CAN decoding broken in latest beta.

Forum for discussing PicoScope version 6 (non-automotive version)
Post Reply
bennog
Active User
Active User
Posts: 33
Joined: Mon Nov 26, 2012 9:16 am
Location: Netherlands

CAN decoding broken in latest beta.

Post by bennog » Tue Apr 30, 2013 8:09 am

I am using the latest beta of PicoScope program. 6.7.19.2
And noticed the can decoding is not working as expected.

Image

Some info on the scope file.
Probe was differential probe on can-H and can-L more than 20 devices on the bus.
Baudrate 500kbps sample rate 4MS/sec (is the 4MS/sec to slow ?)

psdata file (72MB) http://www.ans-net.nl/picotech/20130430 ... set.psdata

The decoder seems mis the most packages.
Also the PicoScope program hang when changing the Decode threshold voltage in above psdata file.

Regards,
Benno

bennog
Active User
Active User
Posts: 33
Joined: Mon Nov 26, 2012 9:16 am
Location: Netherlands

Re: CAN decoding broken in latest beta.

Post by bennog » Tue Apr 30, 2013 5:17 pm

Test with the latest release candidate 20 is the same problem of not correctly decoding CAN.
Same sample used as above, also tested with a sample with 10MSamples/sec but same problem.

Benno

Martyn
Site Admin
Site Admin
Posts: 3346
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: CAN decoding broken in latest beta.

Post by Martyn » Tue May 28, 2013 9:48 am

You appear to be using a x10 probe, what are the characteristics of the signal you are looking at ?

If you just capture a few frames is the decode correct ?
Martyn
Technical Support Manager

bennog
Active User
Active User
Posts: 33
Joined: Mon Nov 26, 2012 9:16 am
Location: Netherlands

Re: CAN decoding broken in latest beta.

Post by bennog » Thu May 30, 2013 2:17 pm

I use a yokogawa 700925 DIFFERENTIAL PROBE probe.

http://tmi.yokogawa.com/products/oscill ... ial-probe/

This is probably the same as your TA041 differential probe.

On the decoding of small samples I will test this when I have acces to a machine.

Regards,
Benno

bennog
Active User
Active User
Posts: 33
Joined: Mon Nov 26, 2012 9:16 am
Location: Netherlands

Re: CAN decoding broken in latest beta.

Post by bennog » Thu May 30, 2013 4:35 pm

The anomaly's I am looking for are as shown in below image.

Image

In the above sample a frequency inverter was bringing the CAN bus to its knees, but only during starting of the motor.
Caused by a lose shielding of the motor cable.

A 60MB psdata file can be found below.
http://www.ans-net.nl/picotech/20130530-0001.psdata

P.S. Picosope crashes on closing after clicking once on the first error frame.

Benno

Martyn
Site Admin
Site Admin
Posts: 3346
Joined: Fri Jun 10, 2011 8:15 am
Location: St. Neots

Re: CAN decoding broken in latest beta.

Post by Martyn » Thu Jun 06, 2013 1:51 pm

You may have better luck with 6.7.21 or later, although some people are having issues with this not detecting the scope. I have seen sensible data except for where there is obvious noise on the signals.

I am surprised that most of the hash doesn't disappear using both CanH and CanL
Martyn
Technical Support Manager

bennog
Active User
Active User
Posts: 33
Joined: Mon Nov 26, 2012 9:16 am
Location: Netherlands

Re: CAN decoding broken in latest beta.

Post by bennog » Thu Jun 06, 2013 5:10 pm

this was due a frequency inverter injecting noise into the CAN bus. The freq. inverter had a CAN on board controller.

Benno

Post Reply