SN65HVD256: detection condition of Undervoltatge Lockout time

Hi Shinichi,

Just a heads up we have many newer devices that are drop in replacements to the SN65HVD256. I would recommend looking at the TCAN1057A. This is a cheaper solution and it is much more robust. It also will have better supply going forward.

1. Hi Shinichi, there should be a filter on the device that stops glitches from causing an under-voltage event. However, this wasn't characterized on this datasheet. I will check with design on this but since this is an older device I am not sure if I will get an answer. If the device goes under voltage it will automatically go into protected mode. For the SN65HVD256 it will take 300 us to go back to normal operation once the voltage is no longer under-voltage.

On our newer devices we do have filter times on under voltage events.

What do you mean by the UVCC deviation? What we provide are the possible min and max values so you must design for the worst case scenario which is that your UVCC threshold is 4.45V and ensure that your VCC does not dip below 4.45V.

2. I'm not sure what condition you are looking for, this is an older device and did not have this value characterized. The TCAN1057A does characterize this value and it is t_MODE which is at max 20 us.

Yes 300 us will even be the recovery time for a momentary drop.

No the falling threshold is specific value that changes from device to device based on the process parameters of the device and the temperature conditions. The under-voltage threshold does have hysteresis however so the rising threshold is 200 mV above your falling threshold to clear the under voltage event on VCC.

You may not be seeing an undevoltage event for multiple reasons however keep in mind that 4.45V is the upper limit for the undevoltage threshold. There is a very good chance that your device's undevoltage threshold is much lower. Have you lowered VCC slowly to see where the threshold is for your device? How are you checking that the device didn't go into protected mode? Are you constantly sending data on TXD and seeing that the CAN bus continues to send data.

Best,

Chris

Shinichi,

Are you an FAE for TI? If not does your customer have an FAE from TI?

If possible please have your FAE email me and we can discuss this offline. If not then this information is NDA.

What we show on the datasheet is the absolute minimum and maximum that the threshold can be for all temperature cases. You have to design your system around these mins and maxes. For instance, your VCC should never drop below the maximum UVCC threshold (4.45V) in all temperature cases. Then if want to under voltage your device you should always bring your VCC below the minimum UVCC threshold (3.5V).

Even if I could give you the information on how much the threshold varies it would not be a good way to design your system because you will get a device that is an outlier due to silicon processes and it will not be in the exact threshold range that you were told it was. What you see in the datasheet are guarantees for how our device will operate under any variance of silicon with the different conditions and temperature ratings laid out in the datasheet. The information that you are asking for would just be an estimate of how the device's threshold would change with temp. That would not be a good way to design your system.

If you need a device with a smaller threshold range for the UVCC I can recommend the TCAN1057AV. It would be drop in to the device you are using now. The big difference is that the VIO pin on the TCAN1057AV level shifts the TXD and RXD pin so you can feed it whatever voltage rail your MCU uses so you no longer need to level shift your TXD pin. It also has a smaller UVCC falling threshold so that if your VCC dips below 4.5V you have more room for it not to under voltage:

Best,

Chris