Best Chip/Step Up Circuit Design for the new 5W Luxeon Star?

LEDBE
R

remuen

Enlightened
Joined
Nov 26, 2001
Messages
600
Location
a place surrounded by Europe
In a few months the new 5W LS should be available. As far as I know it sucks about 750 mA at 6.8 volts. Therefore we need a DC-DC converter (booster) for all 2/3/4 cells applications and also for the 5 cells if using NiMH. Btw, I do know that 2C/2D cells will not be a really good solution because of the high current they had to deliver. So we have to concentrate more on the 3/4/5 cells.

As we all know it should be a current regulated booster. So we do have a new challange here in the CPF. I'm wondering who will solve this problem?

Has anybody already thought about which chip could be used best? The ZLT+ (with the ZXSC300) in its current design can't be used. We need at least a ZLT++ with an improved design if this is possible (other output transistor etc.)

Most of the Maxim IC's (eg. MAX1674) can't be used either to drive this new LS as they have a max. output voltage of 5 to 5.5 volts (voltage regulators) and cannot deliver the needed high output power. As far as I have seen only the MAX1698 would almost match our requirements. The min supply voltage of 2.7 volts would be no hindrance for a 3/4/5 cell flashlight but maybe the max supply voltage of 6 volts (dependig on whether alkalines or NiMH's are used).

Maybe we should take a closer look on some battery charger IC's which will probably come closer to what we are searching for?
 
Great idea René,

with a voltage range of 2.7 to 6V this chip could be the ideal solution for 3-4 cells - no matter what chemistry IMO.

Klaus
 
You mean the two in parallel like MrAl once alluded to in the orignal ZLT forum? Might be time to read through those 36 pages again...
 
I just found two other Maxim IC's which could probably be used to drive the new 5W LS:

- MAX771
- MAX1771

They both are originally voltage controlled but there is a possibility to easy convert them into current-mode controlled boosters without any additional parts.
 
Seems that right now nobody else is thinking about this topic and therefore this thread is a kind of monologue
grin.gif
. Or has anyone else already designed a step up converter which could be used?

As far as I have seen a lot of CPF fellows are interested in this new 5W LS. According to the known specs (750mA at 6.8V) in most applications this 5W LS can't be driven directly so most of us do need such a booster to be able to use it.

Btw, the booster circuit needed for the 5W LS could also be used to drive two of the available 'old' LS in series or even for two parallel strings of two in series switched LS. Maybe this is a better motivation to post your ideas and suggestions here?
 
Hi René,

After reading your post, please , it's not going to get cold, infact I'm watching seriously especially on every charge booster topic ! Btw, this is my 1st post ! 🙂

Anyway, found something might interest you :

http://dbserv.maxim-ic.com/appnotes.cfm?appnote_number=940

and try grab Linear LT1307 data sheet (yes, it's LT1308's brother) , it got another example of converting a constant voltage charge booster to current source.

Again, just want you to know you're not alone, keep posting up the goodies ! 🙂
 
Hi vicbin

Welcome in the CPF
smile.gif
smile.gif
! And thanks a lot for your mental support.

The LT1307 is not really usable in this application because of its low load current. Its max. switch current is 1.25 A and therefore its max. output current is not more than about 600mA under best conditions. According to the datasheet (see the diagrams) the output current is even lower than these 600mA which seems to be realistic. So it must the little brother of the LT1308.

It's quite a bit tricky to find some good chips which could be used for the new 5W LS. I was searching for the mentioned ones quite a while. You have to study a lot of these technical specs and the diagrams. And I'm not really an expert in these things but try to learn as much as I can .....

Concerning the Maxim's app notes: Here are two links dealing with the same topic but are closer related to the chips I mentioned in my earlier posts:
http://dbserv.maxim-ic.com/appnotes.cfm?appnote_number=211
and
http://dbserv.maxim-ic.com/appnotes.cfm?appnote_number=1054

It seems that our CPF fellows do not really like to look into this Electronics Forum. Hope that soon someone writes here in his post: Moved to the LED Forum .....
grin.gif
 
Rene,

I have been looking. There are a few ways to go. The easiest path is to do a step down converter since there are many out there.

To step up there are many ICs that can do this, but, very few of them if next to none that are small enough to fit in the 14.5mm requirement of the mini mag.

Now, if you don't target the mini mag. Say the 2D Mag Light then the LM2621 will fit in the lamp area. It's pretty large cavity area.

But, the LM2621 is a voltage regulator. It would need additional circuitry to make it regulate current.
 
Hi Wayne

Welcome to our first expert
smile.gif
!

I think the 5W LS can't be used in the Minimag. I don't know anything about the heat it will produce but IMO it will be too much for the Minimag means one will need quite a big heatsink. So the size of the the booster circuit is not as important as for the standard LS.

Concerning the step down converters:
There is only one problem - most flashlights have two, three or four cells means the input voltage is too low for a step down converter (4 x 1.5 volts = 6 volts, the 5W LS 'needs' 6.8 volts). So a step down converter could be used only for a few of the bigger flashlights (eg. 5D and 6D Maglites).

Btw, I have added another step down regulator chip in your thread .....

I just decided to make a list of the chips which could be canditates for this booster circuit. The LM2621 is already included.


ZXSC300 (ZLT++ in a improved version? ZLT+ experts: Please help!)
MAX1698
MAX771
MAX1771
LT1308B
LM2621


IMO one very important criteria must be: The circuit should be as simple and cheep as possible!

Today I have ordered samples of the MAX1698 and the MAX1771. According to my earlier experience it will take about 2 weeks (or more) until I will get them. For me it would be to expensive to order samples of LT1308B because of the shipping costs from USA to Europe.

Who can add some more chips to this list? Who has better ideas? Who is able to rate these chips better than I can?
 
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by remuen:
Hi Wayne

Welcome to our first expert
smile.gif
!

I think the 5W LS can't be used in the Minimag. I don't know anything about the heat it will produce but IMO it will be too much for the Minimag means one will need quite a big heatsink. So the size of the the booster circuit is not as important as for the standard LS.

Concerning the step down converters:
There is only one problem - most flashlights have two, three or four cells means the input voltage is too low for a step down converter (4 x 1.5 volts = 6 volts, the 5W LS 'needs' 6.8 volts). So a step down converter could be used only for a few of the bigger flashlights (eg. 5D and 6D Maglites).

Btw, I have added another step down regulator chip in your thread .....

I just decided to make a list of the chips which could be canditates for this booster circuit. The LM2621 is already included.


ZXSC300 (ZLT++ in a improved version? ZLT+ experts: Please help!)
MAX1698
MAX771
MAX1771
LT1308B
LM2621


IMO one very important criteria must be: The circuit should be as simple and cheep as possible!

<HR></BLOCKQUOTE>

ZXSC300 - I think this one is maxed out. Even though this is a current mode it is not regulated current.

MAX1698 - the 300mV sense is better OK, but, the I SW max of 360mA limits the output to something less than 360mA. (Maybe 60% of that).

-------
MAX771
MAX1771
LT1308B
-------
I have not looked at these in detail. The 1308B I think is similar to the 1308 and can't drive the required current.

LM2621 - Voltage regulator. Prefer something like this but current mode.

I'm adding the LT1937. It's similar to the LT1308, but, with a lower sense voltage improving the efficiency. Good for next generation 2-N 5mm driver.

I think the mini mag has sufficient thermal mass to distribute the heat over a large area. I'm currently driving approximately 500mA at 3.2V or 1.6W in my mini mag and it gets warm on the upper 1/2 of the battery tube. I'm going to estimate that 5W will heat the tube up more and if I thermally bond the head to the tube increasing the heat dissipation that the overall increase will be minimal. Although at 5W the head and body tube both will get warm.

I have a test coming up hopefully this week or next where I will be overdriving a mini mag LS to 800mA. That's my target for now. If I get that going I will see what the heating does at 2.6W.
 
dat2zip,
Even if we were able to fit the circuit into a minimag, how do you think the batteries would hold up to the high drain? 6.8 V @ 750 mA seems a little excessive for 2 AA batteries, doesn't it? What would you estimate the runtime at those specs?
And what circuit are you using in your minimag to drive the LS at >500mA?

By the way, thanks for your help on the Zetex circuit, i finally got all the bugs worked out and so far my light is running great.
 
dat2zip

As far as I can remember all these switcher chips except the MAX1698 and the ZXSC300 should be able to supply the needed current. But I'm not really sure of this. I have to take a closer look into the datasheets and will come back on this (maybe in two or three days). In every case I will give the MAX1698 a try because it could be used for some other applications. It's a pity that its Vcc is to high for a single or two cell application. This would need some kind of bootstrapping.

Another point: How could we best simulate the new LS to make some tests with new circuits? Something like MrAl's LS simulator with 1N4xxx rectifiers? We then could fry eggs on the diodes ....
grin.gif
grin.gif


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

Even if we were able to fit the circuit into a minimag, how do you think the batteries would hold up to the high drain? 6.8 V @ 750 mA seems a little excessive for 2 AA batteries, doesn't it?
<HR></BLOCKQUOTE>

Thanks, InTheDark. This is what I have forgotten to mention in my former post. As already said in my opening post IMO the two cell flashlights will have some problems with the high current. I think two AA cells cannot deliver the required current of 1.5 to 2 amps or more needed to get an output current of 750mA.
 
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by TripleDouble:

Remuen, rest assured, this topic is not dead, I just think no one has made real progress that they feel they should be posting. We all have ideas to test, and try, but I'm not sure I want to sacrifice 2 LS's in series at this point.
<HR></BLOCKQUOTE>

Hi TripleDouble, thanks for joining in. At least my 'cry for help' has motivated some of you ....
smile.gif


I think there are only few CPF members here which do have a solution for this challange or do have the knowledge and knowhow to solve it (I really do not belong to them). I do hope they will all join us here. And all together could design the best possible solution (= simple, cheap, easy to build and top performance). As we all cannot invest all our time in this project it certainly needs some time to design such a circuit - except someone has already designed it and will share it with us.

Btw, it would be great if we eventually would have more than only one good step up converter! There are always pros and contras.

Concerning the two LS in series: There are at least a few posts here in the CPF where some of our fellows have discussed this or already have build such 2 LS lights.


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

2AA nimh's could deliver the needed current, and give you surefire style runtime.
<HR></BLOCKQUOTE>

Just to make sure that we talk about the same:
According to the simplified energy formula Pin = Pout / efficiency the input current needed must be more than twice as high as the output current. With two fresh AA alkalines and an estimated efficiency of 85% the input current needed for an output of 750mA at 6.8 volts is already 2 amps.

With NiMH the input current must even be higher than with alkalines because of the lower voltage (3.0 volts versus 2.4 volts for NiMH). Even worse: Using a regulated step up converter means that the lower the input voltage is the higher the input current must be. But frankly I didn't look into the specs of all these AA batteries/cells and so I cannot really say what current they can supply. It is more my feeling which says AA cells/batteries are not able to supply this current ...
grin.gif


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

I think we should not restrict this dialogue to 750 ma, as I'm sure there are probably some useful applications running at lower currents. I don't always need 120 lumens, I'd be happy with 60 lumens blasting out of a minimag body.
<HR></BLOCKQUOTE>

I do fully agree with you! Personally in most situations I prefer the longer runtime over the brightness. My ZLT+ is adjusted to 'only' 300mA. But there are at least two opposite opinions about this. The main point is that you can always underdrive a LS but you cannot drive it with a higher current if the converter is not able to supply it ....
 
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by remuen:
In a few months the new 5W LS should be available. As far as I know it sucks about 750 mA at 6.8 volts. Therefore we need a DC-DC converter (booster) for all 2/3/4 cells applications and also for the 5 cells if using NiMH. Btw, I do know that 2C/2D cells will not be a really good solution because of the high current they had to deliver. So we have to concentrate more on the 3/4/5 cells.

As we all know it should be a current regulated booster. So we do have a new challange here in the CPF. I'm wondering who will solve this problem?

Has anybody already thought about which chip could be used best? The ZLT+ (with the ZXSC300) in its current design can't be used. We need at least a ZLT++ with an improved design if this is possible (other output transistor etc.)

Most of the Maxim IC's (eg. MAX1674) can't be used either to drive this new LS as they have a max. output voltage of 5 to 5.5 volts (voltage regulators) and cannot deliver the needed high output power. As far as I have seen only the MAX1698 would almost match our requirements. The min supply voltage of 2.7 volts would be no hindrance for a 3/4/5 cell flashlight but maybe the max supply voltage of 6 volts (dependig on whether alkalines or NiMH's are used).

Maybe we should take a closer look on some battery charger IC's which will probably come closer to what we are searching for?<HR></BLOCKQUOTE>


Hello Rene,

Well, as for me, I'll be using 5 'C' cells, in a 3 'D' cell flashlight. Perhaps a little resistance. These batteries provide about 15 amp hours of current, and should drive the 5 watt very fine indeed, and get possibly 20 hours or 24 hours of full brightness?

The heat sinking characteristics of the aluminum flashlight should adequately handle the heat sink requirements.

Very very simple and effective, and cost efficient.

Wayne www.elektrolumens.com
grin.gif
 
Runtime calc would be current LS - parts / 4

So it would theoretically go from 1/2 hour or maybe less with Lambda´s overdriven Mad-Max LS - to 45min with an ZLT equivalent circuit to well over 1 hour with an ARC-LS style setup using 2AAs as a comparison.

But I agree that less power / less lumen @ longer runtime might be nice too - I guess latest with the 5W luxeons a dimmer will be introduced on some designs
grin.gif


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

.... Very very simple and effective, and cost efficient
<HR></BLOCKQUOTE>

Hi Wayne

As usually one of your excellent directly driven solutions!
smile.gif
Almost perfect!

Only almost perfect because some people do prefer NiMH or NiCd cells with only 1.2 volts per cell ....

Btw, what's about the LS? I didn't get your answer yet or I must have missed it in my flooded email inbox.


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

I guess latest with the 5W luxeons a dimmer will be introduced on some designs
<HR></BLOCKQUOTE>

A dimmer for you? In a 'Flab-Scheinwerfer'? Never heard about this before
grin.gif
grin.gif
grin.gif
grin.gif
grin.gif


But it is a good idea for all others. So we should take it on the wish list. What do you all here think about this?
 
Dimmer's are great, but add complexity to something we don't have yet.

Wayne, have you ever considering drilling a hole into your mag conversion heatsink, popping in a toggle switch from radioshack, and allowing people to toggle between 0 ohms and 2 ohms?

15 amps hours of current? What brand of c batteries are these?

I really, really wish these LS's could handle more/produce less heat, as I'm dying to slap one into a UK SL4 or SL6. I just might try with an underdriven one anyways.
 
Rene,

I took a look at this again and it has both constant current and constant voltage mode regulation and a 1.5A?? I SW limit. It could easily be re-wired as a straight boost circuit and simulated using LT Switch CAD software. I think this one has possibilities.

You posted this in my quest for my 4AA step down. That's fizzed out and I'm more focused on developing something for the new 5W model.

Maybe I'll do some simulations on this soon.


<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>dat2zip
Even you 'tend to dislike most Linear ICs' here another Linear chip with interesting features: The Linear LT1512. It is a SEPIC constant current/constant voltage regulator with a max. switch current of 1.5A. Would be interesting what efficiency on e would get with this IC?
--------------------

René

<HR></BLOCKQUOTE>

InTheDark,

<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR> And what circuit are you using in your minimag to drive the LS at >500mA? <HR></BLOCKQUOTE>

I'm must be a compulsive murderer at large. Gotta push the limit is my motto.

I've got my MAX1674 running full throttle on the LS. I'm still tweeking components and layout and expect to see another 100mA for the same setup. My mini mag gets warm but not too bad.

I'm thinking the 5W model will require thermally bonding the head as well to increase the heat sink area.

Yeah, Yeah... Pictures... I know. As soon as I get the light meter I ordered I'll take some photos and post my luxs...

Besides my mini mag is out on loan.
 
You must log in or register to reply here.

Latest posts

Top