LD blowing up fun

[jeejeedr]

I had some fun yesterday. I blew up about 4 laser diodes. I still didn't figure out why, so I decided to make a post about it. Maybe one of you has a suggestion.

But first my 'project'. I'm building a 100mW 658nm laser module.
I use rohm RLD65PZB5 diode, a die4drive (http://www.die4laser.com/dvd-rec/Die4Drive.htm) LD driver and some collimating optics.


Laser diode glued to the collimator.

The two wires coming from the mount are for TEC, which I did not figure out how to use for the moment.

So, what happens: I connect the LD driver to the didoe, I power up and eveything works... for some minutes. Then whoops, an expensive laser pointer diode. Putting out a milliwatt or so.

I drive the diode at 146mA (well below the max 200mA).
I 'm wired to gound the whole time.
I'm NOT cooling the thing down, but when I put my finger on the diode when it just died, there is no apperent heat .

So I just don't know why they die.

Before killing some others, I need some idea's because it's getting a bit expensive... ($36 on diodes killed so far).

By the way, should I seperatly ground the collimator assembly and the eventual housing of the module?
Should I isolate the LD from the assembly?

[JimBo]

With the size of the diode I expect the heat will disipate within an extremely short time after the diode blew.
I think you need to keep it cool all the time so it doesnt blow.

Jim

[jeejeedr]

With the size of the diode I expect the heat will disipate within an extremely short time after the diode blew.
I think you need to keep it cool all the time so it doesnt blow.

Jim

Any suggestions how? I've been thinking on drilling a 5.6mm hole in a cooling plate and fix it with that white stuff they use for thermically attach a processor to its cooling device. Thus using passive cooling.

[marconi]

Hi Jeejeedr

Fun... :0 Hehe

I think I hold that record!!!

Your mount should have worked fairly good as far as heatsinking is concerned.
If you do use some silicone grease, be sure it is a tiny amount..
too much can hurt it as well.

Sounds like static got to them to me.
Or, They were spiked by your supply.

Be sure you and the devices are grounded during handling..
A good work area may help as well..I am a believer of grounded Anti-static mats .
Also, Keep the diode wires shorted until mounted and you are ready for connections.

I have also tried these Rohm diodes out..at one time they were going to be used in a product, But the failure rate was too high.

Oh, And Robins driver, Be carefull with that..It can still overcurrent the diode
if you put more than 5 volts on the mod input.
Also, That circuit needs just a bit more work as it may spike the diode by overshoot in the pulses..
The diode is floating above ground...be careful not to ground the mount...it will draw full current unless your diode is isolated.

Good luck with your project.

[jeejeedr]

Hi Jeejeedr

Fun... :0 Hehe

I think I hold that record!!!

How many? :lol:

Oh, And Robins driver, Be carefull with that..It can still overcurrent the diode
if you put more than 5 volts on the mod input.

I had 5.15V, maybe a bit too high. :?
I'll put an 7805 between my power supply and the driver.

Also, That circuit needs just a bit more work as it may spike the diode by overshoot in the pulses..

I have an updated version. He made some modifications lately.

The diode is floating above ground...be careful not to ground the mount...it will draw full current unless your diode is isolated.

Ah, I think there could be an issue here.
I think the diode was touching the housing and I touched the housing, and I was grounded... 8)

Good luck with your project.

Thanx, I post things when I have some better results.

Oh,
I also noticed that there is a peltier in my housing.
a 4mm hole in the middle and about 10mmx14.5 (14.7?)mm

Anyone an idea how to find the voltage and maximum curernt to drive this thing, or how to identify the manufacturer (nothing printed on the damn thing).
?

[marconi]

Hehe..
I could buy a nice used car for all the diodes I've done-in

Yep...Ive seen the new board...I think its still is missing a cap.
Maybe it just might have been the test circuit I made up to try it
but I had to add a .22 mf cap across the diode for starters.

Regulated supply is a must but I was talking about the modulation input..
Since most lasershow boards only put out up to 4.99volts
this usually isnt a problem, but on the bench running a generator or
just a supply to run it analog or TTL ..that is the problem.
To get around the problem would involve using another op-amp
to isolate the input overvoltage to just the op-amps limits are..(the rail).

Sounds like it may not be the case of the mount touching but maybe
static..its hard to say without seeing your hookup arrangement.
You might want to add a high value resistor like 22k up to 1meg ohm..
to the base of the mount to the board ground.
That will eliminate any leakage or hand/static effects ..mostly.
I dont remember that circuit going into oscillation but that op-amp is a very high freq one..so its possible.
wire dressing and use of sheilded wire might help too...altho its rare to
pick-up rf or static that way.

Best way to test the peltier is on a heatsink..with a small plate on top of the device..Proper polarity will be the small plate will get cold.
Use a digital thermometer and watch for temp difference ratio.
if you get at least a 30deg c difference I think its usable..
Dont go too far as to freeze the diode or get condensation on the window...hi.
Sounds like the one you have may run at a few volts up to an amp.
Start slow with a volt or two and see what happens..

Seeing that the peltier is isolating the diode from the mount..
I think I would get the TEC operational first so to cool the diode.

Sounds like your going to have some fun learning

[RedlumX]

Nice mount, where did you get that from?

The Peltier can easily take 1A or more, and if you limit the current to that, not much can happen (except if you interrupt the feedback loop). However be careful not to cool too much while driving the diode at the limit, because the max allowable current decreases with lower temperatures.
On the other hand, unless you really need a highly stable laser output, there is not much reason to TEC stabilize the diode anyway, it's not worth the trouble.

I am developing a TEC stabilized laser module with the same diode (no surprise in view of that ebay offer ;-) for holography applications, and so far just one died due to a surge in the power line. I regularly drive them to 175mA and they put out >110mW after collimator and so far none died while running it hard. 145mA should be completely safe and I guess there is indeed something wrong with your driver, there may be spikes or perhaps oscillations if it is not properly designed; a scope may come handy here.

[jeejeedr]

I'll have another try later on this week with these changes:
I'm going to hook up a scope to the driver circuit to see how it reacts in time.
I'm goind to stabilise the input power before feeding it to the driver.
I'll isolate the diode from the collimator housing. (I'll mount it on the peltier)
I'm goind to add that resistor as suggested by marconi.
I'll try to hook-up the peltier. However, using a thermystor and stabilise temp at about 25°C will be a challenge.

About the case, yes a nice thing. Collimates well too. I ripped it from an industrial barcode scanner. I make baggage sortation software for a living so we use a lot of SICK and other barcode scanners. (too bad they're only 10mW)

some more pics of the collimator, peltier and barcode scanner donor.



Collimator


peltier


complete barcode donor

[chw9999]

Hi, regarding multiple losses of diodes: lucerne (lucernelights.com) from the laserfreak forum (link to thread) buit a driver shutting off the diode power in case you short it out. No more killing of innocent diodes! :-)

Send him an email if you are interested in further details (he certainly speaks english).

Cheers
Christoph

[The_Doctor]

Today I killed a diode while trying to gauge the focal length of an AR coated lens I found, one of nine I never used due to longer FL and too wide a beam...

The lens was 20 feet downrange on a very well collimated red beam 2mm wide at that point. I was (foolishly) using a shiny cardboard surface I picked up as target without thinking first. When the focal point was found, at that instant enough light returned back along the 20 foot path to destroy the cavity on the other side of the room.

I'm still trying to work out whether this is cool, or depressing...

Is there such a thing as an 8th wave plate? I have this vague idea that if I could rotate the polar plane by 45°, the return journey adding another 45° might make sure that reflected light can't do this. Is there a cheap, efficient way to make it so retroreflection is not fatal?