ATtiny25 damaged after shipping

Hi guys,

I'm having a strange issue with some ATtiny25 SOIC chips programmed with Arduino IDE and flashed with a USBasp. I made a few kits, each kit involving 4 mcus.

These kits were designed for a specific car application (LED lightning) and I tested every circuit after assembly. Everything was OK on my side, on the bench.
Design is kinda simple... 4 boards per kit, one LM1117-5 and ATtiny25 on each board, transistors, LEDs etc... They are designed to work "in pair" (2 boards sandwiched together") sharing a common ground and having their own +12v input.

A while ago I sent the first kit from France to Germany. The other kits leaved for the US. Once arrived in Germany the guy there tested the boards before installing them in his car, and all boards were defective. He assured me he did respect polarities, and I'd like to trust him.

He sent the boards back to me and this is what I discovered:
Pair 1 :
PCB 1: ATtiny KO - not responding to USBasp/Arduino as ISP programmers
PCB 2: ATtiny KO - not responding to USBasp/Arduino as ISP programmers

Pair 2 :
PCB 3: ATtiny KO - not responding to USBasp/Arduino as ISP programmers
PCB 4: ATtiny KO - overheating and shorted (didn't got on fire I guess thanks to the LM1117 overcurrent protection)

On PCB 4 a high power LED was also fried and I had to replace it. That leads me to say that particular board was shorted or polarity was reversed one way or another.

So I sent my own boards to that guy, and started repairing his boards for myself.

Then he received the 2nd kit and again... They are defective:
Pair 1
PCB 1 - Working
PCB 2 - KO

Pair 2
PCB 3 - KO
PCB 4 - Working

What the hell could fry an ATtiny like this? The LM1117-5 have their decoupling capacitors on the input and output. Also I tested my own board on my own car and it works, being the same car I doubt the car itself could be the root cause.
I also doubt French or German posts use gamma rays on parcels. X-rays maybe but would this be a problem anyway? We all order stuff from China and items usually work OK :wink:
All boards were sealed in antistatic bags and protected with paper towel for shipping and they don't look physically damaged.

Did anyone here have a similar issue while shipping programmed chips by mail?

Can you give a full schematic please, also, did you put a diode on the input to protect against reverse polarity?

What decoupling caps do you use? The LM1117 needs some ESR (0.3 Ohm minimum) on the output cap. I heard using too low ESR cap (i.e. ceramic) may lead to poor function - most notably large voltage overshoot at power up. The guy that said this discovered it by killing his MCUs - they died and appeared as short circuit when tested by a DMM. It looks similar.

I have shipped over a hundred PCBs with attinys mounted on them, packed in an antistatic bag inside a bubble mailer. I have never had a board arrive DoA. So it's something you're doing or something about how he's using them.

Is that being powered off car's +12v? Have you taken appropriate actions to deal with the appallingly noisy power source? Automobile +12v is notoriously noisy due to the spark coil, motors for pumps and fans, etc. It requires special consideration to protect your circuit, and will often damage electronics designed to work with a nice clean 12v from a mains-powered adapter.

In addition to what DrAzzy says, automotive electrical systems produce occasional voltage spikes of several hundred volts. TVS diodes and clamping diodes can be used to protect against those.

Wow I didn't see your answers! Thank you guys :slight_smile:

proto-pic:
Can you give a full schematic please, also, did you put a diode on the input to protect against reverse polarity?

I don't have the schematic anymore but I could redraw one. There's no diode because there's no room where the boards are installed.

Smajdalf:
What decoupling caps do you use? The LM1117 needs some ESR (0.3 Ohm minimum) on the output cap. I heard using too low ESR cap (i.e. ceramic) may lead to poor function - most notably large voltage overshoot at power up. The guy that said this discovered it by killing his MCUs - they died and appeared as short circuit when tested by a DMM. It looks similar.

Very good question... While I was planning on using a specific reference I forgot to order it and used another one instead. I could test them with a component tester. There are 10µF anyway.

DrAzzy:
I have shipped over a hundred PCBs with attinys mounted on them, packed in an antistatic bag inside a bubble mailer. I have never had a board arrive DoA. So it's something you're doing or something about how he's using them.

Is that being powered off car's +12v? Have you taken appropriate actions to deal with the appallingly noisy power source? Automobile +12v is notoriously noisy due to the spark coil, motors for pumps and fans, etc. It requires special consideration to protect your circuit, and will often damage electronics designed to work with a nice clean 12v from a mains-powered adapter.

jremington:
In addition to what DrAzzy says, automotive electrical systems produce occasional voltage spikes of several hundred volts. TVS diodes and clamping diodes can be used to protect against those.

I decided to install the repaired kit in my car and it worked as expected.
The first board is powered by the turn signals relay's output, the 2nd board get's its power from a RUN out of the ignition switch. Thing is, I have used bare Arduinos in these cars more than once without any issue. I know the voltage isn't clean but it's OK. There are capacitors on the coils circuit. Moreover I've taken some OEM accessories appart already and I can assure you they are not protected in anyway aside a fancy low dropout 12v regulator and a capacitor. Which I couldn't implement because of clearance issue.

I guess he did something wrong. He also has "mirror see-through turn signals". Maybe they interfere with my devices. I'll wait for the US guys to install their devices on their own. We'll see.
Having mine working today made me feel a lot better and confident :wink:

I can assure you they are not protected in anyway

I can assure you that manufacturers of quality equipment intended for an automotive electrical environment DO protect their circuitry against voltage reversals, brownouts and high voltage spikes.

You simply haven't read enough on the topic, or looked hard enough at quality equipment. I strongly recommend that you read the material in the link I posted in reply #4.

You are asking for trouble if you manufacture equipment that fails to take this very serious threat into account.