TS3A44159RSVR analog inputs clamp diodes?

Hello Volker,

The datasheet gives very specific max ratings on each pin as shown in the below table. Exceeding these ranges can damage the part.

Can you provide a block diagram of what you are trying to accomplish?

Regards,

Andrew

Hi Andrew,

I'm aware of the data sheet specifications. But the given voltage range is normally a value implying a low resistant source beeing capable of delivering high currents. It should be the current destroyig, not the voltage, by deliveringthe energy for power dissipation over input protection circuities like clamp diodes. This is normally specified by parameters like "max. sink or source current to any input pin". Hence the common protection for micro inputs is to use serial resistances to reduce any fault currents.

However in our measurements we do find that there must be some kind of clamping circuit inside the TS3A44159RSVR whch is between the input crcuit shown in the drawing and our STM micro, as a high impedance 5 to 6 volt source is clamped down to about 3.6 V by TS3A44159RSVR inputs. My question aims to get more information about this input protection circuit to estimate its clamping limits.

here is our input circuit (the output does not connect directly to the STM as stated in the lable ut instead runs trough an TS3A44159RSVR analog switch):

Aim is to protect our 0-10 V input up to 24V fault voltage. The voltage divider scales down to 0-3 V input for the STM MPU (which gets a precision ref voltage of 3V). D2 protects against any ESD and clamps down fault voltages to a max of about 5 to 6 V. D1 was planned to reduce the faut voltage further down to 3.3V minus Vf of the Schottky which is less than the specified V+ +/- 0.5V o the analog switch. BUT: the Vr of the Schottky will detroy our accuracy as it is non linear and highly temperature dependend and hence can't be eliminated by calcultion of the MPU. So we tested without D1 and got the astonishing result that the voltage at the TS3A44159RSVR input was only 3.6V when we applied 24V at our analog input. Testing different devices for times over 1h did neither destroy the analg switches or the MPUs nor did it reduce the precision of the switch or analog MPU input stage (we use PGA and ADC). So by test we would say we could simply leave out D1 but by data sheet specification we would need any further kind of protection. This was the reason for my question about the switch input protection circuit in order to get more reliable information before using any unspecified properties of the device. E.g. we could simply add another serial resistance to reduce possible sink or source currents to the switche's input if needed. But we would not do so deliberately without need as it would increase input impedance thus increasing sensitivity to HF inductions. 3V inpt at 12 bit means we are talking about 0.73 mV accuracy and with our input voltage divider we need to care about 520 nA fault currents keepng up 12 bit accuracy.

Hello Volker,

Design is looking into this.

Regards,

Andrew

Hello Volker,

I apologize for the delay with this but the design team needed to run a simulation to be sure and here is there response on this: 

"There is a diode that is clamping the input.  6V is greater than the abs max ratings for the IO pins and due to the process this kind of voltage will be over-stressing the transistors and will degrade the part.  Because of this we cannot guarantee proper operation over the lifetime of the device.

Regards,

Andrew