Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

  • State
  • Replies 8 replies
  • Subscribers 32 subscribers
  • Views 139 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

SN65HVD1050: Activation delay of up to 30ms after powering up

Steve Hättasch
Prodigy 10 points
Part Number: SN65HVD1050
Other Parts Discussed in Thread: TCAN1057A-Q1

Tool/software:

Hello,

In our designs we usually use the NXP TJA1050 CAN Transceiver. For a short while now we are using the TI SN65HVD1050 as a replacement and problems seem to arise from this change.

Given operating conditions:

We are triggering our controller to boot loader mode via special CAN messages. Thus, the controller resets and clears the needed flash pages. Due to safety reasons, our circuit disables the power supply of the CAN transceiver during flash erase. Once done, the boot loader re-enables the transceiver supply and waits for 3ms (supply is OK after 1.5ms) until it tries to connect to the bus. 

Observation - bad case (SN65HVD1050):

After powering up the transceiver, messages on the bus get through to the RX-Pin after approx. 30ms. See screenshot - yellow curve: 5V supply of transceiver; blue curve: Rx-Pin of the transceiver SN65HVD1050

With this the CAN-Controller transitions into bit-dominant-error and then to bus-off.

Observation - good case (TJA1050):

After powering up the transceiver, messages on the bus get through to the RX-Pin after approx. 2ms. See screenshot - yellow curve: 5V supply of transceiver; blue curve: Rx-Pin of the transceiver TJA1050

Questions:

Can anyone confirm, if there is some kind of delay circuit in the TI transceiver? And under which circumstances?

Is there a chance, that it might be a bug inside the transceiver?

Am I missing some features or conditions to be met before the transceiver starts?

Any ideas how to avoid it in the first hand (instead of just increasing the delay for continuing the transmission) ?

To make it clear: we have the possibility to change the software in a future release to work around this issue, but we need to fully understand the behavior. However, we have several devices out at the customers with this problem which get bricked by this behavior and need to get sent in for repair, which is more of a pain.

Hope anybody can help.

Best regards

Steve

  • 0
    TI__Guru 70762 points

    Steve,

    The startup delay shouldn't be this long, do you see this on all SN65HVD1050 devices in your testing? This is an older device, and the TCAN1057A-Q1 can be used in its place for better performance, cost structure, and supply long term. I'd recommend moving to that device if you can. In the meantime, I will look at what validation data we have to see if there's anything indicating start-up delay.

    Is it possible for you to also share a schematic?

    Regards,

    Eric Hackett 

  • 0 in reply to Eric Hackett
    Prodigy 10 points

    Hi Eric,

    thanks for the impulse for the future, I will definitely safe this idea for the upcoming projects. But for the devices already shipped, this won't help.

    So I still try to understand what we are doing wrong, since these problems never occurred in the past with TJA1050. I'd really appreciate, if you could advice anything to cope this on existing devices.

    As you asked for the schematics, here is an excerpt with the affected parts. As you can see, the Tx and Rx lines are split as they are connected to two microprocessors. The standby line is routed to one of the controllers but is not used and hence static low. The CAN_Enable line deactivates the 5V-supply during flash erase and is re-enabled afterwards.

  • 0 in reply to Eric Hackett
    Prodigy 10 points

    we figured out, that the startup delay increases with lower temperatures. so at room temperature and below it's around 20-30ms. Rising above ~45°C ambient eliminates the problem. This holds for parts bought from 2023 to now and for older ones (we had in the lab for 2+ years)

  • 0 in reply to Eric Hackett
    Prodigy 10 points

    Hello Eric,

    any ideas oder any news on this topic?

  • 0 in reply to Steve Hättasch
    TI__Guru 70762 points

    Steve,

    Sorry about the delay on this one, please give until Wednesday for a proper response. Thanks for your patience.

    Regards,

    Eric Hackett 

  • 0 in reply to Eric Hackett
    TI__Guru 70762 points

    Steve,

    It's been difficult to get some design assistance on this device since it is older, but your schematic doesn't show any reason why this problem would be occurring. What does the startup time change to once the temperature is raised to 45C?

    Regards,

    Eric Hackett 

  • 0 in reply to Eric Hackett
    Prodigy 10 points

    Hi Eric,

    once the temperature is raised to 45C the startup time reduces to approximately 1ms (+/-).

    Meanwhile we got some more devices back from customers with the same problem. On some the delay at room ambient is even as high als 60ms and more. It seems to get worse at negative temperatures (e.g. -10C)

  • 0 in reply to Steve Hättasch
    TI__Guru 70762 points

    Steve,

    Let me take this back to design and get back to you on Monday. Apologies for the delay in between responses, with the holidays in full swing we have limited resources at the moment.

    Regards,

    Eric Hackett