• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 33 subscribers
  • Views 208 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TCA9535: TCA9535 Unused I/O Pullup/Pulldown needed?

Thomas Acquistapace
Prodigy 70 points
Part Number: TCA9535
Other Parts Discussed in Thread: TCA9555

Hello

we are using the TCA9535 and we have about 8 of the 16 pins as unused in our design. The unused pins to not have the Pullup or Pull down through 10k (GND or VCC). In our design we see the USED pins being driven ON at Power Up. Indicating that the device is powering up with its pins partially driven (seems to be weak drive) to a HIGH state. This behavior is of course not what the datasheet indicates (all pins are configured as INPUTs at power up). Can you tell us if we add the resistor pullup/down (10k) to these unused pins that the partially driven ON state on the USED pins will be alliiviated?

 I have attached our connectivity of the part in our application below

TCA9535 configuration 

We are using TCA9535RTWR

Pin 20 :	SDA	: 10K Pullup to +3.3V : Driven By FPGA (sink drive) after power up 
Pin 19 :	SDL	: 10K Pullup to +3.3V : Driven By FPGA (sink drive) after power up
Pin 22 :	INT	: 10K Pullup to +3.3V : No other connection
Pin 21 :	VCC	: Direct connecto +3.3V : FPGA uses same supply 
Pin 24,23,18:	A2:A0	: Connected to GND
Pin 9,EPAD:	GND:EPAD: Connected to GND

Pin 1:		P00	: Connected to LVTTL xcvr input pin (+3.3V). No other connection 
Pin 2:		P01	: Connected to LVTTL xcvr input pin (+3.3V). No other connection 
Pin 3:		P02	: Connected to 10K series R to 2N2222 BJT Base. A 100k base resistor to GND is also used. No other connections 
Pin 4:		P03	: Connected to parallel a 10k to GND. P03 drives a LVC logic input (+3.3V). No other connections  
Pin 5:		P04	: P04 drives a LVC logic input (+3.3V). No other connections
Pin 6,7,8:	P05:P07	: No connections 

Pin 10:		P10	: Drives Gate of 2N7002 N-Ch MOS
Pin 11:		P11	: Drives Gate of 2N7002 N-Ch MOS
Pin 12:		P12	: Drives Gate of 2N7002 N-Ch MOS
Pin 13-17	P13:P17	: No connections

Can you assist

Thanks, Tom

  • 0
    TI__Mastermind 49010 points

    Hi Tom,

    Are you able to do a read from the configuration register and output register to verify if the device is indeed powering up in an output state? 

    My guess at the moment is that the device may just have some leakage current which is giving you a some kind of voltage bias at those pins. Typically with unused p-ports, I recommend they be biased to either GND or Vcc with resistors OR set the unused p-ports to be outputs (High or Low doesn't matter) inorder to lower the supply current when the pins float or see noise (shoot through currents due the p-ports being left as inputs and partially turning on CMOS structures around mid rail). 

    -Bobby

  • 0 in reply to BOBBY
    Prodigy 70 points

    Hello Bobby

    I have not programmed my Xilinx Artix 7 series FPGA connected to TCA9535 yet. I am just powering on to examine the power up state of the pins connected to the TCA9535 and they drive devices which cannot have these pins set HIGH at power up. So, I see significant drive on the BJT base circuit (10k in series with the BJT base !!). I see the BJT base at over 0.6V and the BJT is turned ON. So, this means the TCA9535 is driving hard after a power up.

    The 3.3V rail for the FPGA and evertyhing else the TCA9535 touches at the same time. Same power plane.

    I understand that the datashet calls out that unused pins should be connected through 10k to GND or +Vcc.

    Can you tell me if this is the reason the TCA9535 comes up with the output driven Staong HIGH ?  This is not the info I have seen on the other forums which says that the device really does come up with pins all inputs (HIGH-Z). Our observations show the device coming up with some or all outputs HIGH.

    Now regarding your comment "...... inorder to lower the supply current when the pins float or see noise (shoot through currents due the p-ports being left as inputs and partially turning on CMOS structures around mid rail).

    Does this mean that the inputs can actually Latch Hard at a state near Vdd/2 ? If so, this is not what we would expect from a device which has its inputs set to High-Z at startup.

    -Tom

  • 0 in reply to Thomas Acquistapace
    Guru 242810 points

    The TCA9535 definitely should have Hi-Z pins at startup. Are you sure that you do not have a TCA9555 (which does have internal pull-ups)? What is the voltage at P03?

  • 0 in reply to Clemens Ladisch
    Prodigy 70 points

    Hello We check the device code and it is a TCA9535. The internal pullups of 100k in the 9555 does not expalin the hard drive at some voltage lower than the 3.3V rail. We are seeing signficant current  driving from the TCA8535 to several device (the NPN BJT base drive through series 10k, above). The voltage at P03 is  2.0 V after power up.

  • 0 in reply to Thomas Acquistapace
    Guru 242810 points

    2 V over 10 kΩ implies a current of 200 µA out of pin 4, or an internal pull-up resistor of 6.5 kΩ. This is not what a TCA9535 would normally do, so I'd have to guess that the device is damaged or fake, or that there is a solder bridge.

    There are ESD protection diodes at all pins; do you measure anything else when the chip is completely powered down?