
woah, dude! maybe try using a bunch of 9 volt batteries ruung in paralell, or popping a 9-volt LDO regulator between batteries and bulb?
my 2 cents fell in da gutta

neg

[EDIT: changed Series to PAralell, it make a difference...]
 
Sounds like 5S2P (5 series cells, in 2 parallel strings) is too hot a source to stuff through the '1185. If you're bent on 123's, try 4S2P and see if it still flashes on you.

Chris
 
I only successfully use 12 CR2 batteries with my 1185 in a 4x3 configuration. The 5x2 123 stick I use with all kinds of 12V bulbs that run around 2A. I have 2 3D Mags using it now; one has a 12V 20W MR16 and the other has a WA1166, my personal favorite bulb from Carley or Welch Allyn. The WA1166 has the highest light efficacy I've seen. It's actually slightly underdriven with 10 123s in a 5x2 config. It's perfectly driven by 6 123s in series. I had to figure this out with trial and error and my multimeter. I'm using Voltage per Cell = -0.4159 * Current + 3 in all my 123 calculations. Using this formula you are getting 11.55V and probably more amps then the bulb can handle. But I'm sure it was bright! /ubbthreads/images/graemlins/thumbsup.gif
 
Icey,
The bulb blew from being overly stressed. I had a failure in the same mode, complete with intact filament, powder and fragged capsule. I suspect that the capsule became stressed due to the extreme heat generated in the small space. Bulbs of this power and larger tend to use larger capsules to dissipate the heat. The fact that the filament is intact is a clue that while it could take the power, the glass could not.
WA00185Ginseng.sized.jpg


Here's my original post on this failure.
WA01185 Explosive Failure

BTW, welcome to the "Inner Circle" /ubbthreads/images/graemlins/grin.gif

Wilkey
 
Oh, the dust is likely to be the various tungsten halides that precipitated out from the rapid cooling and decompression of the capsule. These would normally recycle back onto the filament under heat.

Wilkey
 
illumiGeek -

Thanks for your effort and valuable guess.

So, it wasn't a short of any kind. That's very helpful.

Nerd -

I appreciate the calculations and condolences.

When the second bulb blew up in the PMR, I started speaking a different language. It had white and chartreuse smoke damage all in it. After a very careful cleaning it shows some tiny impact marks which, upon close inspection, are imbedded with that chartreuse stuff. The marks are minor and I'm happy to say the PMR is not ruined. The Mag reflector didn't fair well at all.

I'll hold onto the remains of the bulbs in case a ModMan wants the PR bases.

Neg -

Regulation? There's an idea. Probably past my low skill level though.

Tweek -

Using soloco's -0.4159 * 1.67a = -.69v + 3v = 2.31v * 5 = 11.55v, 4S2P would be 9.24v. I would think that would work. Thanks for the suggestion.

soloco -

Thank you for the 4x3 CR2 configuration and especially to calculation to get the voltage drop on 123s.

1166 sounds great. With the minimum buys at WA and CA this stuff can get a little expensive. I'm going have to bite the bullet soon. Oh well, it's fun and all that power can be incredibly useful.

Wilkey -

How did I miss that post? Great pics and explanation. Tungsten halides, eh? That makes sense. Pretty color.

BTW, I sent you some emails with attachments a few days ago.

Thanks again, gentlemen. Where else in the world could I get all this accurate information so fast?

- Jeff
 
actually, regulation is simple. all you need is a reg chip that can handle loads in excess of 2 amps ( i think they dont exist......

SCRAP THAT.....
reg chips only go up to 100 milli amps......
SCRAP REGULATION!!! its even out of MY range, and im pretty good......


as far as i can say, either go with 4x2 or 4x3 to fix ya prob, but i cant think of anything else. how much do these Superbulbs cost? if they are reasonably cheap, buy some more and experiment. damn, i want one of these babies.....
 
They aren't that expensive but they are a pain to get. I got a few on a group buy a very kind member organized. Organizing this kind of group buy is challenging to say the least. Don't quote me but I think Welch Allyn has a $100.00 minimun, Carley has a $50.00 minimun. I've got to start taking better notes. Anyway, you buy from Welch Allyn then have Carley pot them in PR bases. Links? Of course I didn't save those either. Shouldn't be to hard to find though.

Theeeeen to run at this heat level you've got to use a PMR or make a NSPMR.
 
[ QUOTE ]
Neg2LED said:
actually, regulation is simple. all you need is a reg chip that can handle loads in excess of 2 amps ( i think they dont exist......

SCRAP THAT.....
reg chips only go up to 100 milli amps......
SCRAP REGULATION!!! its even out of MY range, and im pretty good......


as far as i can say, either go with 4x2 or 4x3 to fix ya prob, but i cant think of anything else. how much do these Superbulbs cost? if they are reasonably cheap, buy some more and experiment. damn, i want one of these babies.....

[/ QUOTE ]

Sure they do. The LM350 will regulate 3.5A. The LM338 will regulate 5A. Both have a dropout voltage of around 2.5V at maximum current.
 
Is that a home potted bulb into a PR housing using thermal compound? Seems to me, most the high power bulbs I've seen either take the heat out through the leads or have a ceramic potting material, which is a whole different beast in the first place.

Any possibility, it could be cooling the bulb housing unevenly, causing extreme stress on the bulb glass.

Boom there goes another bulb.


Neg2LED said:
SCRAP THAT.....
reg chips only go up to 100 milli amps......
SCRAP REGULATION!!! its even out of MY range, and im pretty good......

Humm, plenty of chips exist that go into the hundreds of amps, here is a photo of one switcher I built that does +50A.

http://www.molalla.net/~leeper/jarps2.jpg

Each one of those coils is rated for 14.2A (28.8A at 30% inductance reduction), I put two in parallel, its a dual phase supply, built on 4 oz. copper,each MOSFET is rated for 75A each. You'll notice the sense resistors are made from litz wire.

I've even got a few switching supplies I built around here that will run 200A continous. Total output is 1000W (you guessed it, it was a 5V switching power supply with at 200A continous output). (what do I use for a load bank? A large sheet of copper which gets hot as heck)

Let me see if I can find a photo of the load bank somewhere.
Here we go:
http://www.molalla.net/~leeper/jarpsu.jpg

You'll see a 1/4" copper plate with 9 25W resistors with thermal paste to transfer the heat. Even so, you can boil water on the plate with no problems, when it is being used. With good heatsinking, you can push it further than the resistors themselves are rated for.

Ah, found a better shot:
http://www.molalla.net/~leeper/load.jpg

You'll note the near mirror finish is now oxidized, and you can see a spot above the third resistor from the left, where I melted solder on the plate when I really loaded it hard.

I think the biggest issue is finding low cost power supplies that I can use to power my switching power supplies from.

Guess where I get my efficiency kick from (hounding on power supply efficiency)? Remember, any loss in the power supply goes up as heat, notice that +50A supply in the photo doesn't even need a heatsink on it...you can hold your fingers on it without a problem.

Heheh, what do you do for a fun hobby?

(***thread hijack***)
 

Latest posts

Back
Top