TCA9535: Grounding A1 causes overheating

Vincent Castellano

Part Number: TCA9535

I have been having great success using the TCA9535 (TSSOP-24 package) as I/O expanders for the Raspberry Pi but I've isolated one peculiar problem.

When the A1 address line is jumpered directly to ground the device will overheat. I haven't characterized the specifics of the current rate but it takes a matter of hours for the temperature buildup to reach the point where the device is permanently damaged.

I have several PCBs with DIP-24 sockets to accept the 9535 on a breakout board, each PCB set to a different I2C address. I can swap in different 9535's and they all exhibit the problem on the "grounded A1" boards, but behave fine on a board where the A1 pin is held high. To seal the deal, I reworked a board so A1 is connected to +5V and it works fine.

I haven't tried grounding A1 with a resistor. There seems to be no difference whether I'm just running the device standalone (with no ports or I2C lines connected) or when it's operational (device is recognized by the bus master and all ports are connected and commanded to output mode).

I'm moving on since having only four addresses available is plenty but I'd like to know what I might be doing wrong.

"Thanks in advance for any insight you can provide."

over 3 years ago

0 Tyler Townsend over 3 years ago

TI__Mastermind 47998 points

Hi Vincent,

Do you have a schematic available? What is supply voltage (VCC) are you operating the device at?

Regards,

-Tyler

0 Vincent Castellano over 3 years ago

Prodigy 10 points

Here is the schematic: http://bulk.wgvc.com/PiBus.pdf

(There are not actually two of each device, I have both SMT and SMT footprints so I can choose to use either a DIP-adapter breakout or the actual SMD when I assemble.)

Vcc is 5V and lots of other devices are happy with it including the MAX5825 and the Raspberry Pi.

Thanks for looking.

0 Tyler Townsend over 3 years ago in reply to Vincent Castellano

TI__Mastermind 47998 points

Hi Vincent,

Do you have a screenshot instead? The website you have attached is restricted access.

Regards,

-Tyler

0 Tyler Townsend over 3 years ago in reply to Vincent Castellano

TI__Mastermind 47998 points

Hi Vincent,

I was able to access the link. Looking at the schematic I notice you have two TCA9535's. Do you see both of them fail overtime, or does only one fail at a time? And if so, what one on the schematic specifically fails?

Regards,

-Tyler

0 Vincent Castellano over 3 years ago in reply to Tyler Townsend

Prodigy 10 points

The two devices have different footprints (which you can't tell from the schematic), only one is ever connected. They each have identical connections.

0 Tyler Townsend over 3 years ago in reply to Vincent Castellano

TI__Mastermind 47998 points

Vincent,

I would mention that the total current sourced by all I/O's must be limited to 160mA, (80mA for P07-P00 and 80mA for P17-P10) but it seems that this problem occurs even when the I/O's are not in use. The schematic seems to be fine. I don't see a problem with schematic connections at this point. The only thing I could think of would be a solder bridge connection that accidently shorts pins on the device on the PCB. Another thought might be the trace width on the layout but I highly doubt this because the address pins have very low leakage current. I ordered some samples for this device, and I will try testing the ambient temperature over time. I'll try to recreate the problem, and will get back with you once I retrieve results from lab.

Regards,

-Tyler

Interface

Interface forum

TCA9535: Grounding A1 causes overheating