Inexpensive ESD (static) protection.

[Markus Unread]

Inexpensive ESD protection.

If you are good at soldering very small things, read on!

The first place you should look when thinking about ESD protection for your laser diodes (LDs) is in the TVS category. TVS's are designed for ESD and transient protection. The standoff voltage of a TVS is the voltage that, with some variation, falls below the breakdown voltage (Vbr) of the device. The "head room" is the difference the two ratings. TVS headroom is often set to accommodate the world of digital logic which is a bit more forgiving when it comes to transients. This is why I selected TVS's by their minimum breakdown voltage instead. The minimum breakdown voltage is basically their zener voltage. But TVS's trade toughness for accuracy so their Vbr isn't as precise as a zener diode or voltage reference part.

When you look into these parts, download the data sheets and look both at the specs and the pinouts. Unfortunately, for the SOT parts, they aren't standardized for polarity. I connect the two diodes in the SOT package in parallel to provide a small amount of reverse polarity protection. If you are worried about reverse voltage, throw a Schottky diode in between your LD and power supply (any rectifier, but Schottky has the lower voltage drop and will dissipate less heat at high current). The TVS's below are UNIDIRECTIONAL and conduct at ~0.7V if voltage is applied backwards. If you don't want to care about the polarity of the part, there are bidirectional versions of these parts.

If possible, check your LD voltage at the current that you intend to run - IMPORTANT! Measure when the LD is COLD! The voltage goes down as the LD heats up. This will help you determine which voltage of TVS to use.

Here are a few examples. There are a LOT more choices, but I did a search on what was in stock at Digikey.com

For Nichia 445nm:
NXP PESD5V0U1UT,215 ... 568-4058-1-ND ............ 7.0V $0.53
NXP PESD3V3U1UT,215 .... 568-4047-1-ND .............. 5.8V $0.53 (1)

For BluRay 405nm
Vishay VESD08-02V-GS08 751-1438-1-ND ............. 10.2V $0.42
NXP PESD5V0U1UT,215 _ _ 568-4058-1-ND _ ._ ._ 7.0V $0.53 (2)
MCC SMCJ6.5A-TP _ _ _ _ SMCJ6.5A-TPMSCT-ND _ 7.2V $0.76 (2b)

(1) I measured my 445nm at 4.7V for 1250ma, so 5.8Vbr would work.
(2) I used this on a 405nm LD running at 200ma. Check your LD voltage at the current that you intend to run - IMPORTANT! Measure when the LD is COLD! The voltage goes down as the LD heats up.
(2b) If you want more "head room" this might be better.

Testing a SOT TVS - 30ga wires attached. In the case of the TVS's for 445nm and 405nm LD's, the side of the TVS with the two legs is the plus (+) side. The single pin side attaches to the minus (-).

TVS soldered onto a 405nm LD in an Aixiz module.

Wires on.

Shrunk up.

 

[LazyBeam]

Excellent info... +1

Wow, this device takes up almost NO room at all. It could be used in even the most compact builds. Can it be placed across the driver outputs as well? That way you only have to heat the LD once in stead of twice. The less the LD is handled and heated the better.

How well does this work in comparison to a bridge Capacitor & resistor?
How well does this work in comparison to a Lasorb?

 

[HIMNL9]

You know that, for ESD protection, is enough to solder an 1Kohm SMD resistor between the pins, right ? .....

 

[FireMyLaser]

I use a 5V zener diode for ESD protection. I don't know how good of a protection it is, but not even a piezoelectric igniter from a lighter could kill two 5mW red diodes connected in series in series (Vf ~4V.) I say that was successful.

 

[Flaminpyro]

The very first thing you should look at is reducing stray static in the area of work and this can simply be done with a can of anti static spray, it realy is that simple
Get a can and keep it at your work area, when you get ready to mess with bare laser diodes spray the area where you are working, your chair, the floor under you and your self some.
This $4.00 can of spray will go a long way to help preventing you from loosing a $50.00 diode to static damage

Peace All...

 

[LazyBeam]

I like antistatic brushes, personally.

 

[Benm]

You know that, for ESD protection, is enough to solder an 1Kohm SMD resistor between the pins, right ? .....

I usually do that, or a 1n4148 diode in opposite polarity of the laser diode. Using the resistor makes the LD a nicer load for current sources too, as it will always draw a bit of current, even when far below the lasers operation voltage. It does throw of precies current readings though, but who cares about those few mA?

The main ESD danger lies in too high reverse voltage across the laser diode. In its conducting direction the diode will dissipate the charge itself.

 

[Markus Unread]

I usually do that, or a 1n4148 diode in opposite polarity of the laser diode. Using the resistor makes the LD a nicer load for current sources too, as it will always draw a bit of current, even when far below the lasers operation voltage. It does throw of precies current readings though, but who cares about those few mA?

The main ESD danger lies in too high reverse voltage across the laser diode. In its conducting direction the diode will dissipate the charge itself.

Protection from reverse bias is good, but depending on the structure of the LD itself, an ESD pulse in the forward direction can damage the device before it can start to conduct. That's why a TVS is like a fast response version of a zener - also structurally a bit more robust. I'd use a zener if there was no TVS available.

Anyway, I'll get back to the top and answer some questions when I get back later. In the mean time, have a look at this -

http://www.conformity.com/PDFs/0809/0809_F3.pdf

Even though it focuses on one part of the problem, the scope shots of the tests are pretty representative. And yes, if you can use a wrist strap and an ESD proof work surface, you are golden. Depending on the host, once surrounded by a Faraday cage of solid aluminum, things should be pretty much bullet proof.

One of these days I'm going to take a look at the startup output of some cheap boost power supplies (like the O-Like) and see what surprises turn up

Later,
Mu

 

[Markus Unread]

I like antistatic brushes, personally.

Holy crap! A radiation source!
Yeah that, uhm... yeah.

I've always been wary about radiation above UV...

 

[LazyBeam]

Alpha emitter... not a big deal.

 

[Markus Unread]

Excellent info... +1

Wow, this device takes up almost NO room at all. It could be used in even the most compact builds. Can it be placed across the driver outputs as well? That way you only have to heat the LD once in stead of twice. The less the LD is handled and heated the better.

How well does this work in comparison to a bridge Capacitor & resistor?
How well does this work in comparison to a Lasorb?

It really depends on how much the laser module is going to get handled.
If you are working out a host design and fiddling with stuff constantly, then I'd stick with the TVS down at the diode. The farther from the diode, the less effective it'll be.

A capacitor across it can't hurt, but may not help either. The ESR & ESL of a cap limits how well it deals with ESD and other transients.

Depending on the environment (low humidity winter, synthetic winter clothes) the ESD jolt can cause a huge overvoltage into a virtual short. So resistor/capacitor fixes are a good supplement to signal lines that go in and out of boxes, but not so much for low impedance power lines.
[See the PDF I referenced before]

Lazorbs. They're made for this, and I'm sure that there are TVS's in them, but they are pretty expensive. It's hard to know.
Here's a URL for the "Blue" diode Lazorb.
http://www.lasorb.com/LASORB_L44-683-X.pdf

They have done a bunch of testing, but I'd rather spec the Vbr to be closer to the operating voltage of the part. I think 7.5V is awfully high for a 445nm diode.

Hm.. Did I answer your questions?

 

[HaloBlu]

Too bad those antistatic brushes are so costly. As long as you don't eat an alpha emitter (& can't inhale it) your safe. Some brushes are quite tasty, just not that one

The lil TVS looks right at home on your diode there :san:

 

[Benm]

What is needed depends on conditions indeed. If you handle your components carefully and take minimum precautions, the risk of ESD is not very large for things like laser diodes.

Obviously you could slide across the carpet on a dry day, holding the laser diode by its body and then touch the pins on a radiator and the end of your slide, in which case the TVS might save it, while a resistor definitely will not

 

[jib77]

How about multilayer varistors? Ive seen them used lately in lue of TVS's.

 

[Markus Unread]

How about multilayer varistors? Ive seen them used lately in lue of TVS's.

They have a slower response time and a larger gap between the standoff and breakdown voltage. They generally don't spec the range of the breakdown voltage for any given part. Whereas, the TVS's might say 5.8min,6.0nom,6.2max in the spec. This gives the designer a much better idea as to how the part will perform in real life.

I like straight up MOV's for power supply rail limiting. If it's a switching supply with overload shutdown, I then go for the crowbar. Now I'm getting off topic.

 

[Pangolin]

Hehe.

How many people have tested the solutions being proposed here in the real world?

Common ESD devices made to protect semiconductor products won't work for Laser Diodes. If you want to understand why, see this page:
LASORB - ESD absorber for laser diodes

We have tested TVS as well as multi-layer varistors and other approaches. They won't work for laser diodes. LASORB is the only thing proven to be 100% effective at preventing laser diode failure caused by ESD and power surges. And it's only $2 if purchased in large quantities...

There is a wealth of information on the LASORB web site, including videos showing how well it works, and pictures showing our test bench.

Best regards,

William Benner