Re: How to identify pins on telecom laser?
Wolfram, Sat Jun 10 2006, 02:49PM

According to this page output power is 2mw at 1310nm. Max current is 30mA.

How many do you have? Would opening one up to take a look inside be an alternative?

Re: How to identify pins on telecom laser?
The Wumpus, Sat Jun 10 2006, 10:41PM

Many thanks for that, now at least I know how much power to feed it! I have four of this specific type, I might have a go at dissecting one, though I'm not sure if I could make much sense out of it when open

Re: How to identify pins on telecom laser?
..., Sun Jun 11 2006, 04:51AM

My best guess would be that the pins connected to the case are the + of the laser (that is pretty much the standard)

From there, I would carefully apply power to each of the other pins... Use like 1v at 3ma to see which ones are connected to the case (I would expect the laser and possible the sense). This narrows down the number of possible pins for the laser.

Now use a supply limited at 30ma (or add a resistor to limit current around there) and carefully apply voltage to the remaining pins. One of them should show very little current draw up to a volt or 2, then should suddenly draw the full 30ma over a range of few tenths of a volt. This is the laser XD

From there... Try taking a volt meter from the case to the remaining pins, one of them should read a voltage proportional to the output power...

Once you have the laser/monitor... Measuring the power using the voltmeter slowly give it more current, and find where the max output occurs (it should be around or below 30ma if the datasheet is right). Back of the current to give 10% less output power (just to be safe) and record that number. Always use a power supply rated for that. Be careful though, the temp sensor might also be attached to the case. The power sensor will have a very fast response to the to diode current, while the temp sensor will have a much slower output. Don't mix it up, or you will just keep adding power until you blow up the diode

At this point, I would expect there to be a tec and a temp sensor. The tec will act like a normal resistor... But the temp sensor will too, the difference will be the value. The tec should draw about an amp at a volt or 2, while the temp will only draw a few ma it that. If you find both, try monitoring the resistance of the temp sensor while giving the tec some power... You should see the temp sensor's resistance vary with tec power. Now remember that the temp sensor will *probably* read a resistance inversely to temp, so as the resistance drops you are heating the diode...

I would say with a 99.9% accuracy that doesn't have a modulator (my dad grows the lasers for a living, so I have seen more than my share of them).

From there, you can find the optimal temp for the laser...
Hook up a voltmeter on the power sensor and give it a regulated current (the amount doesn't really matter, go for the value you get earlier) and put a ohmmeter across the temp sensor. Then in theory, you should see the resistance across the temp sensor slowly dropping (if it is slowly rising them you got a weird temp sensor, and its resistance is directly proportional to the temp... so reverse everything about the tec accordingly). Now give the tec a little power... You should see the temp sensor's rate of change differ... If it slows down or reverses you found the direction to cool the diode. If it goes faster it is the direction that heats the diode... Now look back at the meter and heat the diode a little. If the power decreases then cool the diode. Mess with it until you find the max power... Now record the value of the temp sensor, this is where you are supposed to run the diode.


So at this point you should have a pretty complete pinout, the operating current/voltage of the diode, the operating temp of the diode (if you want the value in a real number like degrees C, power everything down and let it sit for a few hours to normalise, then slowly heat/cool the module until you get the same reading for the temp sensor), and the relative power output. What more could you want?

A few things...

I doubt that it is only 2mw, since it is at an eyesafe wavelength and rated class IIIB... Seems like you should get at least 5mw Also that case should be able to dissipate a few watts, so unless that is just the power of the tec... I would also bolt it down to a good sized heat sink...

I have a calibrated power meter that has the same connector if you want me to calibrate one of your lasers... From there you could create a little adapter to couple one into a phototdiode as a power meter... Or I could just do all of them I have to warn you, it reads in DB, so I don't know how that would convert to mw...


Have fun and be safe!

Re: How to identify pins on telecom laser?
The Wumpus, Sun Jun 11 2006, 02:14PM

Thanks for the detailed instructions, I will try to identify the pins this way. I'm just using a photodiode and voltmeter to see if anything is coming out of the fiber, I have no way of measuring the power.

Re: How to identify pins on telecom laser?
Steve Conner, Sun Jun 11 2006, 04:01PM

There is no standard regarding which end of the laser diode is connected to the case. Some are grounded anode, some grounded cathode, and some floating. I design drivers for telecom lasers in my day job, and this causes a huge headache.

The temperature sensor will almost certainly be a "10k NTC" thermistor. It measures 10k at 25ºC and decreases as it gets hotter.

The TEC will measure as a very low resistance (an ohm or two) between two pins.

There won't be a modulator in the case, but there may be a "Bias tee". This is a little inductor and DC blocking cap, to allow the DC drive current and the modulation to be fed in on different pins.

Be careful what you use to buzz the circuit out. I heard that the ohms or diode test range on some meters can kick out enough voltage to destroy the laser diode if you reverse bias it, although I've never destroyed one this way myself.

The IR sensitive cards used for testing TV remotes should show the output of a laser like this. A digital or TV camera probably won't, though. This is a low powered 1310nm laser (a milliwatt or two) and it's really pretty boring.

... I guess your power meter ought to read in dBm, not dB. (unless you pressed the "Relative" button some time.) That means dB relative to 1mW, so 0dBm is 1mW, +3dBm is 2mW, -3dBm is 0.5mW etc. If you ever see it reading +60dBm, you're about to get lasered in half Goldfinger style.

Re: How to identify pins on telecom laser?
The Wumpus, Sun Jun 11 2006, 05:38PM


Since it is fiber-coupled and has standard connectors, I was thinking of using it for controlling high-voltage contraptions such as tesla coils. Kind of like a smart chicken stick. I use my 500mw, 750mw and 1W diodes for the less boring stuff .

Re: How to identify pins on telecom laser?
..., Sun Jun 11 2006, 07:16PM

The meter is designed for measuring the losses across a fiber, but it does have a dBm mode And it can handle up to 30dBm

The only way to be sure of everything would be to carefully (I am thinking dremel tool with a cutoff weel or a little carbide tip) remove the lid and trace out the connections... If you post a detailed pic we can probably figure it all out for you...

Re: How to identify pins on telecom laser?
Steve Conner, Sun Jun 11 2006, 08:13PM

Measuring the losses across a fiber? We use an OTDR to do that.

Wumpus, if it helps, the package of that laser is called a "14-Pin DIL" iirc, and the pin-out is AFAIK more or less standard (see attached diagram)

The optical connector is called a "FC/PC" (or "Flattie" for short ) and the fibre will be singlemode.



If you're just using it as part of a fibre optic link, you might even get away without driving the TEC.

Re: How to identify pins on telecom laser?
The Wumpus, Sun Jun 11 2006, 10:52PM

Many thanks for that pinout! I fed 30 mA to the laser diode pins and I got a voltage across the PD, so it seems to be working. However, the light does not show up on my digital camera CCD (probably it is completely blocked by the IR filter), so I have yet to verify that something is coming out of the fiber. I'll try some photodiodes...

And yes, I don't think I will bother driving the TEC, since I just plan on using this as a "very high insulation optocoupler"

Re: How to identify pins on telecom laser?
Steve Conner, Mon Jun 12 2006, 09:09AM

I forgot to say, you need an InGaAs or germanium photodiode. Ordinary silicon ones are barely sensitive at all to 1310nm. These photodiodes are somewhat pricey, we pay about $70 each.

Also, my boss says that some of the IR testing cards may not work at this wavelength (but others do)