Best Step down converter IC for 4 AA design?
- Thread starter dat2zip
- Start date
Help Support Candle Power Flashlight Forum
- Status
remuen
Enlightened
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by dat2zip:
Subject says it all.<HR></BLOCKQUOTE>
It depends on what you want to drive with it
For a couple of Nichias
I would recommand the LT1932 from Linear Technology. Its advantages: Only a few SMD parts needed (no SEPIC) and a high efficiency up to 95%. Can also be used for step up conversion without any changes in the design. And it is current controlled and has a dimming function. And your brilliant idea with the Zener diode can also be applied to this chip.
Ok, now the next one with a better idea ......
Subject says it all.<HR></BLOCKQUOTE>
It depends on what you want to drive with it
For a couple of Nichias
Ok, now the next one with a better idea ......
remuen
Enlightened
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by dat2zip:
It might be interesting to see if the bare LED mounted good to a very large heat sink might take quite a bit of overdrive. <HR></BLOCKQUOTE>
Two days ago a have driven a bare LS Emitter with 600mA more than one hour. I mounted it on such a huge heatsink it is used for power amplifier transistors (BD xxx). The heatsink got only a bit warm but not hot. After a few minutes the current was quite stable.
It might be interesting to see if the bare LED mounted good to a very large heat sink might take quite a bit of overdrive. <HR></BLOCKQUOTE>
Two days ago a have driven a bare LS Emitter with 600mA more than one hour. I mounted it on such a huge heatsink it is used for power amplifier transistors (BD xxx). The heatsink got only a bit warm but not hot. After a few minutes the current was quite stable.
Hopefully the step down could drive one LS with plenty of overdrive capabiltiy and in the future drive the newer LS models with higher current demands.
If I get some of the Bare LEDs I'm going to go for the kill. Abuse the bare LED and see how high I can drive the current up. It might be interesting to see if the bare LED mounted good to a very large heat sink might take quite a bit of overdrive. I need to review the thermal app note from Luxeon and see what how high theoretically we can push the poor LED.
If I get some of the Bare LEDs I'm going to go for the kill. Abuse the bare LED and see how high I can drive the current up. It might be interesting to see if the bare LED mounted good to a very large heat sink might take quite a bit of overdrive. I need to review the thermal app note from Luxeon and see what how high theoretically we can push the poor LED.
remuen
Enlightened
No ideas from other flashaholics ........
dat2zip, have a look onto the LTC1778 from Linear Technology even it is in a 16-Pin narrow SSOP package.
It looks quite interesting to me.
dat2zip, have a look onto the LTC1778 from Linear Technology even it is in a 16-Pin narrow SSOP package.
It looks quite interesting to me.
two things, is there a completed circut out there that you can buy, I am too lazy to put one together myself...
Talk to lambda about overdriving luxeons... he has had excellent luck with this and he has a very clean 3d mag mod. he is here on the forums.
or go to
lambda10.tripod.com
Talk to lambda about overdriving luxeons... he has had excellent luck with this and he has a very clean 3d mag mod. he is here on the forums.
or go to
lambda10.tripod.com
Brock
Flashaholic
moving over to the "Electronics Forum. Batteries Included"
Must be quiet out here. Spring break?
I'm still filtering through datasheets from Maxim, Linear and National. I've found quite a few that fit the bill. Most have quite a large physical footprint and or require external components.
Some of them have both voltage feedback and current feedback or current limiting of some form. It might be possible to set the voltage feedback to try to regulate at 3.6 Volts for example and use only the current control to regulate the current. In this way the 2.9V to 3.5V LS voltage would put the converter in current regulation only. And, it would be safeguarded to output 3.6 Volts if there was a no load condition. That's what I'm shooting for right now.
I was trying to hit all the requirements at once but, I think I'm going to dump the 1.5A output (nextgen LS) and just shoot for a regulator that will do Normal (0.35) and overdrive (up to 1A). There are many that meet this and will achieve 90% or better.
It will take me another week to munch through all the different datasheets.
I tend to dislike most Linear ICs since most of them are constant frequency with throttle back mode. They are hard to tame and keep in control. The one shot methods and or combination one shots with comparator feedbacks are far better suited for this type of load range, Vin range and Vout range.
If Vout was truely fixed at 3.3V and only the load current varied as well as Vin. Then, most DC/DC converters would do a decent job. But, having a varying Vout condition such as a standard LS with a large Voltage range throws and addition variable into the mix and most ICs don't handle this well with good efficiency. Especially in the Boost where optimization is hard to come by.
I'm hoping to keep a small real estate board design for boost and a buck design. The boost I've played with are the ZLT+ and the MAX167X series.
I'm also looking for a single cell boost converter to drive 1-3 Nichias efficiently.
My schedule to work on flashlight mods is now more limited since I have other activities kicking in. I probably won't finish another converter mod till mid year or so.
I'm still filtering through datasheets from Maxim, Linear and National. I've found quite a few that fit the bill. Most have quite a large physical footprint and or require external components.
Some of them have both voltage feedback and current feedback or current limiting of some form. It might be possible to set the voltage feedback to try to regulate at 3.6 Volts for example and use only the current control to regulate the current. In this way the 2.9V to 3.5V LS voltage would put the converter in current regulation only. And, it would be safeguarded to output 3.6 Volts if there was a no load condition. That's what I'm shooting for right now.
I was trying to hit all the requirements at once but, I think I'm going to dump the 1.5A output (nextgen LS) and just shoot for a regulator that will do Normal (0.35) and overdrive (up to 1A). There are many that meet this and will achieve 90% or better.
It will take me another week to munch through all the different datasheets.
I tend to dislike most Linear ICs since most of them are constant frequency with throttle back mode. They are hard to tame and keep in control. The one shot methods and or combination one shots with comparator feedbacks are far better suited for this type of load range, Vin range and Vout range.
If Vout was truely fixed at 3.3V and only the load current varied as well as Vin. Then, most DC/DC converters would do a decent job. But, having a varying Vout condition such as a standard LS with a large Voltage range throws and addition variable into the mix and most ICs don't handle this well with good efficiency. Especially in the Boost where optimization is hard to come by.
I'm hoping to keep a small real estate board design for boost and a buck design. The boost I've played with are the ZLT+ and the MAX167X series.
I'm also looking for a single cell boost converter to drive 1-3 Nichias efficiently.
My schedule to work on flashlight mods is now more limited since I have other activities kicking in. I probably won't finish another converter mod till mid year or so.
- Status
