Idea for driver/host

[Grix]

Ok, this might sound extremely stupid to experienced builders, but I'm just throwing this out there.

So the XJ-A1*0 diodes have proved to survive at currents like 1A, which is similar to the currents LED's from flashlights uses. So do you think it would be possible to use the drivers from flashlights to power these diodes?

For example buy this kit: 18650 build kit , and then skip the part where you strip the board and install a flexdrive, but connect the supplied board directly to the diode instead? (Maybe with a simple resistance in series if the current is too high). There are plenty of these drives at DX for very cheap: DealExtreme: $2.36 2.7V~6V 3W Cree Circuit Board for Flashlights (15mm*2.7mm)

 

[ReNNo]

If this is 3xAMC7135 then current should be about 1A , maybe too high for these diodes without TEC.

Yes, AMC should give stable current to 445nm diode with 2xLi-ion (18350 or 18650).
I use AMC in my Red and IR build and everything works just fine.

2xAMC7135 chips should give about 700mA current to diode and looks even better solution than Lm1117, because of its size and great PCB that can be used as battery contact.

 

[Grix]

So these would work well?
DealExtreme: $20.18 Quality AMC7135 700mA Regulated Circuit Board for DIY Flashlights 20-Pack

Can you confirm that just placing these in a standard flashlight would work to power the diode? If so, that would be very cheap and easy!

 

[HIMNL9]

Don't forgot to add some capacitors ..... they don't care too much with LEDs, that are almost 100 times more resistent than LDs, about spikes and so on ..... but for use them with LDs, imho, is better be more cautious and add also them .....

 

[Krutz]

i agree, those amc7135 chips are amazing! used two in parallel for a greenie, with no extra components at all. if you have the space (and those chips are tiny!) go for it, add a capacitor. there is a note on that on the amc7135 datasheet.

each chip gives 350mA, I found them to give me around 330mA though.

no, you cant really do other currents than 350, 700, 150mA etc. A resistor in series does NOT reduce the current, the current is the same for every component in the line, its just the voltage that changes, which doesnt help us here. you could theoretically put a resistor parallel to the laserdiode. this is highly dangerous (for the diode), since the diode changes its "voltage current ratio" all the time. that means that very little changes in voltage have drastic changes in current, which is why we regulate the current, after all.
in short: it could be possible to change the current of AMC7135 chips, but you should drive them very conservatively then! not that impossible that with rising temperature your diode (which lowers its resistance) gets more and more current, up to almost the full current, as if the resistor wasnt even there..

finally: the AMC is a common-positive connected driver. that means that you will short your battery out if your diode has internally connected the negative wire with case. thats not the case with the "floating" 445 diodes.
and, big extra, you can drive IR and 635nm diodes with that chip, even though those diodes have the case connected to positive!

-you cant use the AMC with red diodes
-you cant use the flexdrive with IR or 635nm diodes

oh, those AMC ones have a voltagedrop of just 0.2v!

repeating myself: i just love them! AMC7135 chips are a gift for drivers just as those 445nm diodes are for projectors. both have downsides, but besides that, they cant be beaten!

manuel

 

[ReNNo]

You can use AMC for red diodes but diode must be electrically isolated. It can be done with thermal tape without problems.

Only problem would be dissipated heat.
If 445nm diode has 4.5V FV on 700mA and batteries are 8.4V when fully charged then AMC chips need to dissipate 2.73W of heat.
Chips need to be heatsinked.

 

[Krutz]

both true, forgot about those!
isolating: I was thinking to put two strips of tape at the ends of a module or heatsink, to keep distance to the host. then fill that tiny gap with epoxy or thermal paste.

true again, the AMC is a linear regulator, burning the excess "voltage" to heat. the battery will supply exactly the current you need for your diode (350mA, 700mA etc), the AMC reduces the voltage to the diode.
from what I heard, they are really tough. someone said that one of his AMC chips got so hot it melted the solder contacts, the chip fell right off! ..and worked again, when resoldered back! ..would have expected them to have a 105°C switchoff (did I read this in the datasheet?), but anyway..

thanks for hinting, renno!

manuel

 

[Grix]

Hmm..

It says it cant take Vin of 8 volts anyways, one 3,6V battery is recommended.

What about this driver (one is cheaper than a 20-pack)? DealExtreme: $2.36 2.7V~6V 3W Cree Circuit Board for Flashlights (15mm*2.7mm)
Is it possible to just desolder one of the chips to make it 700mA?

From a review:

It doesnt heat up with extended use at all which means that it wont need to be mounted to a heatsink

 

[ReNNo]

both true, forgot about those!
isolating: I was thinking to put two strips of tape at the ends of a module or heatsink, to keep distance to the host. then fill that tiny gap with epoxy or thermal paste.

I use very thin thermal tape (not standard thermal tape) that is wrapped around whole module.
Thermal contact is very good because there is minimal distance between module and heatsink. This thin thermal tape works just like thermal grease but just better for electrical isolation.
There is no need to use thermal epoxy. Just put small blob of epoxy at the end of heatisnk.
Works great for me.

 

[ReNNo]

Hmm..


What about this driver (one is cheaper than a 20-pack)? DealExtreme: $2.36 2.7V~6V 3W Cree Circuit Board for Flashlights (15mm*2.7mm)
Is it possible to just desolder one of the chips to make it 700mA?

From a review:

If it's AMC7135 then you can desolder one chip for less current.

Datasheet says that AMC can take maximum of 7V at input, but I was sure that it worked with 2x18650 and IR diode (1.4A) for few seconds... didn't tested more.

 

[SpyderFire]

interesting idea . I have some of these

DealExtreme: $16.58 AMC7135 1050mA Regulated Circuit Board for DIY Flashlights 10-Pack

I will try one out with a diode. They work off of one Li-ion batt.

 

[ReNNo]

First we need to test AMC7135 how does it work with 2xLi-ion and silicon diodes (4.5V FV).
I don't have any at the moment, but I'll try to find one and test it.

 

[Krutz]

running these with one li-ion is right at the border.. the AMC7135 drops 0.2 volts, the diode needs more than 4 volts (i think!)

so you wont have the full runtime, at least. once the battery-voltage is too low, the diode wont draw any (useful) current at all, even though there is still much energy left in the battery.. sure, li-ion recharge to over 4 volts! gibe it a try!

manuel

 

[ReNNo]

One Li-Ion is FAIL.

At 700mA forward voltage of 445nm LD is more than 4.2V, and it will not work as it should be.

That is why I'm going to test AMC7135 with 2xLi-Ion...

Here is one more idea...
I can also put some silicon diodes in series between batteries and host. (In tailcap) That can lower voltage on AMC.

It can be done, I know, but we should test it to be sure.

 

[SpyderFire]

You were right, it did take 2 li-ion batteries to get 1W of 445nm out of a amc7135 @ 1.050A. it's definitely bright. With a aixiz regular glass lens im getting ~800mw and with a 405g-1 im getting 995-998mw

 

[ReNNo]

That is very good result.
Is AMC overheating too much?