broken TOF sensor vl53l0x

[Jvan.10]

I am using the vl53l0x tof sensor in one of my products but I am currently experiencing a huge amount of fallout on this product.

On a batch of 30, I had 10 broken VLX53l0x TOF sensors after soldering.

What occured, the sensor started to pull down itself automatically on the reset pin.

My PCB assembler told he used the recommended solder profile for placement of this part.

All of this is done in an ESD safe environment.

The shock precaution, the PCBs that worked I put on a wafer applying 70G of force to them trying to get them to fail but all of the 20 leftover PCB  keeps working. We also think the hasn't been any of the components reels that had been dropped on the floor.

Moisture sensitivity, all of the sensor were in between the boundries as described in IPC/JEDEC JSTD-020-C during assembly. After that we tried again with a dry room for 24Hours at 125degrees Celcius.

We ordered this component at mouser electronics, so we assume that's also good.

The TOF sensors that are were working from the beginning are still functional.

We have no idea what is going wrong and why the components were failing. Does anyone know what is going wrong?

[John E KVAM]

The only thing you can do wrong with this chip is stave it of power or not put large enough pull-ups on the I2C and interrupt lines.

The chip uses 20mA when ranging, but it turns the laser on at about a 30% duty cycle. So for very short bursts assume we need 60mAs.

I've seen power issues with very small power traces.

Could that be it?

And do try putting larger pull-ups on the lines. It's a cheap test and won't hurt anything.

[Jvan.10]

The weird thing, is not that the tof sensors arent working, because I have 20 working sensors on my boards. The problem is that some of the vl53l0x are broken after the soldering process. After we replace them with new sensors they are working. The problem is that if we are going into mass production we are affright that we are facing an fallout percentage of 60% because of the sensor failing during fabrication.

For your question about the design, I am using 1K8 resistors as pull ups and the trace width is 10 mil with a copper weight of 1 oz per Square feet. According to my calc that should be suffiecent for 1 Amp of current if the sensor needs it.

I hope you you could help us with an solution that can garantee that the sensor doesnt break during soldering.

[John E KVAM]

We have shipped these sensors in the multi-millions to cellphone companies, and had almost no fallout.

That being said, the companies that are building millions of phones are pretty careful about their manufacturing tolerances and although we sell them in vast quantities, there are a relatively few designs.

We put a solder-flow temp recommendation in the data-sheet, but I've seen some pretty wild excursions from that with little fallout.

Every time I've taken a 'faulty' chip back it always works on our test jig.

The last time I got a complaint like this the problem was solved after he went with a 4-layer design and beefed up his power rails. But he didn't tell me what else he did - if anything.

[Jvan.10]

Dear John,

I tried to beef up the power Rails with a parallel extra wire and even tried adding some 18 pF bypass capacitor on the vcc rail, still no positive results.

I shall make an small report with pictures of the board , wire routings, scematics and X ray pictures the the designed system so you will have more insight of the current problem.

Sincerely,

Jim