What driver can I use to supply 3A?

[Andratos95]

Hello,
I want to drive a NUBM08 laser diode at 3A and make sure that it doesn't get destroyed. Originally, my plan was to do this with a linear regulator using the LM338, but then I have had many people tell me that it is quite stupid since using it at 3A would dissipate a lot of power as heat on the feedback resistor.

It was then suggested to me to use a "switching regulator" as a driver. Now, from my understanding this means that I could use a buck converter in some configuration that allows me to choose the current limit to 3A (but maybe this is completely wrong?). I am not interested at all on keeping the laser driver small, I would actually like to create the driver myself on a perfboard, but I don't know exactly what would be a good circuit for this task.

I have seen on AliExpress that they sell these cheap buck converters with a potentiometer to adjust the current limit (see image). Can I use these as a laser driver? If not, why is that?

Could you suggest a good laser driver that can provide a constant 3A which I can buy on AliExpress under 10$? (It doesn't matter if it's large).


Thank you!

 

[Dusty_Lenses]

Be careful with those buck regulators from china, some of them throw massive spikes under certain conditions. Cant remember which one, but I do believe it was the lm2596 based that were doing it, and it was during power up even after a safe current was set.

Some experimentation with a dummy load will be a must.

 

[RedCowboy]

Send DTR a message and tell him you want it set for 3.0A >

https://www.ebay.com/itm/171690251341?hash=item27f989544d:g:tHcAAOSwzhVWsTPZ

or

Astral SuperDrive V - 5A Adjustable Buck / Constant Current Driver w/ 1 kHz PWM | eBay

Astral SuperDrive V. PWM Frequency: 1kHz. Mode: Buck Driver (step-down). Ripple: Depends on ratio of V-out / V-in. At 5A and 4.5Vf output, ripple is about 1-2%. V-out (V): 2.5 - 13V (depends on current output and laser diode’s Vf).

www.ebay.com

NOTE : Either bucking driver must be heat sinked.

If you only want to spend 10 dollars and don't care about size there's this > https://www.ebay.com/itm/261969315148?hash=item3cfe96854c:g:PBEAAOSwsc1bHyLe

or this

3A Step-Down Constant Circuit Laser Diode Driver/For 1.6W-3.5W Blue Laser Diode | eBay

Find many great new & used options and get the best deals for 3A Step-Down Constant Circuit Laser Diode Driver/For 1.6W-3.5W Blue Laser Diode at the best online prices at eBay! Free shipping for many products!

www.ebay.com

 

[Andratos95]

Send DTR a message and tell him you want it set for 3.0A >



or


NOTE : Either bucking driver must be heat sinked.

If you only want to spend 10 dollars and don't care about size there's this >

or this

First of all, you're awesome, thank you for the help. Second of all... Could you give me a short explanation to why the chinese buck converter from AliExpress I showed would be bad? I mean, what do the ones you posted have that make them be almost 10x more expensive? Also, the last one you posted on ebay says only up to 3.5W so I suppose that I couldn't drive a powerful diode such as the NUBM08 or Sharp gh04c05b9g which are around 5W, right? (Or does that wattage refer to something else?). Again thank you so much for helping me

 

[RedCowboy]

Laser diodes are sensitive to spikes, so drivers designed for lasers have protections built in that the general purpose dc to dc converters don't.

Here's one good for 1-5A you can drive your nubm08 @ 4.5A as long as you put it ( your nubm08 ) in a heat sink.

NEW 5A Laser Diode Constant Current Driver - 1-5W Engraving - TTL - 12V - Fan | eBay

Find many great new & used options and get the best deals for NEW 5A Laser Diode Constant Current Driver - 1-5W Engraving - TTL - 12V - Fan at the best online prices at eBay! Free shipping for many products!

www.ebay.com

 

[Giannis_TDM]

First of all, you're awesome, thank you for the help. Second of all... Could you give me a short explanation to why the chinese buck converter from AliExpress I showed would be bad? I mean, what do the ones you posted have that make them be almost 10x more expensive? Also, the last one you posted on ebay says only up to 3.5W so I suppose that I couldn't drive a powerful diode such as the NUBM08 or Sharp gh04c05b9g which are around 5W, right? (Or does that wattage refer to something else?). Again thank you so much for helping me

2 things, I have a dedicated driver that can do 3A, you can find it on my website along with all of its features: https://sites.google.com/view/giannislasershack/laserled-drivers datasheet is available.

And an explanation: The actual topology that the aliexpress switch mode supply uses is not fundamentally wrong, rather the component choice and cost-cutting measures are bringing it down from its full potential. What do I mean by that? Well, it is a normal CV (constant voltage) IC that has an additional circuit with it to add a constant current feature(here is a video on how it works and how you can DIY such a circuit.

) Now there is nothing wrong in doing that and in fact, that's what most hobbyist grade CC drivers do but the trick is that it requires very careful component selection and math else it is going to be very unstable and have obscenely high current ripple and voltage ripple(voltage ripple = voltage change in millivolts or volts over a period, same goes for current ripple)

Now that the area where chinesium falls short, usually the chiniseum switch-mode power supplies have very high ripple, I have tested some to be as high as 1v p-p (very bad for diodes, you at minimum want sub 100mv p-p to ensure diode longevity and 'happiness') and that's down to the cost-cutting that enables them to cost only dollars a pop. Also, dedicated drivers will have a feature called soft start where they raise the current up to their set instead of starting at their set current, that soft ramp negates current overshoot.

PS, Regarding the question you asked, Yes and no, W is a measure of optical power(when regarding lasers), rating a driver by that is plain wrong, drivers are and should be rated by their maximum current output, diodes too if you look at the datasheets the manufacturer gives you a maximum operating current.

 

[Andratos95]

Laser diodes are sensitive to spikes, so drivers designed for lasers have protections built in that the general purpose dc to dc converters don't.

Here's one good for 1-5A you can drive your nubm08 @ 4.5A as long as you put it ( your nubm08 ) in a heat sink.

Thanks! I read that linear regulators have much less ripple than buck converters, but the trade-off it's that they're less efficient. That driver you sent seems nice! The only "problem" is that it costs much more than it would if I could build it myself, but I definitely had overestimated the complexity of these types of drivers

 

[Andratos95]

2 things, I have a dedicated driver that can do 3A, you can find it on my website along with all of its features: https://sites.google.com/view/giannislasershack/laserled-drivers datasheet is available.

And an explanation: The actual topology that the aliexpress switch mode supply uses is not fundamentally wrong, rather the component choice and cost-cutting measures are bringing it down from its full potential. What do I mean by that? Well, it is a normal CV (constant voltage) IC that has an additional circuit with it to add a constant current feature(here is a video on how it works and how you can DIY such a circuit.

) Now there is nothing wrong in doing that and in fact, that's what most hobbyist grade CC drivers do but the trick is that it requires very careful component selection and math else it is going to be very unstable and have obscenely high current ripple and voltage ripple(voltage ripple = voltage change in millivolts or volts over a period, same goes for current ripple)

Now that the area where chinesium falls short, usually the chiniseum switch-mode power supplies have very high ripple, I have tested some to be as high as 1v p-p (very bad for diodes, you at minimum want sub 100mv p-p to ensure diode longevity and 'happiness') and that's down to the cost-cutting that enables them to cost only dollars a pop. Also, dedicated drivers will have a feature called soft start where they raise the current up to their set instead of starting at their set current, that soft ramp negates current overshoot.

PS, Regarding the question you asked, Yes and no, W is a measure of optical power(when regarding lasers), rating a driver by that is plain wrong, drivers are and should be rated by their maximum current output, diodes too if you look at the datasheets the manufacturer gives you a maximum operating current.

Thank you so much for this explanation, it really helps! If space is not of any concern, would adding a bunch of capacitors in parallel on the output of of one of those cheap chinese drivers be of any help to reduce ripple?

 

[Giannis_TDM]

Thank you so much for this explanation, it really helps! If space is not of any concern, would adding a bunch of capacitors in parallel on the output of of one of those cheap chinese drivers be of any help to reduce ripple?

Not really and it is not really voltage ripple being the problem here (although it is a factor) Rather, the current ripple is the problem and that cant be mitigated by capacitors.

 

[Giannis_TDM]

Thanks! I read that linear regulators have much less ripple than buck converters, but the trade-off it's that they're less efficient. That driver you sent seems nice! The only "problem" is that it costs much more than it would if I could build it myself, but I definitely had overestimated the complexity of these types of drivers

Not really, I price my drivers about 5$ less than the market price while offering a lot more in testing. Each driver comes with a standard test report and if you were to build one yourself it would be way more expensive. 5 PCBs (the minimum) at the specs I have chosen with the cheapest shipping from jlcpcb will cost you 45$ + ~8$ In components + shipping for them. If you don't mass order then they will be way more expensive than my already low prices(to clarify that's for 1 driver).
And it depends really, good buck regulators have nearly the same ripple as linear ones, for instance my 3A buck when set at low currents (around sub 500mA) will have very comparable ripple to a linear reg (sub 6mv p-p)

 

[Andratos95]

Not really, I price my drivers about 5$ less than the market price while offering a lot more in testing. Each driver comes with a standard test report and if you were to build one yourself it would be way more expensive. 5 PCBs (the minimum) at the specs I have chosen with the cheapest shipping from jlcpcb will cost you 45$ + ~8$ In components + shipping for them. If you don't mass order then they will be way more expensive than my already low prices.
And it depends really, good buck regulators have nearly the same ripple as linear ones, for instance my 3A buck when set at low currents (around sub 500mA) will have very comparable ripple to a linear reg (sub 6mv p-p)

Oh I see... But if I include shipping to Europe I imagine it will cost quite a lot

 

[Giannis_TDM]

Oh I see... But if I include shipping to Europe I imagine it will cost quite a lot

Not really, I am in Greece so no customs and shipping is 5.5 euros with tracking.

 

[Andratos95]

Not really, I am in Greece so no customs and shipping is 5.5 euros with tracking.

Oh!!! Well then that changes things... But one thing: would using your driver constantly at 3A damage it in the long run?

 

[RedCowboy]

How about an LT3083 ? Anyone ever used one to make an adjustable linear ?

Here's a video that may help you build your own adjustable linear if that's what you want.

 

[thanhtung]

here bro only 5-6$ run well this link aliexpress
I suggest when you run with 2 x 18650 high current charge it can up more than 2.4A! It its good for life-time of your diode! if you choice exactly current 3A , sometimes it will be jumped to 3.4A or 3.5A for your NUBM08

 

[Giannis_TDM]

Oh!!! Well then that changes things... But one thing: would using your driver constantly at 3A damage it in the long run?

Oh easily, if heatsinked, If not the recommended duty cycle is 30s on 1min off.