LEDs and lasers get way more efficient when they are cold. But their junction drop increases too. So it seems that driving them from some magical Thevenin source would temperature compensate the light output. Has anyone done this? I need to hold some dBm level from a laser source from -40 to +85C and that might be a simple way to do it.
temperature compensating a laser diode
Started by ●October 16, 2023
Reply by ●October 16, 20232023-10-16
On Monday, October 16, 2023 at 11:41:58 AM UTC-5, John Larkin wrote:> LEDs and lasers get way more efficient when they are cold. But their > junction drop increases too. So it seems that driving them from some > magical Thevenin source would temperature compensate the light output. > > Has anyone done this? > > I need to hold some dBm level from a laser source from -40 to +85C and > that might be a simple way to do it.With LEDs, yes. With lasers, no. I ran an experiment back in the early 2000s to do exactly that. I found that, if I held the voltage constant across the diode the current would automagically increase as the LED heated, thus somewhat compensating for the heating. The light output remained fairly constant. The data is lost and the experimental test equipment has been lost as well. I wish I could help you more, but at this point I can only encourage you to proceed with any research you have in mind. I wish you well. John Smiht
Reply by ●October 16, 20232023-10-16
John Larkin <jl@997PotHill.com> wrote:> LEDs and lasers get way more efficient when they are cold. But their > junction drop increases too. So it seems that driving them from some > magical Thevenin source would temperature compensate the light output. > > Has anyone done this? > > I need to hold some dBm level from a laser source from -40 to +85C and > that might be a simple way to do it. > >Wouldn’t recommend it in general. There are a number of not-too-well-understood instabilities to worry about. Runaway and mode hopping are the most common, but there are also some bistable situations, especially with VCSELs. Most diode lasers are tanking pretty quickly at 85C. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics
Reply by ●October 16, 20232023-10-16
On Mon, 16 Oct 2023 22:34:24 -0000 (UTC), Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>John Larkin <jl@997PotHill.com> wrote: >> LEDs and lasers get way more efficient when they are cold. But their >> junction drop increases too. So it seems that driving them from some >> magical Thevenin source would temperature compensate the light output. >> >> Has anyone done this? >> >> I need to hold some dBm level from a laser source from -40 to +85C and >> that might be a simple way to do it. >> >> > >Wouldn�t recommend it in general. There are a number of >not-too-well-understood instabilities to worry about. Runaway and mode >hopping are the most common, but there are also some bistable situations, >especially with VCSELs. > >Most diode lasers are tanking pretty quickly at 85C. > >Cheers > >Phil HobbsYeah, I need to talk the customer down on his temperature range. My gadget will go on his board and I bet his stuff won't work from -40 to 85. I could do a more scientific temperature comp, a sensor and a programmed current source, but that's not as cute. This will be data over multimode fiber, nothing fancy.
Reply by ●October 16, 20232023-10-16
On Mon, 16 Oct 2023 10:41:32 -0700 (PDT), John Smiht <utube.jocjo@xoxy.net> wrote:>On Monday, October 16, 2023 at 11:41:58?AM UTC-5, John Larkin wrote: >> LEDs and lasers get way more efficient when they are cold. But their >> junction drop increases too. So it seems that driving them from some >> magical Thevenin source would temperature compensate the light output. >> >> Has anyone done this? >> >> I need to hold some dBm level from a laser source from -40 to +85C and >> that might be a simple way to do it. > >With LEDs, yes. With lasers, no. >I ran an experiment back in the early 2000s to do exactly that. I found that, if I held the voltage constant across the diode the current would automagically increase as the LED heated, thus somewhat compensating for the heating. The light output remained fairly constant. The data is lost and the experimental test equipment has been lost as well. I wish I could help you more, but at this point I can only encourage you to proceed with any research you have in mind. I wish you well. > >John SmihtSomething like this? https://www.dropbox.com/scl/fi/q06fupx4nrhglpw1k55ip/Lx_Laser_Driver_1.jpg?rlkey=otwsvb9dmtvkgxhhw01c3rcbs&raw=1 If Rd is biggish, I can temperature compensate Vd and not try to get too cute. That LVDS receiver has a typ rise/fall time skew of 100 ps, and I can tweak R1 and R2 to null that to nominal zero. The phemt threshold will be maybe 0.4 or so. I suppose the LVDS receiver could drive the laser diode through just a resistor but that leaves less knobs to turn.
Reply by ●October 17, 20232023-10-17
On Mon, 16 Oct 2023 19:23:05 -0700, John Larkin <jl@997PotHill.com> wrote:>On Mon, 16 Oct 2023 22:34:24 -0000 (UTC), Phil Hobbs ><pcdhSpamMeSenseless@electrooptical.net> wrote: > >>John Larkin <jl@997PotHill.com> wrote: >>> LEDs and lasers get way more efficient when they are cold. But their >>> junction drop increases too. So it seems that driving them from some >>> magical Thevenin source would temperature compensate the light output. >>> >>> Has anyone done this? >>> >>> I need to hold some dBm level from a laser source from -40 to +85C and >>> that might be a simple way to do it. >>> >>> >> >>Wouldn�t recommend it in general. There are a number of >>not-too-well-understood instabilities to worry about. Runaway and mode >>hopping are the most common, but there are also some bistable situations, >>especially with VCSELs. >> >>Most diode lasers are tanking pretty quickly at 85C. >> >>Cheers >> >>Phil Hobbs > >Yeah, I need to talk the customer down on his temperature range. My >gadget will go on his board and I bet his stuff won't work from -40 to >85. > >I could do a more scientific temperature comp, a sensor and a >programmed current source, but that's not as cute. > >This will be data over multimode fiber, nothing fancy.Have you checked the ageing characteristics for both the light source and the receiver ? Use sufficient margins when new so that the link will work also after a few years.
