Using a DC/DC converter to power the laser

[IgorT]

Here's a graphic example of what I was trying to explain.

I understood what you meant, the second time i read it.. Good writeup, BTW...

I also noticed this behaviour in my LD. The current was slowly climbing, when the LiPo was full. But i guess i have good cooling, since this change in current was very small. I didn't understand it untill i read your explanation tho..


So if you regulate the current, the output power will drop, as the LD heats up..

But if you feed it constant voltage, the output power can even go up, as it heats up? Or would it also drop and the power difference would only result in even more heat, possibly killing the LD?

 

[chimo]

I understood what you meant, the second time i read it.. Good writeup, BTW...[/qoute]

Thanks.

So if you regulate the current, the output power will drop, as the LD heats up..[/qoute]

Yep.  However, in some regulation schemes actually use the photodiode on the optical output to regulate the LD. These are more complex in that they must also take the LD drive limits into consideration and have safeguards built in.

But if you feed it constant voltage, the output power can even go up, as it heats up? Or would it also drop and the power difference would only result in even more heat, possibly killing the LD?

It's possible, but the two elements work against each other.  It will be difficult to predict which one will "win".

 

[IgorT]

It's possible, but the two elements work against each other.  It will be difficult to predict which one will "win".

Yeah, i thought so.. It's hard to guesstimate.

Thanks for the info!

 

[IgorT]

Ok, i've asked my electronics engineer to draw me a step up converter circuit, that would put out a regulated voltage over the most of the battery capacity, even when the battery voltage drops..

If i get something interesting, i'll post it.


Otherwise, MatajumotorS told me those DX converters have two boards. One contains the converter and the coil, the other contains a PIC which does all the stupid blinking.. He also told me they can be used separately, so getting rid of the blinking would not be a problem. I just need to find out what that IC is, and find the datasheet. It could be all that is needed.

 

[chimo]

IgorT, why not go directly to a boost (or buck or buck/boost) current regulator? It will be the most efficient.

If you are unfamiliar with the buck boost terminology:
Boost: source voltage is lower than the output voltage
Buck: source voltage is higher than the output voltage
Buck/Boost: source voltage can be either higher or or lower than the output voltage

Current regulated drivers or these types can be purchased from the Sandwich Shoppe (link below).  The set currents can usually be modified by swapping the sense resistor on the driver board.
http://theledguy.chainreactionweb.com/index.php?cPath=48_49_61

Paul

Ok, i've asked my electronics engineer to draw me a step up converter circuit, that would put out a regulated voltage over the most of the battery capacity, even when the battery voltage drops..

If i get something interesting, i'll post it.


Otherwise, MatajumotorS told me those DX converters have two boards. One contains the converter and the coil, the other contains a PIC which does all the stupid blinking.. He also told me they can be used separately, so getting rid of the blinking would not be a problem. I just need to find out what that IC is, and find the datasheet. It could be all that is needed.

 

[IgorT]

IgorT, why not go directly to a boost (or buck or buck/boost) current regulator? It will be the most efficient.

My electronics engineer should be capable of finding the most efficient solution.

I'm familiar with these converters, since i've used one before in one of my devices.. It's just that Gazoo made me a little unsure, when saying he has not seen a boost regulator that would provide a constant current.. That's why i asked my EE to draw me one, so Gazoo can finally say he saw it..

But more importantly, so i can try to make a laser powered by just one AA NiMh cell.




P.S. I took a look at the link you posted, and those seem interesting. They say minimum voltage 1.6V, so they would need two AA cells, but would completely drain them, before the current would start to drop.. Gazoo, take a look..

You have to understand tho, that since i'm in electronics developement and manufacturing, i prefer to build my own things to buying them for example. If i get something good, i'll post it so everyone can make if they wish. But it might be easyer to go with the DX ones..

 

[chimo]

It's just that Gazoo made me a little unsure, when saying he has not seen a boost regulator that would provide a constant current.. That's why i asked my EE to draw me one, so Gazoo can finally say he saw it..

But more importantly, so i can try to make a laser powered by just one AA NiMh cell.

You have to understand tho, that since i'm in electronics developement and manufacturing, i prefer to build my own things to buying them for example.

The only ones on that page that are not current regulated are the Madmax series. It's not too complicated to make a switching voltage regulator into a current regulator (just sometimes tricky to make it stable). You just have to convert the output current to a voltage for the switcher/controller chip's feedback pin. That can be done a number of ways.

I hear you on your last sentence - it's not always the destination but sometimes the journey that's more rewarding. You may want to check the site below for loads of info:

http://www.molalla.net/~leeper/led.htm

 

[IgorT]

I hear you on your last sentence - it's not always the destination but sometimes the journey that's more rewarding.

Exactly. BTW.. Do you work with electronics professionally? You really seem to know a lot about it.

Thanks for the link!

 

[Gazoo]

Yes, I have looked at the ones from the sandwich shop before. I don't want to pay $18.00 for one. And when I said I had not seen any that would maintain current regulation I was referring to the DX ones, and those were the ones tested by Eprom. Sorry for the confusion. I know of another one that is available that will maintain current but it is a multimode and costs around $40.00...but I can't remember the link for it.

I have had no personal experience with boost or boost/buck converters. So I can only go by what others have said about them. The DX solution might be ok, but I will be looking forward to what your EE says. Thank for checking with him..

 

[chimo]

[quote author=chimo link=1194375801/15#22 date=1194537999]I hear you on your last sentence - it's not always the destination but sometimes the journey that's more rewarding.

Exactly. BTW.. Do you work with electronics professionally? You really seem to know a lot about it.

Thanks for the link![/quote]

I'm an EE but don't get to dabble in it for my work - just for fun.

 

[MatajumotorS]

I wonder what it would take to disable the unwanted modes??

Regards rog8811

Yes You can!   remove the controlling mosfet and solder drain and source together!   Then You will have no modes.


I am waiting for this: www dot dealextreme dot com/details.dx/sku.7882 board, it is capable putting out about 800ma (with this: www dot dealextreme dot com/details.dx/sku.7302, with fresh batt i get 240mA 3,06V on my LD) for constant current i am using AMC7135, as you can see it is not working with first board ( only 240ma), but IF there wil be some spikes it will limit it to 350mA..

About an effeciency, with second board you can set an output voltage. I you mesure the voltage on the LD at desired current ( in my case 350mA or about), you can cet the output of converter only 0,1 - 0,2V higer to allow the LDO AMC7135 to work, and that is why effeciency will no suffer mutch.
As EpRoM wrote:
''1 NiMh Test on High - Lowered the boards Output to ~1,2W
On Battery
1,22 V * 1,38 A = 1,684 W

On Emiter
3,15V * 360 ma = 1,134 W

Efficiency = 1,134 / 1,684 = 67% "

I think it is acceptable, and i think it is lower than using LM317. Tell me, if i am wrong.


p.s.  ''i do not have premission to post LINKS in this thread'' ..

 

[MatajumotorS]

Why do You guy's care about filtering current to the LD? I think You need only to limit the max current flowing trought the LD. Imagine - what does it like - waweforms, when a DVD disc is recorded? :

 

[chimo]

You need only to limit the max current flowing trought the LD.

That's exactly what we are concerned about. The regulator will limit the steady-state current, but it is often the transients that will kill the LD.

 

[IgorT]

And when I said I had not seen any that would maintain current regulation I was referring to the DX ones, and those were the ones tested by Eprom.  Sorry for the confusion.

In that case i misunderstood you.. I apologize.

I agree with you about the price of those from the sandwich shop.. For the price of two, i could most likely order a bunch of coils and step up ICs and make my own..

But the price of the DX ones seems quite good to me.. Can't order anything else right now tho.. It'll have to wait a week or two.

 

[IgorT]

About an effeciency, with second board you can set an output voltage. I you mesure the voltage on the LD at desired current ( in my case 350mA or about), you can cet the output of converter only 0,1 - 0,2V higer to allow the LDO AMC7135 to work, and that is why effeciency will no suffer mutch.

That would be the best posibility.. To set it only slightly higher than the minimum. This way less of the excess power would have to be converted to heat. It's the best option for this combination.

The only thing that would be better for battery efficiency would be to make a current regulated step up converter. But it's probably overkill. The boards from DX are not that expensive.


Did you test the lowest voltage at which the first one stops putting out 5V?


BTW: Nice of you to show up.. You'll be able to post links after a few more posts.. (10 post minimum i think)

 

[IgorT]

Efficiency = 1,134 / 1,684 = 67% "

I think it is acceptable, and i think it is lower than using LM317. Tell me, if i am wrong.

The LM317 has an efficiency of 30 - 40%. So this is better.


But i'm not really all that obsessed with efficiency.. I just want to make it small and working.. It's gonna be rechargable anyway..