COM-15208 Buck-Boosts Killing themselves?

Hi there,
I’m currently building a GPS Lap tracker using a Raspberry Pi, going through what were good options to take inputs from both a battery bank so less than 5v and from directly wired to a 12V automotive source so (11-15V) we decided that the COM-15208 was our best option. So far in our testing they’ve worked great however, seems boards are starting to kill themselves, more specifically the TI tps63070 is melting in the VIN side around the [PS/SYNC, EN, VSEL] pins, Not entirely sure why this is happening, Voltage In is below the advertised 16V, the polarity is correct, and as far as I knew these had a thermal protection mode, so not entirely sure what is happening. Could it be that the current is too much to handle, and it’s killing itself from some sort of inrush? I am not an EE, so I am unsure as to why this wouldn’t work, and I am hoping perhaps someone could assist me into why these are failing in this way. Also, here are some photos on how these have failed spectacularly in three separate devices

(sorry couldn’t upload all three here, so I uploaded to Imgur) : failing PS/SYNC, EN, VSEL - Album on Imgur

If you are sure that the battery voltage is within the allowed level and polarity is also correct, maybe there is a fault in the hardware.

That was my guess, but ultimately don’t know what I’m not privy to, so was unsure if there was another problem in my design here or something. I would assume it wouldn’t draw more than it could handle, but perhaps I’m wrong there lol. Would suck if I had to individually test all 100+ power boards we have in inventory right now for these devices, I’ve had good luck with other sparkfun circuit designs, but this one is seemingly not working as well

Are you installing it in an operating vehicle? If so, your assumption that the voltage remains between 11-15V may be incorrect.

See this article.

1 Like

Did not expect those types of swings to be that variant, I had naively assumed it would peak at around ~15ish volts as observed from every system I’ve personally wired which is nearly in the hundreds now as it’s my job to do so, however, these that died were on a non charging 12V system that had registered at 11.5V at the time of hooking these devices up to the system. They had been working well on our two shop track cars (BRZ and a NB Miata), and all of a sudden the magic smoke decided to release, lol not sure what would cause this on a dormant 12V system.

How much current are you pulling? Might need a heatsink on it. Small Heatsink - PRT-11510 - SparkFun Electronics

1 Like

There’s a graph on the hookup guide page that shows the output current capacity vs the input voltage.

Around 1/2 an amp or so, using a Pi Zero 2 W with a simple SARA-R5 Hat for sensors that doesn’t really use a ton of power, that’s what’s throwing me for a loop, I’ll give a heat sink a try tho!

Yeah, I can see why you are concerned given the operating conditions you’ve described. It could be an in-rush event given the Pi Zero 2 W power requirements (5V and 2.5A available).

Looking closer at the datasheet for the TPS63070, it looks like it really is a 2A output, table on page 6. A bit concerned that the RPi board may be pulling to hard on this buck-boost.