TCAN1043A-Q1: How to wake up through RXD ?

Hi AE,

      Thanks for your response, the extended question is as follows:

      Loop the can bus data in every 10ms, We transition from Normal mode to Sleep mode by turning off Vcc.

      During the transition, we found that the RXD has been held (keep H or L) and not entering sleep mode.

      1. Keep looping can bus data (10ms)

             1-1. What are the ways to enter sleep mode properly?

             1.2. Please provide the suggestions to prevent RXD from being held during the transition, Thanks.

Best Regards

Hi Bruce,

There is no issue with going to sleep mode by turning off VCC, VCC need to fall below UVCC for tUV in order to enter the sleep mode. During a UVCC event, RXD will remain high as long as VIO is present. If VIO is not present, RXD will be in high impedance state. 

Regards,

Sean

Hi AE,

      Thanks for your response.

       After testing, I try to disable VCC when the can bus data keep sending in every 10ms, the actual results are as follows:

      1. The result show the status will keep in normal mode include the RXD, the nFAULT and the INH. ( can not enter the sleep mode )

      2. The result show the status will keep in Standby mode include the RXD, the nFAULT and the INH.  ( can not enter the sleep mode )

       It seems that during the transition process ( VCC off) , if the can bus continues to transmit, the status can not enter the sleep mode. ( keep normal and standby )

       I would like to understand the root cause of the issue and please provide suggestions to avoid the issue.

       I look forward to your response, Thanks!!!

       Best Regards

Hi Bruce,

It seems like the device entered standby mode from the sleep mode, because the data on CAN bus wake up the device. It is true that UVCC event made the device enter sleep mode, but only a very short time. (Yes, the device can be waked up immediately after going to sleep mode, no need for VCC) Then it was waked up by WUP, thus entered the standby mode. It was very likely that INH output Low only for very little time, so not being captured.

There's a method to confirm the situation I said above. You can perform the same test, but instead of keep sending data every 10ms, try to input a signal on TXD, and see if there is output on CAN bus pins. The CAN driver will be disabled in standby mode so you should expect to see no waveform on CAN bus.

To put the device in sleep mode, make sure there is no CAN communication. Otherwise the device can be waked up by even a 2 bits message.

Regards,

Sean

Hi AE,

      Thanks for your clear response.

      We designed via RXD pull low to wake up system to power VCC on.

      In reality, the waking up from sleep mode may be too fast cause the system couldn't wake up to power VCC on. (the RXD being pulled low too short time)

      Currently, our issue lies in the need to wake up the system to power VCC on via RXD pull low.

      In this regard, could you please provide me the suggestions to improve the RXD issue, Thanks.

        Best Regards

Hi Bruce,

Once a wake-up request (WAKERQ) received, RXD will go low. Did you mean the device stay in standby mode (RXD = Low) for too short time and cannot be detected? In standby mode and under UVCC condition, the device will wait for tUV (100 - 350 ms) before transition back to sleep mode, was that time not long enough?

Would you consider to use INH to power on VCC? INH is pulled low in sleep mode and output high in all other modes. It can be used to control microcontroller or voltage regulator. 

Regards,

Sean

Hi AE,

     Thanks for your clear response.

     Q1. Yes, the system can not be detected by RXD cause UVCC in Standby mode.

     Q2. In theory, RXD should be pulled H (Vio =H) when wait for tUV (100-350 ms) from standby mode, but in practice, it's not observed RXD Low-High-Low.

            Is it possible that other conditions are triggering cause the RXD always keep low ?

     I know that using INH to connect the system wake pin to fix the issue, but we do not have any pins available for use In the existing circuitry.

     I also try to modify the circuit to fix the issue, but I currently do not have any ideas.

     About the circuit design, I need AE to provide me some ideas or suggestions to address this issue.

     Best Regards

Hi Bruce,

Bruce Dai said:

     Q2. In theory, RXD should be pulled H (Vio =H) when wait for tUV (100-350 ms) from standby mode, but in practice, it's not observed RXD Low-High-Low.

RXD will be pulled to Low in standby mode, and after tUV timer expired (because UVCC event), the device enters sleep mode and RXD pulled high.

So the sequence is: RXD High (sleep mode) → RXD Low (wakeup event, goes to standby mode) → RXD High (tUV timer expired, goes to sleep mode), see figure below,

Do you have a waveform capture that the device waked up from sleep mode, but RXD = Low was not observed in standby mode?

Regards,

Sean

Bruce Dai said:

In reality, the waking up from sleep mode may be too fast cause the system couldn't wake up to power VCC on. (the RXD being pulled low too short time)

      Currently, our issue lies in the need to wake up the system to power VCC on via RXD pull low.

Sorry I might misunderstand your question, so you need RXD pulled Low to be observed, correct? 

Bruce Dai said:

Is it possible that other conditions are triggering cause the RXD always keep low ?

But it seems like RXD here is always Low. Rather then RXD pulled Low, you expected to see RXD = High in standby mode?

Regards,

Sean

Hi AE,

     Q: But it seems like RXD here is always Low. Rather then RXD pulled Low, you expected to see RXD = High in standby mode?

      A: I expected to see RXD pull High about 100ms-350ms during the transition between standby mode and sleep mode.  

      This issue is only observed when continuously sending CAN bus every 10ms and simultaneously power off VCC.

      Besides connecting to the INH pin, I would like to explore other ways to solve this problem.

      Best regards

Hi Bruce,

Bruce Dai said:

: I expected to see RXD pull High about 100ms-350ms during the transition between standby mode and sleep mode.  

Because there is CAN communication (WUP every 10ms) on the Bus, the device will be kept waking up immediately after entering sleep mode. 

My recommendation is if you don't want the device to be waked up, ensure the message on the bus has a state time (either recessive or dominant) ＜  tWK(FILTER). In that case it won't be considered as a WUP, thus the device will not wake up and pull RXD to Low.

Regards,

Sean

Hi AE,

     Compare the differences between NXP TJA1043 and TI TCAN1043A in Can bus 10ms cycle.

      I found TJA1043 that can support RXD ( Green line ) wake up  when the Can driver enter the sleep mode form Normal.

      The TJA1043 also spends about 300ms to enter sleep mode, but the system also can wake up via RXD.

      In practical situations about TCAN1043A, the can bus may still be transmitting when entering the sleep mode, the RXD also can not be trigger.

      Our TI design can not avoid the issue and furthermore, others may encounter similar issues later on!

      Is the issue have common way to solve this issue ?

      Please provide me suggestions, Thanks!

      NXP TJA1043 VCC=Yellow, INH=Orange, nSTB=Purple       TCAN1043A

      Best Regards

Bruce,

This is the part TCAN1043A different with TJA1043, you need to toggle nSTB from Low to High in order to exit the sleep mode. Please find the 8.4 state machine in datasheet.

Regards,

Sean