Diode Drivers

Colin Kaminski · Post by **Colin Kaminski** » Wed Jul 03, 2002 10:57 am

I have been using the thorlabs constant current driver. It is a dream. The output is very flat. It has a port to measure diode current. Matched with a peltier chip I have been able to keep temperature drift down to one fringe. Not quite good enough for holography but I still have refinements to make.

I think a simple battery driven constant current driver would work just as well. Tom was saying a resistor and a battery would do well. I thhink it should be a little more complicated than that.

I have been thinking a simple bridge to limit the current with a 12 turn pot to adjust the current and Tom was thinking about a 1 ohm resistor in-line to be able to measure current with out placing an ohm meter in-line.

Would this work?

Tom B. · Post by **Tom B.** » Thu Jul 04, 2002 4:08 am

Which diode are you using? Over what time period do you get a one fringe drift? What is the variation in ambient temperature and of the diode?

My current lash up consists of an LNCQ05 diode on a big heatsink with 27 ohm 1/2W current limiting resistor connected to an old HP 6214A variable power supply. This supply seems very quiet and stable. I have a 1200 uF capacitor across the supply output, a transient
suppression diode (Tranzorb) between power and ground
at the diode/resistor end of the cable, and a 0.1 uF cap and reverse-biased Schottky power diode across the laser diode. I checked the photodiode current versus output power (measured with a Coherent LaserCheck) and found that this paricular diode had (in the absence of back-reflections) almost exactly 1 microamp of photodiode current per milliwatt, so I just use an old 0-100 uA meter movement to monitor power. To minimize risk of transient damage, my startup procedure is to first turn supply knob to minimum, then turn on the supply power switch and raise the voltage while monitoring output power. Shutdown is the reverse. Below 20C, the diode seems stable up to about 40 mW output.

Colin Kaminski · Post by **Colin Kaminski** » Fri Jul 05, 2002 12:15 pm

I am seeing classic hysterisis. It is about one fringe oscillation every 1 second. It corresponds exactly to the temperature. I have some ideas about how to make the thermal assembly better. My ambient seems irrelevant as long as there are no drafts. I am using the Mitsubishi 35mw diode and the thorlabs constant current driver. I am supplying about 80ma of power to the diode. I do not have the photodiode hooked up. I am running in-between mode hops at about 21C.

Tom B. · Post by **Tom B.** » Wed Jul 10, 2002 5:06 am

Just a thought - 1 Hz seems a bit fast to be thermal oscillation. You can get oscillatory behavior in control systems if you have a ground loop - i.e. if the reference ground voltage for the sensing circuitry is getting slightly disturbed by return current from the load. You might look carefully at your ground connections and think about what the effect of the load current will be on the ground voltage as seen at various points, keeping in mind that there will always be at least a few milliohms of resistance in the wires and connections. 1 Amp times a few milliohms is a few millivolts, which can be significant if it shows up in the wrong place. The usual aim is to provide a "single-point" ground - all ground connections run separately to a single common point and tied together rather than sharing return paths.

Colin Kaminski · Post by **Colin Kaminski** » Wed Jul 10, 2002 1:16 pm

I am pretty careful about ground loops. My biggest problem ground wise is I have the TEC running from a separate power supply from the one that powers the diode. The grounds are isolated from each other. I don't really know a way around this without building a custom supply.

I will finally have the parts to isolate the screws more effectively tonight. Hopefully I will get some better results.