TCAN3414: WUP issue on TCAN3414DR

Hi Marcel,

This is as expected as shown in figure 7-5 of the data sheet. Please see 7.4.3.1 for more information, thanks. 

Best Regards,

Michael. 

Hi Michael
I am sorry, but I am not agree. I did some further analysis according your notes. The red highlighted pulse is still not OK

Best regards

Marcel

Hi Marcel,

Yes, the red-highlighted ~170 ns negative pulse is not the expected WUP indication on RXD. However, it is also not necessarily evidence of a device failure. I.e., the data sheet specifies tWK_FILTER as 0.5–1.8 µs. A bus dominant / recessive interval shorter than the minimum filter time should not qualify as a filtered WUP element. 

I.e., since 170 ns is below the 0.5-µs minimum, the pulse may not represent a qualified filtered dominant WUP event. The approximately 1-µs RXD low pulse observed after the valid WUP pattern should be the expected wake-up indication.

I would next determine whether the 170 ns pulse is repeatable for every dominant bit, whether it changes with bus termination, and whether it is present directly at the IC pin with a low-capacitance probe.

If it is consistently generated by every dominant bit, particularly with a clean 60-Ω CAN network and directly at the RXD pin, then I would consider further taking a closer look at the device level, rather than the system level.

• Measure directly at the TCAN3414 RXD pin, preferably with active probe or ≤1 pF probe, short ground connection, sufficient bandwidth and 10× probe if using a conventional passive probe I.e., the 170 ns pulse should be easily affected by probing configuration.
• Check whether the 170 ns pulse occurs on every dominant i.e., if it occurs only occasionally depending on the CAN bit sequence, it may be associated with the internal WUP state machine / filter.
• Because you’re operating at 1 Mbps, this is particularly important. At 1 Mbps: 1bit=1us and the WUP filter minimum is only 0.5 µs. Hence, ringing / overshoot / undershoot around the CAN dominant / recessive transition could potentially interact with the wake-up receiver and would recommend to capture CANH, CANL, plus VDIFF ( CANH−CANL) and RXD all at the IC pin to confirm. Especially since you said two devices are on the CAN bus, both have a 240 Ω termination resistor. So the effective termination is 120 Ω. That’s a relatively light loading compared with the usual 60-Ω CAN bus. It is not inherently wrong for your controlled test, but it means the CAN differential waveform can have different edge / ringing characteristics than a typical 60-Ω network. Since you’re specifically evaluating the WUP detector at 1 Mbps, I would repeat the test with 60 Ω total termination if possible. That gives you a much more representative CAN physical layer.

Furthermore, the ~170ns RXD pulse seem to occur during the first filtered dominant of the WUP sequence and is shorter than the minimum WUP filter time (0.5 µs). Hence, it should not constitute a valid wake-up request by itself. The subsequent ~1-µs RXD pulse occurs after the required filtered dominant–recessive–dominant sequence and is the expected WUP indication. 

• Repeat exactly the same experiment but send only one 1-µs dominant followed by a long recessive, then stop.
  • If you see: first dominant → ~170 ns RXD pulse → no wake, that should establish that the 170 ns pulse is associated specifically with the first filtered dominant / WUP state-machine entry, rather than with CAN communication.
• Then send the complete WUP and confirm: second filtered dominant → ~1-µs RXD pulse → wake request.
  • That sould be a very clean demonstration of the TCAN3414’s WUP behavior, thanks.

Best Regards,

Michael.

Hi Michael

I have found the the issue. For testing purposes, I set the STB pin always to GND. So the TCAN3414DR filter for WUP go never a reset. That s why I could see a negative pulse by the first dominant. But the WUP pulse is still only 170ns. In the datasheet it is not defined what length of negative pulse I can expect. 

Best regards

Marcel