TPS3813K33DBVR - always RESETs even with WDI properly set

Hi Max,

Looking at the RESET pulse and your configuration it does look like the pulse input is not being detected by the WDI pin. Could you please check the continuity from the WDI pin and the GPIO pin? The 10ms pulse should be wide enough for the state machine to detect the rising edge. Do you know the slew rate of the pulse?

Jesse 

Hi Jesse,

we checked and there's continuity, we have repeated the test with two devices on two different boards and we obtain the same results.

Slew rate. From the datasheet, maximum rise rate is 100 ns/V; at the beginning we use 1,05us/V and we decrease to 100ns/V but there is no effect. I don't think it is the main issue, otherwise there should be indicated in datasheet as a mandatory condition and not a hidden data.

Let me know

Hi Max, 

Do you happen to have access to our EVM for this device.https://www.ti.com/tool/TPS3813Q1EVM 

I will have to run some bench test and see if I can replicate your issue.

Jesse 

Hi Jesse,

frankly speaking, I didn't buy the EVM since I trusted in the documentation. The device shall have a simple behaviour and I cannot buy an EVM for each device I use in my project.

Hope that you manage to replicate the issue, but at the moment I do not use the device at all because it's not working.

Regards

Hi Max,

Unfortunately, the only way I am able to replicate your issue is 1.) Severing the connection connection of the WDI or 2.) inputting a pulse smaller than 20ns to the WDI. With your configuration, the device is functioning as intended. WDT=WDR=VDD WDI=10ms every 60s.

Below is the waveform for 20ns wdi input

For your uC_Watch_Dog_Refresh net, is there a switch in between the WDI and the GPIO or is it a solid trace?

Jesse 

Hi Jesse,

thanks for your test. Just to be aligned, you test with 5V Vdd, right? Because we use 3,3V in our application, so voltage level are adjusted to fit with such value.

Let me know, please

We try, however, to test again but using more probes to understand where is the issue.

Regards

Hi Max,

VDD was set to 3.3V

Jesse 

Ok, but I see that the WDI signal amplitude is more than 3,3V. Is it correct or it's just a matter of oscilloscope view.

I wonder that my issue is not due to amplitude higher than 3,3V that I could not generate with a microcontroller of 3,3V.

Regards

Hi Max,

The VIH(min) for the WDI is .7*VDD also the WDI utilizes a rising edge detection. I just want to double, this is being tested on room temp?

Jesse  

Hi Jesse,
yes, we tested it at room temperature. We saw the datasheet with the VIHmin, so I asked you why the voltage level was higher than 3,3V in the picture.

Regards