Sanity Check on Proposed Low-Power Cyan Diode Laser Build

[imallett]

Hi,

I would like a cyan (465–495nm-ish) laser for casual use (i.e., actually pointing with it in public settings where people won't be wearing safety googles)—necessitating an output power of 5mW. Whereas, the (few) cyan laser products that exist are either expensive (example: Laserglow Aquarius, $399–549) or high-powered (example: Sanwu Pocket, 60–100mW).

However, it doesn't seem impossible for me to build such a pointer myself. I have reasonable engineering skills, including soldering, and a smattering of handy electronics equipment. I don't have domain knowledge about laser circuitry or EE in-general, but I've done my best to read up on it and am happy to treat this whole project as a learning exercise. Budget is soft-limit of $150, and I think it can be done for about $100. If I am successful, I might consider selling clones of it at a small markup—reading around, cyan lasers seem to be in high-demand!

I'm looking for a sanity check on my analysis so-far and proposed procedure, before I actually go ahead and try my hand at it.

---

I found a 490nm diode for $35 from DTR, which seems to be the best single-unit price for a cyan diode. The diode+driver+module is $86 and outputs 55mW.

To reduce the power, I considered adding a simple neutral-density filter to cut it down, but the "Right" way, I guess, is to reduce the current the driver maintains. This will significantly reduce heat load and improve diode life. The diode's datasheet suggests that the diode will work at 5mW output: based on the typical 105mA, 55mW, 0.8mW/mA operation, 5mW should be achieved at 42.5mA, while the threshold current is 40mA. This is corroborated by the associated graph. And, if need be, I suppose I could keep the output a higher than 5mW but also put in a filter.

If I go with the pre-built module, I'm somewhat confused: the listing says "2.5V-6V 1A+", and contextually, I guess this means it accepts any 2.5–6V power supply which can deliver at-least 1A? This is more power by at-least 4⨯ than the diode normally consumes. Is this just headroom? It would be really convenient to power via two AAs, although common alkaline ones cannot supply 1A. Drawback: since the driver is not listed, I'm not sure what amperage output it supports. Maybe these are all questions for DTR?

I could alternately assemble the module myself, which seems like it would be both cheaper and more-educational. The LDSE500 driver looks okay—supply voltage 2.9–9V, supporting a load of 2–8V (diode drop is 6–7.5V). the maximum output current (and therefore attainable precision at lower currents) is a bit high, but it should be fine? Says it comes pre-calibrated to 50mA, although I don't know what simulated load that's for.

I didn't immediately find a good host to put the module in, and thought I should get the above sorted out first. I'd appreciate suggestions though.

Also, I considered starting with a somewhat-cheaper diode, but $35 isn't ridiculous, and since other diodes are quite different electrically, I thought it might introduce too many variables to count as salient practice.

So . . . proposed parts list:
• OD 2 or higher glasses appropriate to this wavelength (5mW is target output, but if I'm starting from 55mW . . .)
• 490nm laser diode, $35
• LDSE500 driver, $25
• V5 12mm module with acrylic lens, $9
• Simulated load diode and resistor (how do I spec these?)
• Host, TBD (suggestions?)
• Power supply, TBD (suggestions?)

---

Thoughts? Comments? Corrections? Suggestions?

Ian

 

[absolute]

You've certainly done your homework, what you have looks good so far.
I've recently put together a 488nm build myself and I was going for about that budget too.

I would choose a three element lens instead of the acrylic one even if you're going for low output because the beam characteristics will be better at a small cost increase, but it is not necessary.

For a budget host I can recommend user trinh hong puoc, take a look at his or other host builders sale threads and see if you like anything. You can contact some of them for custom builds or questions.

For power you will have to see what fits your host, the driver will accept 1 cell or 2 cells Li-ion of your choice (10440, 14500, 16340, 18650). I wouldn't go any larger than say an 18650 host because the output will feel puny compared to the size of the host.

For the simulated load I used the most available diode I could find 1N4001 and a resistor in series to get a current voltage ratio in the range of what the laser diode should be receiving.
You'll be running it at the low end of the diode so about 6V and 42.5mA (I'm quoting your numbers here I didn't double check them).
Vf of 1N4001 is about 1.1V, leaves (6-1.1) 4.9 volts to drop at 42.5mA, Ohm's law will give you about 115 Ohm resistance value, nearest common one is 100 Ohm so you can use that. Be sure to use a resistor of the right power so it doesn't start smoking during testing
Measure the current in series with the diode+resistor and set it at the value you want. The current will be the same when you connect the diode even if you're a little off with the load because the driver takes care of that by adjusting the voltage accordingly.

Try not to fry your driver by forgetting to connect the test load and ALWAYS double check polarity of everything you have connected, I learned the hard way haha

 

[DTR]

To reduce the power, I considered adding a simple neutral-density filter to cut it down, but the "Right" way, I guess, is to reduce the current the driver maintains. This will significantly reduce heat load and improve diode life. The diode's datasheet suggests that the diode will work at 5mW output: based on the typical 105mA, 55mW, 0.8mW/mA operation, 5mW should be achieved at 42.5mW, while the threshold current is 40mA. This is corroborated by the associated graph. And, if need be, I suppose I could keep the output a higher than 5mW but also put in a filter.

If I go with the pre-built module, I'm somewhat confused: the listing says "2.5V-6V 1A+", and contextually, I guess this means it accepts any 2.5–6V power supply which can deliver at-least 1A? This is more power by at-least 4⨯ than the diode normally consumes. Is this just headroom? It would be really convenient to power via two AAs, although common alkaline ones cannot supply 1A. Drawback: since the driver is not listed, I'm not sure what amperage output it supports. Maybe these are all questions for DTR?

Yea don't do this. It is the drivers job to provide a fixed current at all times when the power source is in the listed range. These are buck/boost drivers and a linear and buck can be pushed out of current regulation by limiting the current demand the driver is trying to draw but I have noted quiet a few times that boost drivers do not like being starved of current. It is like the it will freak out sputtering with very very fast cycling and I have seen it damage more sensitive diodes like the 405nm so I discourage any current limiting to the laser driver and it should be capable of delivering 1A to keep a large enough overheat for any any scenario in that range which I will give a little more detail on the confusion in the below statement.

the voltage is over the minimum of 2.5V and the voltage must stay under the max input voltage or it can damage the driver usually in the few seconds before the chip burns over the max voltage it pushes it out of constant current regulation and can push a large surge of current into the diode killing right before the driver chip smokes and if the chip smokes first it will probably pass something nasty to the diode which kills it as well. I do have 6V as it is paired with the flexdrive V5. It is a really good pairing. I do set these pretty conservatively with 200mA by default which seems to provide good power and maintains the wavelength pretty well.

I could alternately assemble the module myself, which seems like it would be both cheaper and more-educational. The LDSE500 driver looks okay—supply voltage 2.9–9V, supporting a load of 2–8V (diode drop is 6–7.5V). the maximum output current (and therefore attainable precision at lower currents) is a bit high, but it should be fine? Says it comes pre-calibrated to 50mA, although I don't know what simulated load that's for.

I didn't immediately find a good host to put the module in, and thought I should get the above sorted out first. I'd appreciate suggestions though.

Also, I considered starting with a somewhat-cheaper diode, but $35 isn't ridiculous, and since other diodes are quite different electrically, I thought it might introduce too many variables to count as salient practice.

I can set it to lower currents or even use a LDSE500 if you request but first I will not sell a unit with any confirmation that you can get it to eye safe 5mW and certainly would not set a driver sending one saying it is there even if the diode will lase below 5mW for obvious reasons. On those it would be up to you to set try to get it there. With this specific diode the testing gave 6mW at lasing threshold but also under 20mW on this LPM being a 20W capable meter I would not bet my sight on accurate irresolution till 20mW or so. I actually think that is in the datasheet for the accurate range for this Ophir LPM head and like a 5% variance up to 200mW then 1% up to 1W an say 0.5% to 2W ect.... getting near 0.05% +/- at full 20W. Anyway there is no scenario you can run this diode consider it eye safe. Laser safety glasses that are certified for this type of laser is needed. You can look at the ones SL offers in the link I will give below.



This is a common point of confusion. Being a buck/boost it will either boost the voltage which draws more current to do this or it will buck which would draw less current than delivering to the diode depending on the voltage of power source and the forward voltage demand of the diode. If you are using 2.9V it might drawn up near 700mA-800mA to deliver 200m @ the forward demand of the diode which does vary based on temperature int he junction which is the reason a constant current driver is required. If you are using the LDSE500 and used say 8.5V it might start actually drawing less current than 200mA to deliver 200mA as it start bucking which is when the load demand minus the dropout in the regulator(usually 1V-1.5V with most efficient ic's) under the supplied voltage from the power source or more.

The other common misconception is while yes if you say have 3V given to the driver and it is drawing 700mA that is 2.1V of electrical energy then the laser puts out 200mW optical energy giving 9.5% conversion efficiency(those numbers are off as the driver will be much more efficient than that but give the idea) and the lost 1. 9W of energy is why a copper module is a

So . . . proposed parts list:
• OD 2 or higher glasses appropriate to this wavelength (5mW is target output, but if I'm starting from 55mW . . .)
• 490nm laser diode, $35
• LDSE500 driver, $25
• V5 12mm module with acrylic lens, $9
• Simulated load diode and resistor (how do I spec these?)
• Host, TBD (suggestions?)
• Power supply, TBD (suggestions?)

---

Thoughts? Comments? Corrections? Suggestions?

Ian

That will all work and I do sugget a single Li-ion power source regardless the driver you choose if the end application will be batery powered.

For hosts there is a member here that makes some awesome full custom units that will accept all my 12mm, 20mm or even the 25mm modules. See here.

Sale thread/ Hosts/Heat Sinks/Focus adapters/lasers

:bumpit:Hi, I thought i would combine all the services i offer in one thread. I posted just some but if you have something in mind thats all good also.I have other hosts as well, Heat sinks are in Aluminum/Aluminum with copper sleeves and full copper, all heat sinks are included with focus...

laserpointerforums.com

Then as mentioned above another member here has a great site which offer safety eye ware that is certified and hosts that will accept both my 12mm modules get the extended copper sink or the drilled pill accessory to accommodate the entire module without having to remove the back half on their C5 style hosts as those can bottom out a few mm's before flush with the standard heatsink which also makes it so you can pinch the leads maybe penetrating the wire cover grounding the positive and it will instant short. You might not even know that it did so which can burn out the switch. Or get the 20mm module and the 20mm adapter for his kits and no special considerations on that.

Safety Goggles

For current CE certificates for our Eagle Pair Goggles, please click these links: EP-1 through EP-10 EP-11 through EP-18

www.survivallaserusa.com

 

[paul1598419]

Keep in mind that the data sheet threshold current is rarely the threshold current of your diode. They are all slightly different, so the output power at threshold will be different as well. If you want a truely eye safe laser, you would need to test several to see which will give you 5 mW or less at threshold current.

 

[imallett]

To reduce the power, I considered adding a simple neutral-density filter to cut it down, but the "Right" way, I guess, is to reduce the current the driver maintains. [...]

Yea don't do this. It is the drivers job to provide a fixed current at all times when the power source is in the listed range.

To clarify, I meant to set the driver's output current to lower, so that the optical output is lower. I wasn't considering throttling the driver's inputs. As I understand it, the driver will adjust its output voltage over some small range so that a constant current flows across the diode. The magnitude of that target current can be set with its pot.

Although, I do retain some confusion on how to spec. out a power source that is appropriate for the driver's inputs. Clearly, one picks something in the range of valid input voltages, but how do I pick the amperage the source must be able to provide? At a total guess, something like:
(min source amperage) = (driver max output wattage)/((source voltage)(driver efficiency)) + (margin) . . . ?

I do have 6V as it is paired with the flexdrive V5.

I looked at the Flexdrive V5, but I didn't see how it could be viable? The specsheet for the diode says it can require up to 7.5V (typ. 6V), while the Flexdrive V5 says it can provide 2–5.5V. I figured that would be a problem.

I can set it to lower currents or even use a LDSE500 if you request but first I will not sell a unit with any confirmation that you can get it to eye safe 5mW and certainly would not set a driver sending one saying it is there even if the diode will lase below 5mW for obvious reasons. On those it would be up to you to set try to get it there. With this specific diode the testing gave 6mW at lasing threshold but also under 20mW on this LPM being a 20W capable meter I would not bet my sight on accurate irresolution till 20mW or so. I actually think that is in the datasheet for the accurate range for this Ophir LPM head and like a 5% variance up to 200mW then 1% up to 1W an say 0.5% to 2W ect.... getting near 0.05% +/- at full 20W. Anyway there is no scenario you can run this diode consider it eye safe.

Keep in mind that the data sheet threshold current is rarely the threshold current of your diode. They are all slightly different, so the output power at threshold will be different as well.

What I'm getting out of this is that 5mW is comparable to the lasing threshold for this diode, so it may or may not be possible to get it to 5mW. While DTR would, should I ask, set it to 5mW or close, they would not make any guarantee about it, and I would have to measure the output power with an LPM designed for lower power ranges and make adjustments. Safety glasses would be required during that procedure.

(I'm making notes on the power and host resources suggested.)

Also, in-parallel, I emailed Sanwu, and they said they could manufacture at 5mW by simply asking for that in the order notes. This does not mean I'll be stopping this project—on the contrary, I'd like to continue it, at least for this wavelength. But, I thought I would mention it in-case anyone wanted to go out and just buy something.

 

[ZRaffleticket]

What about a ND filter on the output? That way threshold current isn't as much of an issue... plus that's what CNI does.

 

[imallett]

To reduce the power, I considered adding a simple neutral-density filter to cut it down . . . . I suppose I could keep the output a higher than 5mW but also put in a filter.

What about a ND filter on the output? That way threshold current isn't as much of an issue... plus that's what CNI does.

Seems like a good approach, if the diode in-question doesn't work at 5mW. One might just plan on setting to ~10mW and cutting in half with OD 0.3, meaning the diode is operating more in its design range (more stability?). Although, I looked around a bit and didn't see ND filters that were inexpensive—I expect because I was looking at filters with tightly-controlled, professional tolerances (which shouldn't be necessary, given that the output power is set from the output of the whole system).

Sooo . . . outstanding issues:

• How do I spec. out a driver that is appropriate to the diode? I assumed you pick out a driver that can supply the maximum voltage the diode requires and the maximum current expected to run with (with headroom on both). However, I'm confused by what DTR said; I don't see how the Flexdrive V5 could work since its maximum output (5.5V) is lower than the diode's typical input (6V), let-alone max input (7.5V).
  
  
• How do I spec. out a power supply for the driver? As-above, after picking the source's voltage within the valid input range, the minimum amperage the power source must be able to supply would be something like:
  (min source amperage) = (driver max output wattage)/((source voltage)(driver efficiency)) + (margin) . . . ?
  
  
• Thoughts on sourcing ND filters?

 

[rvrh3]

Sooo . . . outstanding issues:

• How do I spec. out a driver that is appropriate to the diode? I assumed you pick out a driver that can supply the maximum voltage the diode requires and the maximum current expected to run with (with headroom on both). However, I'm confused by what DTR said; I don't see how the Flexdrive V5 could work since its maximum output (5.5V) is lower than the diode's typical input (6V), let-alone max input (7.5V).
  
  
• How do I spec. out a power supply for the driver? As-above, after picking the source's voltage within the valid input range, the minimum amperage the power source must be able to supply would be something like:
  (min source amperage) = (driver max output wattage)/((source voltage)(driver efficiency)) + (margin) . . . ?
  
  
• Thoughts on sourcing ND filters?

yesss this is what I've been looking for!