SN65HVD251 ISO11783 verification

Hi Ingo,

Can you please clarify your problem for me? Is the device not switching at the defined datasheet limits?

Hi Michael,

i really don't know the reason why physical layer test failed. The commissioned test lab tested according ISO 11783 and 99% passed. Also the test "Verify the ECU is able to detect dominant bit Levels at all CAN voltage levels".

But the complementary test "Verify the ECU ist able to detect recessive bit Levels at all CAN voltage Levels" has failed. I don't know the meaning of "CAN voltage". It could be a verification of compatibility 3.3V/5.0V Systems, or it meant function on several CAN bias voltages?

All the criteria ar not adjustable by me, they belong to the values given by the datasheet, but there ISO 11783 is certified.

greatings Ingo

Hi Ingo,

Reading through the ISO 11783 standard, it seems like this requirement just means that the CAN transceiver used in the ECU needs to be able to reliably register a recessive level (i.e., CANH at most 500 mV above CANL) across the full common-mode range. This shouldn't be an issue for HVD251, though, since it has a minimum "negative-going" (dominant-to-recessive) input threshold of 500 mV across its common mode range. (The two furthest extremes of this testing are documented in Table 1 of the HVD251 datasheet.)

Is it possible to look a little deeper into how this test was set up by the test lab, and what criteria they used to determine a failure? I would expect that they are configuring the test per Figure 18 of the ISO 11783 standard ("Test of input threshold for recessive bit detection") and verifying a high-level output on the "R" pin of HVD251, although with this set-up I would expect no issues.

Regards,
Max

Hi Max,

after contacting Test Lab we verified our design in the same way. We could retrace the error. If the common mode voltage becomes above 3.8V the RX signal will become longer as the TX signal; then the transmission error appears.
We replaced HVD251 with TJA1042 - it works to >10V common mode! What's your meaning about this? How is it possible to send screen shots of the wrong signals of HVD251?
greetings
Ingo

Hi Ingo,

When you reply in this forum, you should see an option to "use rich formatting" (towards the bottom of the text input area).  If you click this it will bring you to a text editor with additional features.  You can use that editor to either attach files to your post or insert media (like images) in-line with the text.

Alternatively, you can also send me images through e-mail if you prefer.  My e-mail address is (first initial)-(last name)@ti.com.

It would be interesting to see the waveforms when the RX signal is lengthened - it should help us better understand what is going on and track down the source of the problem.

Regards,
Max Robertson