ZLT+ additional construction tips and solutions to problems

[**DONOTDELETE**]

Sorry to hear about your ZLT+ problem!

The component with the three legs is the transistor.

What power source were you using? MrAl reported similar problems with a 1.5-volt source dropping below one volt.

The transistor may also fry if, for some reason, the 300 does not turn it off as it should. This may happen if the resistance between the transistor emitter and the 300's Isense pin becomes nonzero for some reason, on the order of even a few milliohms.

One possible cause is cold or oxidized solder joints. I guess the best advice is to make sure all the connections are as clean and solid as possible, especially when resoldering.

Or when adjusting Rsense, maybe the lead that goes to ground should be resoldered instead of the lead that goes to the 300 and the transistor. That way, the critical connection only needs to be soldered once, and there's less chance of oxidation.

Another possibility is solder splatter. Once, when I was testing various inductors, a small blob of solder landed between the + pad and the trace that goes to the transistor base. When power was applied, the transistor was quickly toasted. With circuits this small, even a tiny solder blob can wreak havoc.

 

[JollyRoger]

Hmmm...I see. thanks for your input, Duggg.
I was using two fairly fresh nimh AA's.

It could be the solder splatter? I was not resoldering anything on the board...just soldering the trace on the board to adjust the current (the board I was working on had this current adjusting trace built on to the board).

I don't know....it does seem that this is a very very sensitive (and 'fragile') circuit?

Maybe I should set it, pot the whole thing in epoxy...and never touch it again!

 

[**DONOTDELETE**]

During the project development, I probably resoldered Rsense hundreds of times and resoldered inductors dozen of times.

Only two failures---I blew an LS early on, before I realized how much energy can be stored in the cap. After that, I soldered the connection. The other incident was the blown transistor.

So the components themselves are surprisingly robust for their small size. Now, granted, all these tests were on a much larger PC board, with lots of big traces to quickly dissipate the heat from all that resoldering. And it's certainly possible that heat is more of an issue with the smaller boards.

I don't know about the current-adjusting trace you mention. I imagine takes a pretty long and/or skinny copper trace to equal the resistance of 3-6cm of #32AWG wire.

 

[JollyRoger]

THIS WAS THE FORMER THREAD TITLED:
ZLT+....FRIED

Well, I was adjusting the current on my ZLT+ last night...kept soldering on the little board, turn on, turn off, solder, turn on/off, on/off....And then all of a sudden, when I turned it on once...it lit dimly...then the small piece on the board (the one with three connections--I forget what you call it) started smoking...

This is the second time this has happened. The first time, I replaced the part. But now I don't have anymore...

???
What could cause this? Is the ZLT+ just fragile? Does it not like being turned on/off a lot?

 

[Sean]

I thought that this circuit was really fragile too when I built my first one and fried it.

But since then I haven't had any problems! Once I got it into the maglite everything has been flawless. Now that I got a tailcap switch I can't put this thing down!

 

[remuen]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Duggg:

So the components themselves are surprisingly robust for their small size. Now, granted, all these tests were on a much larger PC board, with lots of big traces to quickly dissipate the heat from all that resoldering. And it's certainly possible that heat is more of an issue with the smaller boards.
<HR></BLOCKQUOTE>

As I know from my own small experience and especially from the one of a r&d engineer here in the company heat should be no problem for these components except you are holding the soldering tip a few minutes on their pins and heat the whole part up to 200 - 400°C. Just to give an idea what these SMD parts can bear: The operating and storage temperature of the FMMT617 is defined from -55 ... 150°C.

My friend is just developing a buck/booster circuit where the caps get about 70°C and the schottky rectifier and the 2 MOSFETs in parallel heat up to 100°C and more under full load (this load is certainly quite a bit higher than only a LS). And even all these parts are very close together up to now none of them got damaged after many hours of testing the circuit under hard conditions.

A missing ESD protection could also be a reason for the failure even the ZLT parts are not as sensitive to it as MOSFETs.

 

[InTheDark]

I had the same trouble when soldering my circuit together, but the components are pretty durable. I've probably soldered them 10 or 20 times, and although they look pretty bad, they still work. I did fry a couple of them, but that was definietly because of my poor soldering skills. I did have problems when trying to pot them with glue, I think because of the small Rsense the glue kinda messes everything up and changes the resistor value. So far, this is the best circuit I've seen for this application, it's simple, easy to make, and relatively cheap. It's a pain to put together, but I guess we can't have everything.

 

[machspass]

I've built a couple of them so far, and they have worked very well. My goodness, my eyes would be in bad shape if I tried to solder these without any optical aid!

I solder these things together with a fine tipped, heat controlled soldering iron, under a low-power microscope. I'm an electrical engineer, and I'm very lucky to work in a well-equipped lab, and have these resources at my disposal.

The microscope and the bright, focused light that goes with it are a great help in finding those stray solder blobs that might get left behind.

 

[papasan]

one of the ways i check my work, besides a decent magnifying glass, is to put it over a strong light...the light will pass through the fr-4 just fine but not through the solder or copper...

and these chips seem very robust to me...i put a PCB that i had already soldered before realizing that the PCB was the wrong layout (forgot to flip it before printing) onto an clothes iron that i had clamped bottom-up and turned on to heat up all the solder at once and easily remove everything...well, the wife got to nagging me and i forgot about it for awhile...when i finally remembered abotu it the chips were all badly deformed from heat...warped like crazy...they still seem to work just fine though =)...

 

[InTheDark]

I just fried my LS while playing with the circuit. I guess you have to be really careful when hooking up the circuit to make sure the LS is connected firmly, I just twisted the wires together and was checking to see if the circuit worked when I heard a sizzling sound.

I hooked up the circuit to a voltmeter to check to see what it was reading, and surprisingly it was reading 24 volts. Is this normal, or did I do fry something else? And by the way, where's a good place to buy another LS? Arc is all out of stock for now, I checked Future and Mark Hannah surplus, but I remember somebody writing about getting green LS's from one of them. Is there another place?

 

[Bushman]

In the dark

The exact things happened to me too but i fried two luxeons before i figured out the the cap was storing up that much energy. I have the board going to the ocilloscop to meausre ripple and to see if the cap is fried. email me and i will forward someother good comments from mr al and douggg

 

[remuen]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Bushman:
email me and i will forward someother good comments from mr al and douggg<HR></BLOCKQUOTE>

Bushman, could you give us these comments here in the thread or make a link to the page where they can be found? IMO a LS is much too expensive to get fried so everything which will help to avoid damage is helpfull to all of us. Thanks!

 

[**DONOTDELETE**]

Here is the email conversation between Bushman and I.

I encourage anyone with problems to post them, rather than email, because ordinarily I consider email to be private, and IMHO not the most appropriate medium for multi-user discussion.

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Bushman emailed,
i am not very good at explaining this but it seems that i am getting 20 VOLTS across the tantlum capicitor and also in output. Fried two luxeons before i figured out what was going on.

funny thing is that i had just tested the circut and trimmed the 32g wire to run it at 360mA and all was fine. I was hardwiring it all together after trimming up the luxeon heat sink card and kabloommmmm out when the luxeon.<HR></BLOCKQUOTE>

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>I responded,
The circuit will indeed develop excessive voltage if the connection between the capacitor and the LS is not solid. I overemphasized this several times on the web page and on CPF.

What I suspect happened was this: the connection to the LS came loose during operation. Without the LS in the circuit, the capacitor was forced to absorb all the energy from the inductor. Then when the LS got reconnected, the cap quickly discharged all that energy into the LS, blowing it.

The same thing happened to me early on the development of the circuit.

To add insult to injury, if you indeed measured 20 vdc across the cap, there's a good chance the cap is also damaged, as it was only rated at 10 volts. A damaged cap will result in excess ripple, which in turn could damage other LS's.

If you have an oscilloscope, you might consider hooking up MrAl's dummy LS circuit with the cap in place, and measuring the ripple that way. But to be absolutely safe, I would replace the cap.<HR></BLOCKQUOTE>

 

[Bushman]

gentleman i agree with the above commments i just am not good at getting things from one place to the other on the computerthanks douggg for doing this and i emailed the other note from mr al to klaus to post also. hope it gets up soon.

 

[InTheDark]

I'm not too familiar with circuitry and all that, but basically what you've said is that if the LS is not hooked up correctly and I fried it, there's also the possibility that the capacitor is probably smoked too? And if I hook up the circuit to an oscilloscope, what should I be looking for in terms of output? How do I measure the ripple current and what is a safe number? I don't want to risk another LS to test it. Also, where is everyone currently buying their LS's since arc doesn't have them in stock anymore?

 

[**DONOTDELETE**]

Remuen, we all know that you and Klaus are the same person!

Just like Switzerland, Germany, and Austria are really all the same country

Just kidding,
Doug

 

[**DONOTDELETE**]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by InTheDark:
I'm not too familiar with circuitry and all that, but basically what you've said is that if the LS is not hooked up correctly and I fried it, there's also the possibility that the capacitor is probably smoked too?<HR></BLOCKQUOTE>

The possibility certainly exists, as the capacitor is only rated to 10 volts. With the LS soldered in place, under normal conditions, it never gets much above 3.5 volts.

I'm not exactly sure how one would test a capacitor for overvoltage damage. Even a simple continuity test is tricky due to the polarized nature of the cap. I would think the capacitance would drop, so perhaps a timed RC experiment could determine C, but with a tolerance of +/- 20%, I don't know if that would be particularly informative.

That leaves the oscilloscope test. According to MrAl, peak output voltage ripple at 350mA should be no more than 150mV above the DC component to ensure the LS gets no more than the 500mA rated pulsed current.

 

[remuen]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Duggg:
Remuen, we all know that you and Klaus are the same person!

Just like Switzerland, Germany, and Austria are really all the same country =)
<HR></BLOCKQUOTE>

Hey Duggg, you forgot to say that we all look the same and that we live in another Universe

.... oh, sorry it's not a Universe it is only the jungle here overseas because a few weeks ago a CPF fellow seriously asked in a post if he also could buy AA, C and D batteries in Europe

. Unfortunatly we have only candles - but did you know that a candle light is much more romantic than a beam of an LS?

Klaus and me are just trying to modify a candle. Our target is a smaller, brighter and less yellowish flame which uses less oxygen

 

[Adam S.]

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by JollyRoger:
[QB]Well, I was adjusting the current on my ZLT+ last night...kept soldering on the little board, turn on, turn off, solder, turn on/off, on/off....And then all of a sudden, when I turned it on once...it lit dimly...then the small piece on the board (the one with three connections--I forget what you call it) started smoking...


This is the second time this has happened. The first time, I replaced the part. But now I don't have anymore...
/QB]<HR></BLOCKQUOTE>

When I was puting mine together I got confused about which way the diode went and had the four negative pins bridging the gap.

Because of this when R-sense was not conected the LS lit dimmly but when I conected R-sense it only got dimmer even when I cut it down way short. So I hooked a strand straight across and the LS did not light instead the transister smoked

, BUT it still worked fine after that and once I discovered and fixed the diode the circuit worked great.

 

[Sean]

Well my once great working mod fried tonight, I was turning it on/off not real fast but fast enough I guess

because something (I think the Transistor) fried!! After 2 weeks of flawless operation in my minimag it just went up in smoke!

Now I need more parts!