Downboy converter board problem. (MUST READ)

[dat2zip]

I wish I could tell you more. I have not looked into the situation, but, I just wanted to post and give you heads up that there appears to be a problem with Downboys specifically at higher voltages like some of the posts that have come up recently.

I suspect the inductor is too low causing trouble for the IC to handle the faster swing through the inductor.

I'm not sure and will post here when I get more information.

Until then, please test any Downboys in all you configurations before potting. Most specifically 3X123 and higher voltage configuration should be tested thoroughly before installation to weed any suspect converter boards.

Please post here any observations you have and hopefully I will be able to nail down the specifics of what is happening with the converter boards.

Thanks,

Wayne

 

[this_is_nascar]

Re: Downboy converter board problem.

Interesting. How do I find out if I have the problem? I've now got several hundred dollars tied up in McLux-PR bezels that are using the DB-917.

 

[McGizmo]

Re: Downboy converter board problem.

Ray and others,

I test the drivers prior to potting them in the heads. In the last 30 to 40 builds, I had 3 of these drivers that weren't drawing the expected current from the power supply and I put these aside. In light of some of the other comments and mention of some of the DownBoys not functioning properly, I grabbed my *rejects* and did some further evaluation. See, I figured I had done something wrong when I had soldered the sense resistors on and hadn't considered it to possibly a problem of another nature. I hooked these up and in addition to not drawing as much current as they should, they also mis behaved when I increased the voltage, drawing similar amounts of current as the input voltage went up when they should have been drawing less. I provided this information to Wayne yesterday.

I believe I have caught all of the suspect drivers from the builds I have done. What I assumed was my error might in fact be something else. I tested over a dozen drivers and heads yesterday which had "passed" my initial screening and they all behave properly with the variation of the input voltage. Even the bunk drivers I have that were rejected seem to work fine at input voltage levels of 6 volts but again, they are not pulling as much current as they should so they would have been a bit low in luminous output had they been used. Until we know more, if someone is uncertain, I don't think they would need to be concerned unless they are using 3x123 power paks. If there is a problem, I think it will present itself in lower light output than you would expect.

 

[roguesw]

Re: Downboy converter board problem.

Hi guys, can anyone explain to me how to test a malfunctioning board? i am using 3X123 cells to simulate the required voltage, where to i put the multimeter probes to test for amps and what type of current is expected for a DB700 and DB400
cheers guys
Des

 

[dat2zip]

I've identified the problem with the Downboys.

The loop stability capacitor near the edge of the board sometimes is not connected to GND. This causes the board to behave in a very unstable manner. It's more unstable as Vin goes up.

Here is a picture of the area in concern.

See the copper burr in the circled area. Yes, this needs to be removed, but the thin sliver left on the board must stay intact. Breaking this copper connection will cause the "Downboy stability problem". You can scrap the areas where the arrows indicate and add a wire between these two points.

To test whether the capacitor is connected or not, you can ohmeter from the capacitor to GND. One end of the little capacitor should read zero ohms to GND. If you do not get a zero ohm reading, then it is possible this copper burr area was ground down or is not making connection anymore.

Please fix this before installation in the final product.

Wayne

*Edited* Change the subject to include (MUST READ)

 

[kj]

Hi Wayne,

I opened my DB1200 module, which doesn't work with 3x123, and I confirmed the upper side of the capacitor is connected to GND and the lower side is connected to a pin of IC.

Is there any other possibility that can cause this problem?

 

[Beretta1526]

I just got two DB9167's the other day. One is already built, but doesn't seem to be problematic. I don't intend to use it with 3x123, so it most likely won't be an issue.

Wayne, How's this one look?

 

[Rothrandir]

yikes! a downboy 9167!?
are you driving the luxor? /ubbthreads/images/graemlins/grin.gif

that's an amazing macroshot btw! /ubbthreads/images/graemlins/eek.gif

 

[darkzero]

I'm glad Wayne nailed the problem and pretty fast I might add. Although I'm not running any on 3x123 yet it's
good to know beforehand.

DB917s are my favorite along with DB700s for a little less brightness but longer runtime.

Jim's right, that is one heck of a macro shot Beretta1526's got there! Sorry to get off topic but may I ask what your equipped with or better yet how much did the camera cost?

 

[kj]

I'm also glad he found one of the causes, but unfortunately my problem with 3x123 has not been resolved... I hope Wayne can find another possible cause of this problem in the end.

Yeah! Great macro shot! I also want to know the equipment and technique /ubbthreads/images/graemlins/smile.gif

 

[Beretta1526]

/ubbthreads/images/graemlins/blush.gif gee, I'm flattened, er... flattered.

I'm using a Nikon Coolpix 990, but my wife's Nikon 2100 has the same capability as does most Nikon digi-cams.

It's just lit with my desk lamp and taken from about 2" away with a slight zoom. The pic is only cropped, the full image is 2048x1600.

Can anyone tell if this one's OK by visual inspection?

 

[dat2zip]

Beretta,

That one looks fine.

There is copper on the edge in that picture. The green solder mask covers it.

Wayne

KJ

I'm still looking into it. I wish I had an answer for you.

 

[dat2zip]

While this is not exactly on topic, this applies to all derivitives using the LT1618 IC (BB, DB etc).

I Posted the following information elsewhere on the forum and will paste it here for reference.


============= Paste =====================

I can't prove what killed the two BBs in question, but, I've done some bench testing and from that I can say the following:

1) The IC used in the DB, BB according to the datasheet does not mention thermal protection and as such we can not assume the IC will protect itself when it gets too hot.

2) The converter board in free space tends to run OK in almost all conditions. It will degrade as it gets hot which is somewhat an overtemperature control. But, this in itself is not foolproof.

Adding external heat to the converter board will push the IC higher than self generated heat and with a LuxV bolted to the converter board appears to be pushing the converter board temperature up high enough to where the IC fails and it will do so repeatedly.

I have not proven this theory out, but, I would have a high suspicion this is the case.

============= Paste =====================

Wayne

 

[koala]

Meaning ARC LSx that uses LT1618 will also exhibit the same problem? So what claimed as thermal protection of the LED is not there?