TS5A3160: issue with the bidirectional nature of the IC

Den,

What is driving your 3.3 V rail? If the TS5A3160 V+ rail is held at 0V and not floating the path from NC to COM should be Hi-Z. I suspect that the V+ node is floating in which the 5V rail in your system is able to pass from NC to COM and place 1.5V on the 3.3V node.

The TS5A3160 V+ min of 1.65 is what is guaranteed but as in most semiconductor specs there is some margin to make sure the spec holds true due to temperature and manufacturing variations.

Why do you need to have COM connect to 5V when your system is off?
You could use a 1:1 switch instead of a 2:1 switch like the TS5A3166 device. This would eliminate the 5V connection NC to COM.

Thank you,
Adam

Hi Adam,

thank you for your reply.

The 3.3V are driven by another IC. But it is turned off by the time all this mess is happening. It is basically unimportant for analysis of this situation. Well the datasheet says it should be enough to have 1.5V to turn it off, I have no idea where 1.5V are coming from, but the IC should be off and if I can't rely on the datasheet, what can I rely on? (it is a bit ridiculous, don't you think?). Also, why 1.5V? I seriously played with all resistors and the 1.5V still persist. Do you have two diodes inside of the TS5A3160? ...

"Why do you need to have COM connect to 5V when your system is off? " - good question, why would I? I wouldn't - that's the problem. MTO pin of the U3 is an Open Drain, that's why I have a pull-up there. Basically when I turn the system off, i.e. the 3.3V, the TS5A3160 doesn't turn off completely and the 5V manage to get to the 3.3V rail through the COM port - this is a bug, not a feature :) . Well I guess this circuit can't work cause of TS5A3160 timing issues. I will have to find another IC. 

Now using 1:1 kills the whole idea of being able to use the button for two purposes .... no?

Kind Regards,

Den

Den,

The datasheet guarantees performance stated in the electrical specifications table.  With V+ = 1.5V it is outside the recommended operating conditions of the device and will not guarantee the performance stated in the electrical specifications table.  

The datasheet states the device will be high impedance (leakage <10nA) when V+ = 0V or V+ = 1.65-5.5V and the control logic pin is set  appropriately.  If the V+ node is between >0 and <1.65V the leakage current is undefined.    

1) Are you able to use the 5V rail for the TS5A3160 to keep V+ between 1.65 V to 5.5 V?  This will ensure the datasheet performance and keep one path Hi-Z and the other path low impedance.  

2) If you make V+ = 0V with a weak pull down does the TS5A3160 signal path become Hi-Z? 

3)  Now using 1:1 kills the whole idea of being able to use the button for two purposes .... no? I miss interpreted your comment and thought you were just trying to isolate the touch button node.  I didn't notice the capacitive touch button in the top left. 

 

Thank you,

Adam

Den,

What was the resolution to your issue? Do you need additional support?

Thank you,
Adam

Den,

I am closing this thread due to inactivity. If you wish to re open the thread you may reply here or start a new thread.

Thank you,
Adam