Anybody round here got a number for the typical inductance of a butterfly-package laser mounted on a board with good lead dress? A friend of mine is having trouble slowing his laser driver down below 500 ps, even with 10-ohm ferrite beads. Next thing to try is a cap after the beads, but some reasonable estimate would help. 20 nH or so, maybe?
Thanks
Phil Hobbs
We TDRd a couple of butterfly lasers and didn't see any inductance...
50 ohms all the way to the laser chip, which looks capacitive.Try bigger beads!
Some of these lasers make an optical step that's quite a bit faster than the electrical input. They store up energy and go bang.
Lasers have a bunch of lags in the loop--the driver has to charge up their capacitance till they turn on, then the carrier density (population inversion) has to build up till they lase, then the lower state population starts to increase, which decreases the gain, and so on. Some systems such as nitrogen lasers don't have a stable operating regime--you get nanosecond pulses out of an N2 laser no matter what you do. (Of course you can also make it not lase at all, but you know what I mean.)
Cheers
Phil Hobbs
Certainly. I have extensive experience with making lasers not lase.
Or turning a $1k laser diode into a $0.10 LED :-)
I'll never warm up to those butterfly packages. Yeah, they may be designed to 50ohm but it doesn't make sense. The laser diode is usually in the low single digit ohms and in order to drive it with gusto the other side has to be even less. So no match.
We have an old HP 8146 that has out lived its usefulness at work.. THey told me I can have it, I don't know what use I would have for it at home base.
Jamie
Right. The telecom boys use 50 ohm distributed amps and bias tees to pump NRZ data into these things. In pulse apps, the drivers need to be closer to 1 ohm than 50, so the matching isn't right. But it's still better than having a bunch of series inductance. I want to put the tail end of my driver circuit inside the butterfly, snuggled up against the laser chip.
The butterflies are sort of terrible packages, and the pinouts aren't standardized. The dimensions (pin diameter, height above the baseplate) aren't either. We're making drivers for a bunch of circa $2K lasers, and luckily haven't blown up any. Yet.
We're getting around 100 ps wide light pulses out of this rig, *with* the pin sockets to the butterfly.
We did blow up close to $1000 worth of GaN fets one afternoon.
We pause this discussion to let JF and JT do their ritual whining....
Have something to add? Share your thoughts — no account required.
Ask the community — no account required