400 - 5mW diode configuration questions.

At a guess, with my (as of yet limited) knowledge of electronics, I believe:
Say voltage drop across a diode is 2.7v - you could hook 4 diodes up, with total 10.8v drop. You could drive this chain of diodes with a LM317 circuit set to the appropriate current, and hook that up to the +12v rail on the PSU. Thiis part could then be an interchangable "branch", and the complete machine made up of say 20 "branches" connected to a PSU. Just make sure not to over-stretch the PSU - I have blown power supplies before, and cheap power suppliies may behave unpredictably under odd conditions such as these. Don't actually build this until someone confirms my theory, and remember that the 317 is very inefficient, wastiing energy as heat, and also that the heatsink is actually "bound" to pin 2 of the regulator.

Ok I have been looking around and happened to stumble upon a site that discusses driving multiple diodes.  Here is what the guys says on the subject:

"There are probably other ways to do this but this is one way that works. The laser diodes can NOT simply be hooked in parallel as the diode with the lowest forward resistance will consume most of the current then blow. When it blows the result will most likely be the immediate destruction of the other laser diodes due to the increased current assuming the first diode blows open and does not become shorted. A similar problem exists when hooking laser diodes in series.

The only safe way is to provide each laser diode with it's own current limiting resistor as each diode may require slightly different bias. With this circuit should something happen to one of the laser diodes the others will be unaffected."

I attached a pic of the circuit this guy designed.

I am going to try and prototype this in the next week or so.  I have to purchase some supplies first.

Forgot  to add the last paragraph and props to the guy:

"As you parallel laser diodes you may need to reduce the value of resistor in the source lead of the IRF510. In fact you can simply eliminate it and tie the source lead to ground if you are using square wave modulation where linearity is of no concern. Note that seperate test points are provided to monitor the current to each laser diode. I've used this circuit with up to 16 laser diodes with no problem.

By John, K3PGP 08/04/97 - Rev 1.05"

hapyman, du you need modulation? if not you can leave out the mosfet in this circuit and also its driving (all what is left of the 10uF capacitor ) and the 1n4007's
Then this circuit reduces just to a set of current limiting resistors for each diode, the easiest way to drive them, but no active regulation. In your case the overall current over the 7805 would be 400x30mA = 12A. It will never survive this! Also the resistor directly after the 12V supply has to be made much smaller but had to take roughly 60Watts (12V-5V-2V dropout)*12A !! This would be highly inefficient.
But when you want to go to current limiting resistors only you can again drive all the diodes just in parallel from 3.3V and adjust the resistor accordingly (R = (U_supply-U_drop)/I_diode =(3.3-3.2)V/30mA = 3.3Ohm.
Please recheck my calculations, I always do mistakes.

Andreas

Thanks for the input Andy. I think I would need modulation and I kind of like the fact that it will remove the extra charge from the capacitor. Please correct me if I am wrong but this would basically allow me to run my device even if a diode burns out.

I realize this circuit looks like overkill but with this many diodes this looks like a great option to me for security. I really wouldn't mind the added regulation on top of the current limiting resistors for each diode.

Also I wasn't planning on running all the diodes through this one driver. I should've specified that ;D. Anyhow I would probably require quite a few of these and basically run the diodes in clusters with each cluster having their own driver. The question would be how many, the guys says he ran up to 16 - 5mW diodes I believe.

Quick question though. Is there anyway to EASILY modify this circuit to allow a different input voltage. I think it will be harder to get my hands on a 12V power supply with this much juice. A 5V or 3.3V power supply with 12+ Amps is much easier to find.

Once again thanks for taking the time.

hapyman said:

Quick question though. Is there anyway to EASILY modify this circuit to allow a different input voltage. I think it will be harder to get my hands on a 12V power supply with this much juice. A 5V or 3.3V power supply with 12+ Amps is much easier to find.

Once again thanks for taking the time.

Click to expand...

Problem is that with 5V or 3.3V you can use no linear regulator anymore because the standard ones have around 2V dropout. But a computer switchmode supply should be regulated very well, so the extra voltage regulation is not needed. When you want extra security maybe you can use an additional active current regulation. One possibility would be like in this circuit: http://www.die4laser.com/dvd-rec/Die4Drive_files/Die4DriveRev1-2.pdf
Only the part right of the opamp
You just need the sense resisor, a mosfet and an opamp extra (plus some small capacitors and a reference voltage).

Sorry, forgot about the voltage drop.

So with the circuit I posted earlier I could use an unregulated 12V power supply and not worry?

Or I could just use an ATX style regulated (switching) supply and just use some current limiting resistors attached to each diode?

hapyman said:

So with the circuit I posted earlier I could use an unregulated 12V power supply and not worry?

Click to expand...

Right, but you have to waste a lot of energy (12V-V_diode)*I_diode

hapyman said:

Or I could just use an ATX style regulated (switching) supply and just use some current limiting resistors attached to each diode?

Click to expand...

This would be the easiest way to drive the diodes, and should work well when the power supply is well regulated. The voltage of the supply should be as close to the diode voltage as possible, to use mall resistors and keep power losses low.

Awesome.  You are right that is a huge amount of energy being lost with the diodes we are using. It would probably be lost as heat as well no?

What extent of regulation should I be looking for in a spec sheet for a power supply?  I heard there is sometimes a small voltage spike when a power supply is turned on.  Obviously this will differ from supply to supply and these diodes can be fairly sensitive.  Are we talking a 1% or less range?

Would a soft start circuit help this then?

Also I was thinking then maybe I could place some diodes in serial with them each having their own current limiting resistor but then if one diode would blow then the voltage would spike for the others. That is where this guys circuit has the advantage I guess... he built in protection for the other diodes if one blows.

Thanks again.

I have yet to be convinced that red light heals, but putting that aside and assuming it works...

From what I've heard it's primarily 635 and 800-something that offer the most benefits. Coherence is not a factor, so why not use LEDs? Far cheaper, easier to drive, and easier to mount.

This article sums it up nicely:

Some time ago, this very question was posted to an Internet laser discussion group in response to the assertion that light emitting diodes (LEDs) and other light sources might work as well in the clinic.  This topic was raised again in the June issue of Acupuncture Today. In this article, we will explore how coherence, a property unique to laser light, creates a dynamic, asymmetrical energy distribution within tissue unlike any other light source.

Light waves, which are aligned perfectly in space and time, are coherent. They will unite to increase the amplitude of the combined waveform, and thus, the intensity of laser light.

On the other hand, light waves that are out of phase and opposed will subtract from the strength of the united waveform. If perfectly opposed and equal, they will even cancel one another out.1

Speckling Is Unique to Laser Light

Together, constructive and destructive interference produce the visually stunning phenomenon known as speckling. Try this: Shine an optical wavelength or therapeutic laser, even a laser pointer, on white paper or against a wall, and observe how small bits of light will seem to dance and move with a life of their own. This phenomenon can also be detected at depth in tissue being irradiated by a laser.

As a laser beam penetrates tissue, variations in optical density will bend portions of the beam. Speckles are regions where laser light is reinforced or weakened. This uneven distribution of energy, unique to laser light, is dynamic and vibrant. It is almost as if coherent light were alive. In contrast, if you shine a flashlight or LED against a wall or on a piece of paper, you will note that the energy distribution is flat and motionless.

LEDs Have a Relatively Broad Bandwidth

The light from LEDs is distributed across a much broader spectrum than that of lasers. Whereas LEDs typically emit across a bandwidth of 30-100 nanometers,2 the spectrum of a laser diode is characteristically 1-10 nanometers. Helium-neon lasers that have very long coherence have an extremely narrow spectral distribution of less than one-tenth of a nanometer.3 The narrower bandwidth of lasers may have significant physiological effects.

Lasers vs. LEDs in the Scientific Literature

It should be pointed out that nearly all of the thousands of studies and clinical reports that have investigated the effects of monochromatic light at low intensities have been performed with lasers. The fact that relatively little research has been done with LEDs speaks volumes. Although the physiological effects specific to lasers may be difficult to isolate, the widespread popularity of laser therapy and larger number of laser studies suggests that these differences are significant.

Jan Tuner and Lars Hode have compiled a collection of research comparing therapeutic lasers and LEDs. What follows is a summary of the results of some of these studies paraphrased from their excellent book, Laser Therapy, Clinical Practice and Scientific Background.

Hmmm I am going to look into active current regulation. This would help prevent a chain reaction of diodes burning out right?