Author Topic: Keysight DSOX1200G: unusually powerful CAN training signal  (Read 436 times)

0 Members and 1 Guest are viewing this topic.

Offline RBBVNL9Topic starter

  • Frequent Contributor
  • **
  • Posts: 326
  • Country: nl
Keysight DSOX1200G: unusually powerful CAN training signal
« on: January 25, 2022, 08:11:51 am »
Hi,

The Keysight DSOX1200G series oscilloscopes may not always get as much love as other makes (which is understandable).

Yet, they have one useful (but not really extensively documented) feature that I wish to share with those that are not aware of it. It’s the extensive set of training signals, and in particular, the CAN serial bus training signal.

Typically, if oscilloscopes can send out CAN training signals (some cannot), they simply send repeatedly out the same simple message with their own name. Also, decoding these signals typically results in observing a lot of NACK errors unless a properly configured receiver circuit is connected while measuring.

Not the DSOX. It sends out a long training sequence (16 messages) that includes short (11 bit) and long (29 bit) ID, data and remote packets, and varying message lengths (1, 4, 5 and 8 bytes). Quasi randomly inserted over training sequences there are a variety of different errors (CRC error, bit stuffing error, NACK), very suitable for testing serial decoders. Apart from frames where a NACK error is deliberately inserted, the ACK is already added (unlike most demo/training CAN signals on other oscilloscopes).

While a couple of simple developer boards (like two ARDUINO MKR series), each equipped with a CAN shield, would you allow to quite flexibly do most of the above too, this cannot create the error signals as created by the DSOX.

PS Tip: because it's a rich sequency, you will want to trigger it properly to a stead point. To do sone, you can trigger on a data frame with ID = 0x7F and first data byte 0xE4 and set trigger holdoff to 35ms.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf