FREE DIY open source BOOST driver!!! Tested & working!!

[rhd]

Re: Ben Boost and 12x diodes

Not to bash on Benmw, but everyone using the Ben Boost should read entirely through this thread --> http://laserpointerforums.com/f38/pesky-benboost-drivers-74518.html

The Ben Boost might not be a suitable driver for the sensitive 12x diodes, especially at enthusiastic currents. Apparently, extra capacitance between the Diode Positive and Diode Negative might fix this issue. But just be aware, this could be the very reason why many of us have mysteriously blown our 12x diodes even at reasonable currents of 450-500mA.

Not to bash Ryan, but that suggestion doesn't make sense to me. From memory, the least important (perhaps even an unneeded) capacitor is the one between the diode's +/-. In fact, I'm not sure the datasheet ever called for that one.

If you want to beef up capacitors, beef up the other caps (or all of them), but not the diode cap alone. We kicked the caps down to 10uf from 22uf, but there's no reason (other than ease of availability) not stick with 22uf (or higher).

I've used derivatives of the Ben Boost with SM blues, 12x SO6J and BW512Ls, and I've never encountered issues. That said, I've never purchased an actual "Ben Boost" from someone else, so I can't speak to the actual commercialized versions that are out there now (in fact, I'm not even clear on who IS selling them now), nor to whether there could be any QC issues. I remember before I headed off for the last month or so, that a bunch of the people looking into selling them were also just learning to solder for the first time.

 

[Kizdawg]

Glad you got moved ok! I LED'd another pl450 on one set at 350ma last night. The build went seamless I thaught untill I switched it on. I guess it could have been an inefficient diode I dont know for sure. The board ran stable current on the test load so I figgured the diode would have been just fine. I have one running on a microboost at 412ma in the B50 you sold me. Anyway's it's good to hear what you have to say about them I know you have spent more time with them than most of us have.

 

[tsteele93]

Re: Ben Boost and 12x diodes

Not to bash Ryan, but that suggestion doesn't make sense to me. From memory, the least important (perhaps even an unneeded) capacitor is the one between the diode's +/-. In fact, I'm not sure the datasheet ever called for that one.

I think wannaburn suggested that one in this post:

http://laserpointerforums.com/f38/pesky-benboost-drivers-74518-2.html#post1077777

Not between the LD +/- though, I think there is some confusion there.

Here is what he actually said about the 3417

http://laserpointerforums.com/f38/pesky-benboost-drivers-74518-2.html#post1076590

"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.."

I think Tom cleared it up here:

http://laserpointerforums.com/f38/pesky-benboost-drivers-74518-4.html#post1079923

OK i just noticed everyone says putting extra caps between L+ and L- which is not entirely true.

If you read the datasheet correctly, the layout is VERY Important for achieving correct regulation. The output capacitor (Vout and GND!!) should be placed as close as possible to the GND pin of IC. And if you are adding extra capacitance it should be between the Vout and GND. Not between the L+ (Vout) and L- (FB pin)

If you put a 10uF cap around the L+ and L- should be more than enough to lower the ripple. BUT!!! In order for this to work you must first have more capacitance at the output (Vout and GND! not FB pin)! Otherwise you are just increasing the ripple instead of suppressing it.
Hell the L+/L- cap is not needed at all.

EDIT: also datasheets recommends ceramic caps to be X5R, and X7R
and original caps are Y5V which have way wider tolerances

EDIT: And below too!

I've used derivatives of the Ben Boost with SM blues, 12x SO6J and BW512Ls, and I've never encountered issues. That said, I've never purchased an actual "Ben Boost" from someone else, so I can't speak to the actual commercialized versions that are out there now (in fact, I'm not even clear on who IS selling them now), nor to whether there could be any QC issues. I remember before I headed off for the last month or so, that a bunch of the people looking into selling them were also just learning to solder for the first time.

I don't know for sure about the others, but mine came from some shady guy by the name of Mohrenberg.

 

[foulmist]

Please read the Output capacitor section of the LM3410 datasheet!!!!

here is a short quote:

The output capacitor is selected based upon the desired output ripple and transient resposne. The initial current of a load transient is provided mainly by the output capacitor!
The output impendance will therefore determine the maximum voltage perturbation!

I don't see why everyone is insisting on those L+ L- capactior! (this is not the output capacitor) OUTPUT cap is Between L+ and GND!

this is how all switching regulators work output cap is Vout/GND! The ripple and transient response is determined mostly by the output cap. Not Vout/FB cap. (That's another story)

 

[rhd]

Glad you got moved ok! I LED'd another pl450 on one set at 350ma last night. The build went seamless I thaught untill I switched it on. I guess it could have been an inefficient diode I dont know for sure. The board ran stable current on the test load so I figgured the diode would have been just fine. I have one running on a microboost at 412ma in the B50 you sold me. Anyway's it's good to hear what you have to say about them I know you have spent more time with them than most of us have.

That doesn't shock me in the least. That's way above "safe" current for those diodes. So I wouldn't put much / any stock in LED'ing a diode at currents so much higher than datasheet specs.

Please read the Output capacitor section of the LM3410 datasheet!!!!

I don't see why everyone is insisting on those L+ L- capactior! (this is not the output capacitor) OUTPUT cap is Between L+ and GND!

Yep, I don't get it either. I add them across L+ L- (in ADDITION to the true output caps that are actually required for the circuit), because if you don't, people whine. "Why is there no cap across the diode's L+/L- ???", etc etc etc. I don't understand why people care - but you get members (like Ryan I suppose) repeating this flawed mantra about L+/- caps over and over and over, and eventually people believe that EVERY driver design is supposed to parallel a cap with the LD. So, since it doesn't hurt anything, and takes up maybe 5 square millimeters, I usually add them. But they're absolutely not crucial in a number of driver designs. In fact, if you're passing the driver on to a builder who doesn't have much experience, the L+/- cap is probably more detrimental than anything, because that cap could end up discharging into your diode if you're doing sloppy builds.

 

[Kizdawg]

So in the end here do we need to do anything with these? Should we be changing some caps? I dont know. All the 445's I have done work great. Eh next 450 I will do at a more sane current I guess I got 1 at 400ma so I thaught I was safe at 350ma even though I know the data sheet say's 145ma max...

 

[foulmist]

So in the end here do we need to do anything with these? Should we be changing some caps? I dont know. All the 445's I have done work great. Eh next 450 I will do at a more sane current I guess I got 1 at 400ma so I thaught I was safe at 350ma even though I know the data sheet say's 145ma max...

I would suggest replacing the Y5V caps with 22uF X5R/X7R for the input and output cap. (Vin/GND) and (Vout/GND). the Vout/FB pin isn't all the crucial but X5R is better than Y5V.

Also I think the design has to be changed a bit, the output cap is pretty far trom the IC itself and that's the most important design flaw. It should be as close as possible to the IC ground pin. Other than that I think some additional testing would be required on a scope too.

 

[Kizdawg]

Sounds good... I will still make and use them but I am going to be alot more carefull. Hey this has been fun anyway's. I learned how to reflow! I only really buggered 3 boards with too much paste.. I was going to fix em but I forgot them in the toaster when company showed up. lol. The nice purple boards were chocolate brown when I remembered... Eh I finally got a Lpc-826 working in a host so I am happy.. I killed a couple of them too... leist there cheapies the pl450's are killin me...

 

[rhd]

So in the end here do we need to do anything with these? Should we be changing some caps? I dont know. All the 445's I have done work great. Eh next 450 I will do at a more sane current I guess I got 1 at 400ma so I thaught I was safe at 350ma even though I know the data sheet say's 145ma max...

Somewhere I have a design that uses a 68uF tantalum capacitor and a 0.1uF ceramic in parallel as the output capacitor. I can take a look for the board file tomorrow.

 

[Hiemal]

Again, though, you forget that this IC is made for driving LED'S, not laser diodes.

LED's are much more tolerant of spikes, and being overdriven. Laser diodes however, are not.

The reason people are getting upset is quite obvious, because YOUR driver is killing their diodes for unexplained reasons. Can you blame them for attempting to figure out why?

The diodes are dying upon startup. That usually points to spikes, or some other form of over-driving.

Don't be bashing on members because they're having problems. I understand there are elements of quality control that need be understood, but if multiple members are having issues there is quite obviously something wrong.

Do you even own an oscilloscope RHD? How did you test the driver for spikes?

Did you test them out with 12x blu-ray diodes?

I'm not trying to be an asshole here, I promise, but I'm just curious how in depth your testing truely was.

 

[rhd]

Again, though, you forget that this IC is made for driving LED'S, not laser diodes.

That doesn't change the schematic / circuit layout, nor the reality that the circuit HAS an output capacitor (it's just not the one bridging L+/-, which is, as others have pointed out, not really called for).

The reason people are getting upset is quite obvious, because YOUR driver is killing their diodes for unexplained reasons. Can you blame them for attempting to figure out why?

Not MY driver, and not BEN's driver either. I've never sold these drivers to anyone on the forum. Though I've given a few away to some vets who tested them (including on 12x diodes), to positive results. I've also used them on those same diodes that Ryan and a few other beginner builders were complaining about, without any trouble.

When Moh decided not to build all of his boards, he sold the PCBs to a lot of different people, including members with no experience actually reflowing boards. Again, a lot of very new people grabbed some of these boards. I didn't read that other thread carefully enough to see where the boards in question came from, and I'd prefer not to - so that I can make this general comment without anyone thinking I'm singling them out personally. Put some consideration into who reflowed the boards in question, as well as the user reporting the problem itself.

Don't be bashing on members because they're having problems. I understand there are elements of quality control that need be understood, but if multiple members are having issues there is quite obviously something wrong.

Raise your hand if you've ever killed a diode with a DDL driver:
* Multiple hands go up *

Multiple people kill diodes with every driver that gets used in any sort of wide distribution. You have to keep in mind that Moh made 1,300 Ben Boost PCBs, some of which he assembled and sold, others he passed along as PCBs and others reflowed and sold, etc etc etc. A lot of people have purchased and used Ben Boost drivers. A lot of people WILL kill diodes with them. A lot of people who HAVE killed diodes with the Ben Boost are relatively new builders.

Datasheet current on a PL450 calls for a max of like 165mA. Some members were complaining that their Ben Boost driven diodes died when the drivers were set to "just" 380mA. Why? Because other PL450s that they (or others) had tested were able to handle 400mA.

To a really novice enthusiast, that logic might make sense. But it's logic that is divorced from the reality of this hobby, where diodes die at vastly different currents from each other, especially when being run at 2 and 2.5x the datasheet current (which is among the highest departures from datasheet specs I've seen for any diodes - awesome, but risky).

Do you even own an oscilloscope RHD? How did you test the driver for spikes?

Did you test them out with 12x blu-ray diodes?

Yep, and nope (just on 445s). Though I think the spikes (if there had been any) would have shown up when the driver was used on a 445, even if 445s were better able to handle them.

Granted, the Ben Boost is about the ONLY thing I've actually used my scope for, and I'm not a scope-using expert. But it (driving a 445) was the first thing I tested when I got my scope, and the output looked clean to my eye.

There's room for someone with a better analog scope to jump in here and confirm, but I think you'll find that there's nothing wrong with the drivers (or more accurately, with the driver design). If anyone wants to tackle a scoping, PM me so that I can send you one of the drivers that I've reflowed myself (or grab one from Ben if he has some on hand), as opposed to testing one of the randomly reflowed versions out there.

So again, don't forget that beginners like Ryan kill diodes all the time. That's not "bashing", that's just reality. This is a "death heavy" hobby, particularly when users are only a few months in to learning how to pull off builds (but even after that). It's frankly not "bashing" to suggest that new builders burn diodes. That's just "recognizing reality".

Keep that other thread in context. I skimmed it, and I saw a bunch of members with a few months of presence here, and not very much rep, and without (to be fair) a history of builds, complaining about dead diodes. Some vets (like DTR, TJ, Wannaburn, etc) jumped in and tried to help, but they weren't the users reporting the actual problems (from what I read, though I might have missed someone - apologies if so).

 

[Wolfman29]

I would +rep you, rhd, but the system won't let me.

But god damn - eloquent even on the internet

 

[Blord]

I think it is the nature of the 12x diode. They are hypersensitive. When I killed them that was in the time the Ben Boost wasn't invented yet. They were connected to the Flexdrive, the only available driver then for one battery.
I have also a PL450 running at 373mA with the Ben Boost without any issues. That is a triple overdrive of standard value. And I never killed a PL450 before. ***knocking on the table***

 

[Hiemal]

RHD, I do apologize for sort of bursting out at you like that. I understand you guys put in a lot of time into the driver, and a lot of thought into the layout of it and whatnot. I hope I didn't piss you off too much. My wording probably wasn't the best in the world...

Just hope we can narrow down the problem of the driver to see if there really is a problem or not, that's all.

 

[Kizdawg]

Me to rhd +1 these are experimental DIY drivers that is what I have said all along.. if someone is selling them because they think there fine then so be it but that would be there own fault.. I remember the talk about who to pay what to... lmao.. for a DIY??? that doesnt make sense... No one ever said wether the OVP boards would help or not yet either.. all my burnt diodes have been on round drivers just to clarify.. I can make about 25 OVP boards too.

145ma max on the spec sheet... btw..

 

[rhd]

There's nothing for anyone to apologize for. I'm just conveying my perspective. I'm also cognizant of the fact that while these drivers are named after a member here, the actual drivers everyone is discussing weren't assembled and quality checked by him. That's an important thing to keep in mind. If a Mohgasm driver failed, you wouldn't blame Daedal.

I'm all for putting drivers to the torture test. I generally say that about most things in this hobby, including things I've had some part in creating. That is (or should be) the idea of this hobby. Failure is failure, and nothing irritates me more than when people try to pretend it hasn't happened. That said, the same perspective cuts two ways. When really novice builders blame the equipment or components because they're not willing to believe that they could be at fault themselves, I roll my eyes (and not with an emoticon, I roll the real ones).

I frequently see "vets" (the ones who actually build) owning up to the diodes that they've killed, discussing their pain, mourning their losses, and exploring the sense of emptiness that comes after a particularly bad run. If vets can admit to killing diodes, then why is it so hard for new hobbyists to swallow the realization that they've murdered a diode? The quest for coherance is literature with a lot of dead InGaN. That's life.

Actually, it's not life. It's a wonderful, but completely non-essential hobby, that costs a lot of money, a lot of patience, and a lot of humility to participate in.