Pesky BenBoost Drivers

[DTR]

Re: Pesky 12x diodes

A higher input voltage should only be a problem if it exceeds the vf of the diode..

This could be the issue. My test load drops about 4.4V @ 1A when simulating a 445 diode. So that is certainly plausible it is what is going on and that would not be an issue with a 6V forward voltage and on a 445 the spike from 940mA to 1050 would not be any problem plus the voltage forward would be even higher if running two in parallel for 1.8A or so.

 

[JLSE]

Re: Pesky 12x diodes

From what ive read, the 12's have a vf of 6v > http://laserpointerforums.com/f38/12x-murder-fund-graph-45355-13.html

Not sure if that is the case for the diodes you are using.. but if thats the case
the battery would have to exceed that 6v, which I dont think would be the case..

Id start with looking closer at the driver, but I have never experienced this
with the 3410.. Maybe a design issue of another sort..

 

[ENX]

Re: Pesky 12x diodes

Mine did die while it was turned on.

I tested it for about 2 seconds, after connecting the driver to the diode and then again for 2 seconds, after soldering and mounting the pill to the heatsink.

After getting everything assembled in host, I went for the first run... Just started to burn a piece of black paper and the light went out.

Probably unrelated to your issue, but wanted to commiserate.

Hope you all have a great weekend.

 

[benmwv]

Re: Pesky 12x diodes

Ryan, probably should have tried it out with another driver first to see if it could handle the current, then test it on the first driver to see if it's bad.

I haven't personally tested any benboosts on a 12x. Surely there are other people who have and could tell us something. They've been out and for sale for a while now.

Maybe since 12x diodes are more sensitive the output caps should be doubled up when using them. The caps were originally 22uf, then we found 10uf worked fine (on 445 diodes) so we changed them.

I would be very interested to get them tested on a scope (and one with 22uf caps too). I'm pretty sure rhd has one, but I don't know if he's tested them. Seems like I haven't heard anything from him for a while?

DTR, I also think that current rise you are seeing is because of Vin>Vout.

I really hope people aren't losing 12x's because of benboost drivers!

 

[Jubathoph]

Re: Pesky 12x diodes

I just killed that nice little osram diode DTR sent me. Ben boost was set at 380 and worked fine on test load. Lased great for a few minutes on and off, then it was dead next time I turned it on. I would attribute this to my own mistake however because the osram perhaps should not be pushed that hard.
Too bad my 100th post is a death report I was looking forward to writing up this build then it dies shucks.

 

[DTR]

Re: Pesky 12x diodes

DTR, I also think that current rise you are seeing is because of Vin>Vout.

I really hope people aren't losing 12x's because of benboost drivers!

After thinking about it I am 100% sure that is correct about the forward voltage with 445.

I do have a question for you though. On high voltage forward low current didoes like the Osram and the S06J's have you done any testing on this circuit for these conditions? Also what is the max output voltage on this driver supposed to be?

 

[benmwv]

Re: Pesky 12x diodes

The only testing I've done for higher voltage diode was on a test load. I've never personally used them on BR diodes or the single mode blues. It should work better on higher voltage, lower current because those are closer to the conditions the chip was mainly designed for. In the datasheet most examples are driving strings of LEDs at low current.

I'll have to look on the datasheet to get the exact number, but I know the lm3410 can boost over 20v.
Edit: Looks like 24v is the max.

 

[JLSE]

Re: Pesky 12x diodes

The only testing I've done for higher voltage diode was on a test load. I've never personally used them on BR diodes or the single mode blues. It should work better on higher voltage, lower current because those are closer to the conditions the chip was mainly designed for. In the datasheet most examples are driving strings of LEDs at low current.

I'll have to look on the datasheet to get the exact number, but I know the lm3410 can boost over 20v.

They can go up around 28v, but with a low current application.. A note on the
caps.. I was using over 100uf on the output side. If you look up some notes
in the search from IgorT, he was also using this IC for some time and was
also for 405's from the PHR up to the 12x when he went MIA..

 

[benmwv]

Re: Pesky 12x diodes

They can go up around 28v, but with a low current application.. A note on the
caps.. I was using over 100uf on the output side. If you look up some notes
in the search from IgorT, he was also using this IC for some time and was
also for 405's from the PHR up to the 12x when he went MIA..

It would probably be a good idea to beef up the capacitance for 12x then. Maybe 2x 1206 22uf in parallel. I wonder how big of caps you can get in 1206 size?

You must have been using tantalum or electrolytic caps? I'll give it a search tomorrow, I'm going to bed!

I'll make a revision on the board tomorrow and post it in the thread, recommended for sensitive diodes.

 

[JLSE]

Re: Pesky 12x diodes

IIRC they are 0805 tantalum's, 100uf plus a couple of small ceramic to help suppress ripple.

I came to a combination while running on a scope for the PHR's and 6x diodes... There was talk
that the 405 LD's may be sensitive to ripple as well, so I was being over protective so to speak.

If memory serves, IgorT had a combination of tantalum and ceramic on the output
the tantalum's were around 50uf and that was for 110mA PHR builds.

Also keep in mind that I was driving smaller loads, with a smaller coil etc.. Your best
bet would see on a scope what the issue is first (if any), it may be completely coincidental..

 

[benmwv]

Re: Pesky 12x diodes

I suppose I'll make 3 versions, one standard, one with double 22uf ceramics, and one with a tantalum and ceramic combo. Then I'll find someone with a scope and send them for testing.

If there are issues they seem to be fairly localized to only a few drivers, considering how many have been made/sold so far. I have a feeling that quadrupling the output caps (with 2x 22uf) should be enough to stop any issues without significantly changing prices.

 

[JLSE]

Re: Pesky 12x diodes

If you're looking to suppress at a low current like 200mA or less, that may work..
Theres been extensive testing on that IC by a few members here, just keep in mind
that the specs given in the datasheet is in relation to LED's.. Conditions for sensitive
LD's especially 405nm need special care..

Even with 445's, they may be able to take a lot more current, but running them at upper
limits, these same concerns may become an issue.

 

[tsteele93]

Re: Pesky 12x diodes

Hmm, this might explain my 405 killing spree. Where do I need to put the cap to stop this bloodshed?

I am using 0805 22uF 16v x7r caps, is that good enough or should I stack the caps? And which ones? Thanks!

 

[benmwv]

Re: Pesky 12x diodes

Stack the one between the + and - leads.

 

[JLSE]

Re: Pesky 12x diodes

Hmm, this might explain my 405 killing spree. Where do I need to put the cap to stop this bloodshed?

I am using 0805 22uF 16v x7r caps, is that good enough or should I stack the caps? And which ones? Thanks!

I had numerous discussions with and old member here when I began using the
3410 a couple years back.. I had 30 or 40uf originally and that was for a
phr running at 125mA.. He laughed at me :undecided:

I ended up with 100uf plus ceramic caps (cant remember the values) to
suppress ripple. He was using similar if not more for the same diode..

So if you are powering a 8x or 12x, depending on current, somebody
is going to have to hook up that kind of load and get it on a DSO-scope
and start adding suppression til you get a clean output. The more current to
the load, the bigger the spikes and ripple.

Im really surprised these drivers began selling with no testing? I was using different
layouts and heavy testing for a year or so before I even considered selling
any.

Also, these are being used in parallel which is not even mentioned in the datasheet..
I have no idea what else may be occurring there, but i'd get
some real testing in before hooking up to a 50$+ diode...

 

[Hiemal]

Re: Pesky 12x diodes

You guys need to realize that a lot of capacitance is not as needed.

Think about how filtering works. The more spaces there are between the peaks of the voltage, the more effective the ripple becomes.

Since these ben boost drivers are running at 1.6 mhz, ripple shouldn't be too much of an issue. There's very little "space" between the peaks of the voltage output.

That's not to say the capacitance isn't the issue, it's just something to think of. Typically the higher the frequency the less capacitance is needed to effectively make it DC.

Alternatively it could be input capacitance causing issues. I'm not sure.