PCB designs for laser power regulation?

I've gotten my 3W 445nm diode laser, (Nichia NDB7875) putting together the heatsink block, and did the "proof of life" test using a bench power supply.

Now I'd like to put together a current-controlled power supply that I can control from an Arduino or (preferably) from RAMPS 1.4 electronics. It seems I should be able to use one of the analog pins from the Arduino Mega (if any are available) as resistance for the LM317. There are many articles like this one [www.instructables.com] that describe the basics of building a current-regulated diode laser driver. I really want to be able to control the power output dynamically (in software) depending on the material.

I'm wondering if anyone has a preference for a design that can be controlled from modified Marlin firmware on RAMPS 1.4?

... I'm driving my laserdiodes with currents of up to 9Amps with Arduinos and a board with 5 parallelized LM's (3x LM338, 2x LM317) an switching the power to the LM's with a MOSFET, that is powered from the Arduino-pin over an opto-coupler -- the power modulating is done with PWM ...

Here are some basics: [www.reprap.org]

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Some my old experiments on this thread:
[forums.reprap.org]
I used heatbed connection, little bit modified Marlin for that and special G-code sequences.

Ah, thanks, I probably should have searched the wiki first...

That sounds very much like what I want to do but, being quite a noob with electronics I am still trying to figure out a couple of things. Using the LM317 / LM338 parts seems fairly straightforward and the datasheets provide some examples of circuits that are fairly close to what I want. It seems that basically I need to calculate the resistance needed. It looks like you're using ceramic capacitors and quite a bit of heatsink on the LM parts.

@tertzoid: I thought about using the heatbed supply but the input voltage range on my 445nm diode laser is 3.7V - 5.5V; Thanks for the link, I'll read through the thread more carefully; this is all helpful for me...

... you'll need around 1.6Amps and two different coolers that can handle 5 Watts at the diode and at the LM.

Best would be a LM338 and a sense-resistor of 0.8 Ohms (ceramic, 5Watts).

With really good coolers (TEC or water-cooler) you can go up to 2Amps, what's set with a resistor with 0.6 Ohms ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

The diode laser is press-fitted into a 12mm copper housing, which I was planning to fit into a drilled-out piece of aluminum bar stock using heatsink compound, then fit the heavy aluminum bar stock with fan-cooled CPU heatsinks. I also considered drilling out a thermistor measurement well close to the diode laser to keep track of the heat during initial testing.

Looking at the datasheet it seems a single LM338 should be able to handle as much current as I need. Is there a reason to use more than one in parallel (such as managing heat dissipation better)?

TEC seems like a good option.

Thanks again for the detailed suggestions, they are much appreciated!

... if you'll run with only one amperage setting and modulate the average power with PWM, then a single LM338 is enough - if you want selectable max. powers, then you can either switch different sense-resistors to the LM or you can arrange several LM's with different resistors and select one or more of them (averaged sum) for your wished power setting.

I'm selecting one ore more of the five parallelized LM's with jumpers or switches to get max. settings of 1, 1.5, 2, 2.6, 3, ... to 8.6 Amperes max., as I'm using different diodes with currents betwen 1 to 9 Amps.

But this can be done with 100mA-settings too for even finer selection - have some BluRay-diodes with 200mA too ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Yeah, I used LM type regulator between laser and heatbed 12v connector.
Correct timing then needs that quite "pervert" G code sequences, because heating seems to be quite low timing priority in Marlin firmware.
Other approaches exists, but mine experiment was "quick'n dirty"

OK, I've been thinking about this and reading through some of the designs out there and I think I understand what I want to do...

I should be able to separately supply 5V and 12V (for fans) so there is less wasted current. I could use a separate optocoupler and MOSFET for the 12V rail for fans (lens blower and cooling fans). I was able to get a good deal on 2mm copper plate from my local surplus metal supply so I have a stack of copper plating bolted together to drain heat from the LD - should be almost twice as efficient as aluminum and weighs about the same as an extruder stepper motor. The 5V rail on the ATX power supply I'm using for my Prusa i3 is unused and should easily be able to supply the current needed for the current regulator circuit and LD.

Basically I can use an LM338 with resistors to set the MAXIMUM current. I could also use jumpers or DIP switches to enable selection of resistors for max current.

The optocoupler to MOSFET allows an analog output pin from the Arduino to be used with PWM to modulate the AVERAGE input current to the LM338, thereby allowing lower power. I cannot use PWM modulation with a MOSFET to control max current because it basically gives me an average and even a momentary transient above the maximum current for the LD will fry it. That much seems clear.

The only part I'm not entirely clear on is the use of capacitors (10uf seems to be common) on the output side. It seems there needs to be a diode in there to drain the output side capacitor when power is cut with the LD disconnected. Would a simple 5V LED do the trick? That seems useful anyway...

I'll try putting together a schematic...

Here is my first pass at a working schematic. My design skills are beginner level but the basic idea is to take a 12V input (from a PC power supply) and regulate voltage to around 7.5v using an LM338, then another LM338 with the current sense switchable. The bottom range current is for focusing. I've stuck with the mfr "safe" current range for longer life.

To be added: protection diodes (to guard against capacitor discharge), optoisolator and power transistor for Arduino on-off control.

kicad sources are on github...

... I'm driving my diodes with 12Volts and a good current regulator circuit (no startup/shutdown spikes) -- the diodes 'sees' only the current passing through, not the supply voltage ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Thanks for the quick response. The thing I'm confused about is that the datasheet for the NDC7875 says operating voltage is 3.7 - 5.5vdc.

This model has some thermal protection built in also. When I've accidentally driven it up around 8v or so it begins to blink.

I also looked at using a 2A Murata DC-DC converter to get 5.5V from the 12v input (since I also need that to run the fan) but figured I can do the voltage stepdown with another 338 (I have several).

So am I not understanding the datasheet correctly? I think the datasheet says "operating voltage" meaning "forward voltage". In all my tests with a bench power supply I've set voltage to 5.5 and adjusted current from around 0.05A to 1.7

... as you use the LM chip not in voltage but in current regulator mode, the 'forward voltage' is reduced to the actual current running through the LM and the diode and is so defined by ohms law or the voltage drop across the diodes resistivity.

This is totally different, when you'll control the laser power with a voltage-regulator!

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

You don't need a power regulator. Just use a Flexmod P3 Driver. It is only $35. It takes a constant power supply and uses a trigger. You should always power your laser with a seperate power supply from the one you are using with the ramps or anyother board espically if you are using stepper motors. The back EMF from steppers is where you would get a voltage spike from. I used (2) 12v / 5A power supplies for mine. 1 for ramps and 1 for the laser. Simple, Cheap, and Effective.