Simple "universal" laser diode driver

[The March Muppet]

Happy New Year - although mine started exceedingly poorly - My silly GSD with a ball lodged in his throat at 5AM, wondering why everyone was trying to suddenly kill him trying to get it out. We got it, he's fine, I was rattled but have also recovered.

...

I am doing my "homework" - but there's a lot out there that just looks wrong.

Lots of LM317 based DIY drivers, some wired CC, some CV, some with bizarre component valuers, and some just incorrect.

I suspect the reason that many of them appear to work for their creators is that they're playing with high current diodes that the 317 doesn't have enough output to blow up.

I just want to muck about with a pile of (likely 0.5 - 2 mW max) older LD's in bits of various optical bits I've salvaged over the years - for example a bar code scanning head.

Not really certain how to set up for it though. Seems a typical Vfwd for red is 2 to 2.3 V with current anywhere depending on the device. I don't believe that I have any violets, but probably have some drives that contain them so may want to play in that sand box eventually.

Presumably "any" diode will safely operate (if nowhere near optimal) series current limited to say, 20mA? - so it would appear that a setup with adjustable voltage regulation and a simple series resistor should light anything up without killing it..

But also presumably, once the diode is lit, I'd then want to regulate current.

Of course to what level - as I recall they achieve a certain subjective intensity and then die with an increase in current. You only know you're too far when you just fell through the ice - though seem to also recall that once the device was lasing, additional current had little effect on intensity.

Can you even daisy chain adjustable CC and CV circuits?

I happen to have a nice old 500 Ohm wire wound pot - toss this on the end of a 317 adjustable voltage reg, bring the voltage up until it just lases, and then use the pot to increase current?

Anyone have much experience playing with unknown diode "open loop" play?

As well, are they "tougher" than they once were - or just cheaper? I recall red diodes being in the $450 range in the early 90's: handling precautions as if they were full of nitro, and feedback "mandatory" - pretty sure the dollar pointers must be just running a resistor.

Any help appreciated.

 

[The March Muppet]

Can't see offhand why something like this wouldn't work, but I've never ganged regulators.

 

[ronsimpson]

I've never ganged regulators.

It should work. Left regulator sets the maximum voltage and the right sets the max current. The question is why set the max voltage? If things are working well, the current and diode will set the voltage. (Using a 2 to 2.3V diode at its rated 20mA then 20mA will get you to the right voltage) I think just drive the CC regulator with a reasonable voltage.

 

[Visitor]

Constant current devices can take some thought to wrap your head around.

It might be worthwhile to set up an LM317 as an adjustable constant current source, say 20mA maximum, connected to a 12 volt power supply and experiment with some LEDs of different colors. Start with maybe 5mA, connect an LED and measure the resulting voltage across it.

The LED will draw the fixed current – the resulting voltage depends on the chemistry of the LED and is called the forward voltage. This will vary a little with the supplied current but it's fixed by the LED. You don't have to control the current AND the voltage and in fact, you CAN'T control both at the same time.

 

[The March Muppet]

why set the max voltage? If things are working well, the current and diode will set the voltage.

Will they? I've never heard of an LED that has a maximum voltage spec, and they can of course practically operate at high voltages albeit with impractical series limiting resistors - but several LD's I looked at randomly do - and just a few hundred mV above "operating". As well the appear to be extremely sensitive to spikes and ESD - instantly turning from laser into expensive dim LED's.

Visitor - Laser diode, not LED emitter.

in fact, you CAN'T control both at the same time

This is what I have always thought as well but not entirely certain as to why. If the regulators somehow fight, though I'd expect that they'll do so reducing either E or I, but not sending the LD into magic smoke hell.

 

[Visitor]

Visitor - Laser diode, not LED emitter.

Understand how constant current devices operate – start with something cheap. Once you have the basic understanding, the rest will start to make sense.

 

[gophert]

Visitor - Laser diode, not LED emitter.

The constant current power source can't tell the difference. Who cares?

 

[Nigel Goodwin]

Will they? I've never heard of an LED that has a maximum voltage spec, and they can of course practically operate at high voltages albeit with impractical series limiting resistors - but several LD's I looked at randomly do - and just a few hundred mV above "operating". As well the appear to be extremely sensitive to spikes and ESD - instantly turning from laser into expensive dim LED's.

You need to stop thinking voltage, LED's are current devices, and have a threshold voltage just like a zener diode (and in fact LED's are often used in voltage regulators) - and they don't operate at high voltages, the series resistor 'may' have a high voltage at one end of it, but at the LED end it's just the usual low voltage.

So no need for a voltage regulator, it's completely superfluous.

Laser diodes are just specialised LED's, with more stringent requirements, and much more easily damaged - in old CD Players great lenghs were gone to in order to prevent the LD been killed.

This is what I have always thought as well but not entirely certain as to why. If the regulators somehow fight, though I'd expect that they'll do so reducing either E or I, but not sending the LD into magic smoke hell.

Why can't you understand it?, it's simple ohms law. You can only have constant 'V' and constant 'A' if 'R' is also constant - and you only need either 'V' or 'A' to be constant, the other wil follow automatically because 'R' is constant. But none of this applies to an LED as you don;y have a constant 'R'.

In the series regulator case you drew above there's no suggestion of a constant 'voltage and current', you have a constant 'voltage' feeding the constant current source, and a constant 'current' feeding the diode. The two are not related in any way, and are entirely separate.

 

[The March Muppet]

The constant current power source can't tell the difference. Who cares?

This had not been crystal clear to me. I have been under the impression that LD's contained some sort of fragile voodoo, and were rather different and unique from jelly bean LED E/I curves.

Why can't you understand it? ... In the series regulator case you drew ... not related in any way, and are entirely separate.

Got it. So the circuit would "technically" function - But all that the first reg would do would be to allow the second to dissipate slightly less power. irrelevant as I'm going no where near max specs anyway. So no point.

So why all of this "driver" stuff if a well regulated power supply and nothing more than a pot are required?

 

[gophert]

I have been under the impression that LD's contained some sort of fragile voodoo, and were rather different and unique from jelly bean LED E/I curves.

Sorry, I don't know anything that behaves as if voodoo is involved except for...
- the intermittent performance of a damaged component
- a bad solder joint
- a cracked copper trace on a PCB, or
- girlfriends
...

 

[Haidilao]

Diode (**broken link removed**) is one of the earliest semiconductor devices, and its application is very wide. Especially in various electronic circuits, using diodes and resistors, capacitors, inductors and other components to make reasonable connections to form circuits with different functions can realize rectification of AC power, detection of modulation signals, limiting and clamping, and power supply. Various functions such as voltage stabilization. Traces of diodes can be found in common radio circuits or in other home appliances or industrial control circuits

 

[unclejed613]

This had not been crystal clear to me. I have been under the impression that LD's contained some sort of fragile voodoo, and were rather different and unique from jelly bean LED E/I curves.

i've messed around with lasers removed from defunct blu-ray players. the "blue" laser (looks like a bit more on the violet end, so a bit shorter wavelength than you would expect when comparing it to a blue LED) has a forward voltage of about 3.6-3.8V. where a laser differs from an LED is when you get to a certain threshold current, the laser crystal begins amplifying, and what was a dim light output suddenly changes to very bright. most of the blu-ray lasers i have worked with have a gain threshold at 35-50mA. this is highly temperature dependent. and heat sinks are usually required. making a variable constant current source is quite easy...
https://www.falstad.com/circuit/e-howland.html

you can add a voltage slider to the "-5V" input node, set the minimum voltage to -10V and the maximum to 0V and now the current source is continuously variable.

 

[The March Muppet]

Sorry, I don't know anything that behaves as if voodoo is involved except for...

You forgot classic cars.

 

[gophert]

You forgot classic cars.

No, I didn't forget anything. Note the "..." at the end of my list. I just didn't bother to type it.

 

[CocoaTea]

i've messed around with lasers removed from defunct blu-ray players. the "blue" laser (looks like a bit more on the violet end, so a bit shorter wavelength than you would expect when comparing it to a blue LED) has a forward voltage of about 3.6-3.8V. where a laser differs from an LED is when you get to a certain threshold current, the laser crystal begins amplifying, and what was a dim light output suddenly changes to very bright. most of the blu-ray lasers i have worked with have a gain threshold at 35-50mA. this is highly temperature dependent. and heat sinks are usually required. making a variable constant current source is quite easy...
https://www.falstad.com/circuit/e-howland.html

you can add a voltage slider to the "-5V" input node, set the minimum voltage to -10V and the maximum to 0V and now the current source is continuously variable.

Sorry for reviving, just thanks!

 

[unclejed613]

this one isn't very old.... usually if there's one that's extremely old... like 2015 or something you might have somebody say something about it.... maybe even me posting an animated GIF from young frankenstein, of igor and the doctor digging up the body for the monster...