Thought I post one of my first Laser projects as my first posting.
Some people might be interested or even take a few bits from my results and problems, that's why I share the Info.
I included the description of my laser driver (Hah: LLD - Lirezh-Laser-Driver) which can be used for the 3-mode setting and is a LOT better than using a linear voltage regulator setup (mainly on blue diodes)
In the past 2 months I invested quite some time in experimenting with 3 2W blue laser diodes.
What I wanted:
A 3-mode 2W laser-handheld device that can be left on for at least 20 minutes without overheating.
3 mode means:
Mode 1: 50-100mW optical output
Mode 2: 500-800mW optical output
Mode 3: 1.8-2W optical output
The reasoning behind this is that I want a secure way to point the device or set the focus and then increase power when ready.
It's a security meassure and good to have.
I built 2 lasers so far, the second one (see attached images) is the much better model as I used Solarforce parts as case.
I can really recommend the Solarforce body parts except the Power-switch which is very easy to kill with high currents.
That's why I had to replace the switch with a custom part, will find a better solution to the current 3-way switch on the back soon.
I've tried countless circuits to power the diode, this was by far the hardest part.
What I set myself as *requirements* for the laser driver:
* A tiny circuit which can fit into the small solarforce body (18650 size)
* A stable circuit which is regulated and current can be easily changed
* "High" power required, 2Amp at 4.6V .. so about 10Watt power
* temperature stable for "medium" periods
* somewhat efficient circuit, supply of < 9 V
* 100% selfmade! I did not want to buy something (that's like cheating!)
* All parts should available or from scratch, I did not want to buy anything.
What options did I try ?
The first was the circuit everyone here seems to know
A linear voltage regulator (LM317 for example) used as load regulator.
First problem: LM317 does not work for high currents.
Solved: LM350 is able to supply more than 2 ampere which is fine for the job.
Well the LM circuits are simple and stable but, sorry, bullshit for blue diodes.
My first project used such a circuit because it was easy to make but after all this was a "fail" if looked at in detail.
The LM3** in current regulation mode will have a voltage drop of 4.5V and they are in series to the load.
The Diode will take about 4.5 V and a lithium cell provides 3.6V
2x3.6V = 7.2V. So if you take in account the LM3** eats 4.5 of those you do not have enough voltage left for your diode.
So you need 3 cells/batteries = 10.8V (when full charged 12.6V)
12.6V Supply means that 4.5V are at the diode (let's say 2 ampere) and the remaining voltage drops at the LM circuit (8.1V)
In shorter words: Such a circuit turns 2/3 of your power source into pure heat at the current "regulator" which needs an increddible heatsink and eats your battery away fast. (total power of 26 Watt!)
So there are 2 (and a few others I never got into) options:
a) a switching regulator, this is complicate and requires a coil
b) another circuit that does not require 3 batteries to operate
The small size of the Solaforce body means that I did not want more than 2 TO-220 parts in my circuit (that's the case-size from most power transistors or the LM3** parts)
Ok well I tried at least 15 different circuits for current sources, all the possible variantions found in books and on the net.
I tried using mosfets, jfets, npn and pnp variants.
Not a single one of them was acceptable, most were not good for higher currents like my project and all of them were instable.
Many use diodes as reference source but the books did not take into account that such a circuit is instable:
a) temperature changes the diode voltage curve !
b) the current that goes through the diode changes the voltage curve !!
I finally found a soltution by myself, ignoring all the shit from the books:
Lets call my circuit Lirezh-laser-driver (LLD)
I use a TIP-41A NPN transistor (a medium power transistor providing a good (30+) current amplification at 2 ampere collector current.
I use an LM317 to provide a fixed voltage using a multi-turn potentiometer to set the exact voltage, this voltage is fed through 3 resistors into the base of the NPN.
Two of those resistors can be shortcutted by using a 3-way switch, this creates the 3 modes I talked about:
3 resistors (1k+200Ohm+14Ohm) = 50-100mW
2 resistors (200Ohm+14Ohm) = 500-800mW
1 resistor (14Ohm) = 1.8-2W
The emitter of the NPN is set to ground through a shunt (100-300mOhm) which is built from a resistant-wire and increases the emitter voltage.
The diode is connected between VCC and NPN-collector.
By using a 1-2 ohm resistor in Mode 3 (14Ohm base resistor) as test-object I calibrated the circuit by tuning the base voltage (the LM317 circuit) to match 1.8ampere (which results in ~1.75W optical power) and tested it for a while at different input voltages and temperatures to make sure it is stable.
It's not completely stable, the current can go up 20-40mA when the NPN gets very hot which is still fine.
The NPN will turn any voltage above the laser-diode drop(4.0-4.7V) into heat, only a switching regulator could avoid this. So the NPN needs a heat sink, it can be a small one that fits into the solarforce body but make sure it does not touch the case (which is Ground!)
How it works:
The Shunt acts as simple regulator, the voltage drop at the shunt increases the NPN emitter voltage, when it's high enough the base->emitter voltage difference will be below 0.6V which means the NPN will increase C-E resistance.
So by increasing the base voltage the current that can flow through the shunt until voltage drop is cutting power is increased, very stable and very simple.
The two additional (shortcutable) resistors at the base limit the base current enough to drop the Collector current, current amplification is at about 30-50, so when using Mode-1 or Mode-2 the transistor is controlled by base-current without feedback.
When using Mode-3 the transistor is a regulated current source, the feedback is provided by the shunt.
Conclusion
My circuit (LLD) is likely the most simple and still stable circuit for driving a laser diode, it requires just a few electrical parts and can provide large as well as small currents for laser diodes.
The LLD is a lot more efficient than a linear voltage regulator circuit, it will operate fine at VCC 6V-9V.
Of course any switching circuit will beat the efficiency but it's also hell a lot more complicate to calculate or build.
I was able to fit it into about 3-4cm of the solarforce body.
A small LED can be added to signal when the voltage is dropping too low to protect out LIPOs, at this moment it's not included in my setup.
The mechanical setup
I've two left hands, it's amazing this thing actually does not break into pieces when touching it
As I said, I wanted to use parts from scratch. (the solarforce body is already like cheating
The Solarforce body parts are cheap and allow you to perfectly put them together at the size/length you require.
It's just a few $$ for a high quality case and takes away a lot of work to connect your batteries etc.
So 2x18650 batteries fit into the body, the End-cap was modified by me and I replaced the weak push button (which burned through early) by another switch.
I screwed a hole into one of the solarforce body parts and put a 3-way switch in (3 mode selection).
A 20mm plastic tube is used to hold the LLD circuit, it fits perfectly (and strong) into the front part of the solarforce modules and protects the circuit from shortcutting against ground (the solarforce case itself).
I used a 15 year old CPU cooler as heatsink for the aixiz case, I simply drilled a hole into the center.
Use a few drops of water or oil when drilling, it makes the result very smooth !
I mounted the front-cap of my solarforce front-module on the heatsink using 3 long screws and put the Aixiz case (with glass lens) into it using thermal paste.
The heat sink gets about 30Celsius when running for 20 minutes, the bottleneck is likely the tiny heatsink of the NPN. But it should be possible to run this for 30-60 minutes at high power.
I actually don't know if the circuit will overheat after that, the laserdiode heatsink is good enough for any duration.
That's it
Maybe I'll make a few more shots later.
This thing works very well, it easily burns into wood.
When running at 1.8W and focusing it anything that is touched will turn black in an instant.
Fucking dangerous, I doubt my o-glasses protect against a direct hit for more than a fraction of a second ?
I'm still a noob, please take that into account.
I never touched laser diodes before and just started this out of curiosity and boredom.
Some people might be interested or even take a few bits from my results and problems, that's why I share the Info.
I included the description of my laser driver (Hah: LLD - Lirezh-Laser-Driver) which can be used for the 3-mode setting and is a LOT better than using a linear voltage regulator setup (mainly on blue diodes)
In the past 2 months I invested quite some time in experimenting with 3 2W blue laser diodes.
What I wanted:
A 3-mode 2W laser-handheld device that can be left on for at least 20 minutes without overheating.
3 mode means:
Mode 1: 50-100mW optical output
Mode 2: 500-800mW optical output
Mode 3: 1.8-2W optical output
The reasoning behind this is that I want a secure way to point the device or set the focus and then increase power when ready.
It's a security meassure and good to have.
I built 2 lasers so far, the second one (see attached images) is the much better model as I used Solarforce parts as case.
I can really recommend the Solarforce body parts except the Power-switch which is very easy to kill with high currents.
That's why I had to replace the switch with a custom part, will find a better solution to the current 3-way switch on the back soon.
I've tried countless circuits to power the diode, this was by far the hardest part.
What I set myself as *requirements* for the laser driver:
* A tiny circuit which can fit into the small solarforce body (18650 size)
* A stable circuit which is regulated and current can be easily changed
* "High" power required, 2Amp at 4.6V .. so about 10Watt power
* temperature stable for "medium" periods
* somewhat efficient circuit, supply of < 9 V
* 100% selfmade! I did not want to buy something (that's like cheating!)
* All parts should available or from scratch, I did not want to buy anything.
What options did I try ?
The first was the circuit everyone here seems to know
A linear voltage regulator (LM317 for example) used as load regulator.
First problem: LM317 does not work for high currents.
Solved: LM350 is able to supply more than 2 ampere which is fine for the job.
Well the LM circuits are simple and stable but, sorry, bullshit for blue diodes.
My first project used such a circuit because it was easy to make but after all this was a "fail" if looked at in detail.
The LM3** in current regulation mode will have a voltage drop of 4.5V and they are in series to the load.
The Diode will take about 4.5 V and a lithium cell provides 3.6V
2x3.6V = 7.2V. So if you take in account the LM3** eats 4.5 of those you do not have enough voltage left for your diode.
So you need 3 cells/batteries = 10.8V (when full charged 12.6V)
12.6V Supply means that 4.5V are at the diode (let's say 2 ampere) and the remaining voltage drops at the LM circuit (8.1V)
In shorter words: Such a circuit turns 2/3 of your power source into pure heat at the current "regulator" which needs an increddible heatsink and eats your battery away fast. (total power of 26 Watt!)
So there are 2 (and a few others I never got into) options:
a) a switching regulator, this is complicate and requires a coil
b) another circuit that does not require 3 batteries to operate
The small size of the Solaforce body means that I did not want more than 2 TO-220 parts in my circuit (that's the case-size from most power transistors or the LM3** parts)
Ok well I tried at least 15 different circuits for current sources, all the possible variantions found in books and on the net.
I tried using mosfets, jfets, npn and pnp variants.
Not a single one of them was acceptable, most were not good for higher currents like my project and all of them were instable.
Many use diodes as reference source but the books did not take into account that such a circuit is instable:
a) temperature changes the diode voltage curve !
b) the current that goes through the diode changes the voltage curve !!
I finally found a soltution by myself, ignoring all the shit from the books:
Lets call my circuit Lirezh-laser-driver (LLD)
I use a TIP-41A NPN transistor (a medium power transistor providing a good (30+) current amplification at 2 ampere collector current.
I use an LM317 to provide a fixed voltage using a multi-turn potentiometer to set the exact voltage, this voltage is fed through 3 resistors into the base of the NPN.
Two of those resistors can be shortcutted by using a 3-way switch, this creates the 3 modes I talked about:
3 resistors (1k+200Ohm+14Ohm) = 50-100mW
2 resistors (200Ohm+14Ohm) = 500-800mW
1 resistor (14Ohm) = 1.8-2W
The emitter of the NPN is set to ground through a shunt (100-300mOhm) which is built from a resistant-wire and increases the emitter voltage.
The diode is connected between VCC and NPN-collector.
By using a 1-2 ohm resistor in Mode 3 (14Ohm base resistor) as test-object I calibrated the circuit by tuning the base voltage (the LM317 circuit) to match 1.8ampere (which results in ~1.75W optical power) and tested it for a while at different input voltages and temperatures to make sure it is stable.
It's not completely stable, the current can go up 20-40mA when the NPN gets very hot which is still fine.
The NPN will turn any voltage above the laser-diode drop(4.0-4.7V) into heat, only a switching regulator could avoid this. So the NPN needs a heat sink, it can be a small one that fits into the solarforce body but make sure it does not touch the case (which is Ground!)
How it works:
The Shunt acts as simple regulator, the voltage drop at the shunt increases the NPN emitter voltage, when it's high enough the base->emitter voltage difference will be below 0.6V which means the NPN will increase C-E resistance.
So by increasing the base voltage the current that can flow through the shunt until voltage drop is cutting power is increased, very stable and very simple.
The two additional (shortcutable) resistors at the base limit the base current enough to drop the Collector current, current amplification is at about 30-50, so when using Mode-1 or Mode-2 the transistor is controlled by base-current without feedback.
When using Mode-3 the transistor is a regulated current source, the feedback is provided by the shunt.
Conclusion
My circuit (LLD) is likely the most simple and still stable circuit for driving a laser diode, it requires just a few electrical parts and can provide large as well as small currents for laser diodes.
The LLD is a lot more efficient than a linear voltage regulator circuit, it will operate fine at VCC 6V-9V.
Of course any switching circuit will beat the efficiency but it's also hell a lot more complicate to calculate or build.
I was able to fit it into about 3-4cm of the solarforce body.
A small LED can be added to signal when the voltage is dropping too low to protect out LIPOs, at this moment it's not included in my setup.
The mechanical setup
I've two left hands, it's amazing this thing actually does not break into pieces when touching it
As I said, I wanted to use parts from scratch. (the solarforce body is already like cheating
The Solarforce body parts are cheap and allow you to perfectly put them together at the size/length you require.
It's just a few $$ for a high quality case and takes away a lot of work to connect your batteries etc.
So 2x18650 batteries fit into the body, the End-cap was modified by me and I replaced the weak push button (which burned through early) by another switch.
I screwed a hole into one of the solarforce body parts and put a 3-way switch in (3 mode selection).
A 20mm plastic tube is used to hold the LLD circuit, it fits perfectly (and strong) into the front part of the solarforce modules and protects the circuit from shortcutting against ground (the solarforce case itself).
I used a 15 year old CPU cooler as heatsink for the aixiz case, I simply drilled a hole into the center.
Use a few drops of water or oil when drilling, it makes the result very smooth !
I mounted the front-cap of my solarforce front-module on the heatsink using 3 long screws and put the Aixiz case (with glass lens) into it using thermal paste.
The heat sink gets about 30Celsius when running for 20 minutes, the bottleneck is likely the tiny heatsink of the NPN. But it should be possible to run this for 30-60 minutes at high power.
I actually don't know if the circuit will overheat after that, the laserdiode heatsink is good enough for any duration.
That's it
Maybe I'll make a few more shots later.
This thing works very well, it easily burns into wood.
When running at 1.8W and focusing it anything that is touched will turn black in an instant.
Fucking dangerous, I doubt my o-glasses protect against a direct hit for more than a fraction of a second ?
I'm still a noob, please take that into account.
I never touched laser diodes before and just started this out of curiosity and boredom.
Last edited:
