This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TCAN1044A-Q1: CAN physical layer test waveform confirmation

Part Number: TCAN1044A-Q1

Tool/software:

Hi team,

The customer wants to ask, if the matching terminal resistor of CAN is not attached, will it affect the waveform test of CAN when it is unloaded?
The first picture: The waveform of CAN is very volatile, S-shaped. I checked and found that the 5V wave of my power supply is normal when the CAN is disconnected, but when it is restored, the wave becomes very large. I reconnected the ground, but it still affects the CAN load. I also tested the load capacity of my 5V power supply chip and there is no problem.

The second picture is a no-load test without a terminal resistor on the board. The waveform looks a bit problematic. If it is no-load, the waveform amplitude should be larger. It should not be like the picture.

The third picture shows the test end connected to a 60 ohm resistor. There is still no resistor attached to the board end. The waveform is slightly normal, but the waveform is still not very good.

...

The following pictures are the test terminations with 120 ohm resistors. The waveforms with resistors attached to the board end are the most normal. However, after the waveform is magnified, there are still some problems. The waveform in the middle is not straight and a little curved.

The waveform after enlargement is a bit curved

  • Alan,

    Can you please share a schematic of the customer board? It looks like the 5V supply is not decoupled capacitively and so the voltage has ripple/amplitude dip due to the dynamic current change going from recessive to dominant.

    Also, without the proper termination, waveforms will not be precise, so it's always important to terminate the CAN bus properly. In all of these cases it looks like the RXD waveform is still properly responding though, so there shouldn't be any issue there.

    Regards,

    Eric Hackett