The real high power and full range DIY laser diode driver *new version 3*

[mojo_1234]

:thinking: Update: new more compact layout version 3 added

Hi folks,

in the past months there'd been tons of threads about high power constant current drivers for laser diodes (especially for our new friends the 445nm diodes).

More than a year ago I've developed a precise high power driver for 808nm laser diodes (capable to drive loads up to 8 amps). When I got my first 445nm diode I've tested the driver and it works fine with them. But the original 808nm schematic contains a lot of "unnecessary" protection and precision stuff you don't need for 445nm.

So, I've decided to build a simpler version for the 445nm diodes. Personally I don't like this LM317 based stuff (limited current range, expensive powerful pots required, too much energy is wasted into heat). And DIY switched current drivers are hard to handle for beginners.

Just take a look at the schematics I've attached. With the given resistor values the driver is capable to provide a constant current starting from *zero* 0mA up to 1250mA. With a lower value of the shunt resistor the driver can handle more than 8 amps at all!

HAVE FUN GUYS!!!

Questions? Remarks? Any interests in DIY kits? It's very easy for me to get all the required parts...

Yours,

Mojo

--------------------------------------------------------------------------
:thinking: Update: new more compact layout version 3 added

(more pictures see last reply)




--------------------------------------------------------------------------
TECHNICAL STUFF:
Supply voltage:
min. 6.5V for maximum current with R7 set to 1 Ohm
(or 7V with optional protection shottky diode)

Output current range:
0mA* - 1250mA with shunt resistor R7 set to 1 Ohm
0mA* - 2500mA with shunt resistor R7 set to 0.5 Ohm
* minor leakage of ~1mA may be possible


REMARKS:
The current is set with the pot. Before connecting the driver to the supply voltage always set the pot to *zero*. This is GND (pin 1=2 of pot).

Before connecting your expensive diode to the driver ALWAYS test with a dummy load first. A dummy load for laser diodes just can be build very very easy. Just put 5 or 6 standard diodes in series, connect them with the driver, power up and measure the current.

(+) -->|-->|-->|-->|-->|-- (-)

To measure the current just connect a voltmeter in parallel to R7. With R7 set to 1 Ohm each mV measured corresponds to exactly 1mA current. You can also connect a current meter in series to the voltage supply. Just subtract the current required for the driver itself.

The optional schottky diode protects the driver and *your laser diode* in case of the supply voltage connected with wrong polarity. But you'll also lose about 0.5V which have to be added to the supply voltage.

Add good heat sink for the mosfet (Q2, IRF520) is required. In my second prototype I just used the massive aluminum base plate of my laser build. The shunt resistor will also be quite hot during operation!

If you can't get the IRF520 you can also use a similar type (must be of type N).

All the capacitors are not required but without them it might be possible that some current spikes may occur (and your diode doesn't like spikes at all). Personally, I would never remove C2.

The opamp used for the driver is dual one (two similar opamps in one package). To avoid any unwanted oscillations just connect the remaining pins as shown in the schematic.

R8 is not required. But it will ensure that C4 is discharged when connecting a laser diode.

Most of the voltage is lost in R7 (1V at 1A current). If you choose a smaller resistor value less voltage is lost at same current (e.g. 0.5 Ohm -> 0.5V at 1A). The required supply voltage can be reduced then and less power is wasted into heat! ALWAYS TEST WITH A DUMMY LOAD!

Be aware of the fact that the diode is directly connected to the positive voltage supply and not grounded! But as far as I know the cases of the 445nm diodes are isolated from the diode itself.


PART LIST:
R1 - 1k Ohm (limit current for zener diode)
R2 - 15k Ohm
R3 - 10k Ohm
R4 - 1k Ohm
R5 - 100 Ohm
R6 - 1k Ohm
R7 - 0.56 Ohm / 1W minimum!
R8 - 10k Ohm

C1 - 10uF (optional)
C2 - 1nF (recommended)
C3 - 47pF (optional)
C4 - 100nF (optional)

D1 - Zener Diode 2.7V / 500mW
DS1 - Schottky Diode SB320 (or any other standard type with 2A or more)

Q2 - Power Mosfet N-Type IRF520N (or similar type)

IC1A/2B - Opamp TS912 (rail-to-rail)

To build your own driver just use a drilled copper breadboard and a lot of solder...

 

[Morgan]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

What's the current draw for this driver at 1.2A output? It looks a bit bulky and labby-fied but can it be run in a handheld?

Thanks for this.

M

 

[mojo_1234]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

Hi morgan,
for sure... if you build the driver based on SMD parts there's no problem to reduce the overall size.

With 6V of supply and a 0.1Ohm shunt resistor there's no problem to drive your handheld.

Do you have some more specs of your handheld build?

mojo

 

[aricha]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

hi

can you run 4 diodes with 1 driver?

 

[CrazyMonkey]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

no, the op-amp he chose has a max of 16v.
If you used his schematic with the correct parts, I think so.

 

[Morgan]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

Hi morgan,
for sure... if you build the driver based on SMD parts there's no problem to reduce the overall size.

With 6V of supply and a 0.1Ohm shunt resistor there's no problem to drive your handheld.

Do you have some more specs of your handheld build?

mojo

I didn't have any personal specs to speak of but just wanted to pose the question. I have no objections to an SMD approach, (in fact, only subjections! )

I am no electronics genius but it would be interesting to see how small this could go in it's current form? It's somewhat of a shame its not a boost driver. Would it be a major issue to get it to this? Stop me if I'm being a touch ambitious but size is everything after all! What about a digital pot for microcontroller interface?

M

 

[rev0]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

It can go pretty small. Just take a look at any DX or similar green module

The DX module has a very similar schematic to this driver, it uses a pretty wimpy little SOT-23 series-pass transistor though.
http://laserpointerforums.com/f42/making-ttl-driver-dx-200mw-greenie-51916.html#post720330

I've never used a digital pot before, would be an interesting way to do analog modulation though. I was thinking if you had a microcontroller with a DAC built in, or got an external DAC, you could feed the DAC voltage directly into the positive input of the Opamp to set the current.

 

[danilow]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

for this 445 diodes how much mA and V can run without kill them ?¿?¿

 

[CrazyMonkey]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

for this 445 diodes how much mA and V can run without kill them ?¿?¿

That's a hard question to answer. 5A will kill it..
1.5A may give long life, 2A will slowly degrade over the course of hours(maybe 10's).

voltage is harder to answer, but you don't run them with fixed voltage, only constant current. Max amps should never be above ~4.8ish volts, some diodes may run at a lower voltage for the same current as others.

 

[mojo_1234]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

Yesterday I've build a much smaller version of the driver. The new layout is using the diodes host as the heat sink. Overall size is about 3.5cm x 2cm. But with some tricks (place parts on both sides of the board) it can even go smaller.

Pictures comming soon!!!

Please consider the driver I've mentioned was not designed to fit into a Aixiz housing (or similar). One watt of blue produces to much heat - so you will need a bigger housing to run your laser more than a few seconds.

My own build (see images at the beginning) runs a 950mA for more than TWO HOURS without any problems. So, I was able to make shots of the beam in about 500 meters distance... will add some pics soon.

@aricha
With some modifications the driver is capable to drive four diodes or more. But as already mentioned by CrazyMonkey - the opamp can handle up to 16V - not more!

For 4 diodes you'll require at least 20V of supply (each diode requires about 4.5V). If a add a simple voltage regulator you can reduce the voltage to around 9V to supply the opamp and reference voltage circuit. Shall I post an example circuit sheet for it?

 

[Firefly]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

i'm impatient to see your pics!!!

 

[HIMNL9]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

Good application of the old op-amp / mosfet current sink circuit

Just as suggestions, as voltage reference, you can use a 431A, that gives you a 2,5V constant voltage reference, virtually insensitive from temperature and fluctuations ..... also, for this current, i suggest you to use a 0,22 ohm 1W min sense resistor, cause at over 1A, the power dissipated from it, will be very high, if you use an 1ohm one.

About the op-amps, the LF354 may be better, for stability, and cause it have a 32V max voltage, but always as suggestion, i say you to keep the input voltage as low as you can (7 / 7,5 V or similar), cause, remember, all the part that is not used from the load, must be dissipated from the mosfet as heat, and more input voltage you give it, more power it need to dissipate ..... about this, any one of the "new" 60 / 80 A low voltage mosfets (40V), can be good, having RdsON around 9 or 10 milliohm .....

Just suggestions, anyway :beer:

BTW, a warning for those not used to current sinks drivers ..... the ground of the power supply and the ground of the load are NOT in common, as almost all the types of other drivers have ..... NEVER let the load ground go in contact with any other ground wires, or your diode smokes out more fast of that what you can notice

 

[mojo_1234]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

Thanks for your suggestions.

I'm already using a 0.1 Ohm resistor in my new build but it also requires a lower voltage reference (1 Ohm <> 1mA/mV, 0.1 Ohm <> 10mA/mV). So, the driver is more "sensitive" to noise or "swetty" fingers . The usage (as mentioned by HIMNL9) of a more precise reference voltage is a must (use a 431A or LM336-2.5).

The overall supply voltage (with protection diode) can be reduced to ~ 5.5V at all. In case of 1000mA and 4.5V for the diode only 1 Watt of heat is produced!!! That's more or less nothing...

Doing this with a standard linear regulator like the LM317 you will produce much more heat (min. 2.5V for the LM317 and 1.25V for shunt -> 3.75V * 1A = 3.75 Watts).

To loose less than 1 Watt is only possible with a buck converter (switched regulator). But again building your own buck converter is very very difficult even there are a lot of integrated circuits available on market.

I'm also using a PIC based controller with a 12bit DA converter which can easily connected to the driver. It only requires a simple voltage ...

 

[Meatball]

Re: The real high power and full range DIY laser diode driver *with 445nm pics*

Just a suggestion: If you display this circuit on a clean looking board, and can prove the circuits abilities and specs from top to bottom, people will be asking you to make these for them. You COULD end up selling these here as a kit, or completed circuit. If you would consider this idea, I would suggest sending a completed circuit to a VET so they can make a review on it.

Less risk is involved when selling laser drivers, rather than lasers!

 

[mojo_1234]

Re: The real high power and full range DIY laser diode driver *updated!*

So, here's the latest version of the driver layout. It's much smaller than the original one. If a smaller shunt is used (or mounted on the backside) and the fet is mounted on the chassis... 3.5x2.5 cm or less is possible.

With the current design I've tested it with 6V of supply and a 5A !!!! dummy load... is getting hot... but works for hours !!!


If you want to know more details just contact me PM...

mojo





some specs


the final comparision to the original design



And there's still enough empty space which wants to be used :friend: ...

 

[Solonar]

Re: The real high power and full range DIY laser diode driver *updated!*

Attempted to do a board layout in eagle, talk about a pita.

4 hrs later all I could manage was a single sided board with 1 jumper wire for a trace I could not manage to route.