Using Kvaser CAN products with Labview

[Dasylab User]

Hi,

Has anyone here any (independant!) experience of using Kvaser CAN products with Labview 8? (in particular their Leaf-Light USB interface?)

The demos with their libraries seem ok (albeit basic), but, when venturing out alone & expanding their VIs, the behaviour of the interface/the VIs seem erratic.

Thanks & Regards,

Paul Towle.

Luk Leamington.

http://www.luk-leamington.co.uk/

[Bob_Y.]

I haven't used the leaf-light product, but I have used the memorator.  We normally pick NI or Softing CAN hardware, but we needed a fairly cheap USB (no NI USB CAN?  Why not?) solution and this fit the bill nicely.  It's CAN communication is fine and the supplied VIs worked as expected in our applicaiton.  We had to do a bit of re-work to get everything to fit with our applicaiton, but it was fairly straight-forward.

If you give more details on the problems you are running into, maybe we can help.

Hope that this helps,
Bob Young

[Dasylab User]

Hi Bob!

Thanks for your posting.

I agree, we normally use NICAN-2 PCI cards. In this case, we needed a low cost USB CAN interface .... but NI have nothing to offer!  I suspect that’s all about to change though. NI is finally waking up to USB!

In the meantime, I've been trying to use Labview 8 with the Leaf-Light.

We are simply trying to display a number of CAN channels on a chart in real-time. Normally, I would use Dasylab but, there isn’t a Dasylab Driver for the leaf Light! With Dasylab, this project would be easy!!

The problems are:

1). The CAN channel definitions really need to come from a CANdb file

2). The VI needs to display as many as 16 CAN channel traces on a Chart in real time AND, be able to offer some analytical features (cursor measurements etc)

For the time being, I have postponed 1). and I am concentrated on 2). ….although I would welcome any suggestions you may have on the creation of a VI to configure a Leaf Light from a CANdb file!

I am manually extracting the CAN channels (2 at present) and am plotting the data on a waveform chart. I then have to transfer this plotted data into a waveform graph (because the chart doesn’t provide cursors!!). The VI almost(!) works ….but is;

a). susceptible to timing glitches (usually more so the longer it’s been running)

b). somewhat jerky in its operation.

+ I’m not finding it particularly easy to configure the time axis on the charts/graph (mainly due to my inexperience with LV?)  

The VIs are attached in a zip file although I apologise about their scruffy (part developed) appearance!  The remainder of the VIs are standard Kvaser VIs from their library.

The main inner loop continually reads CAN data from the ‘CAN Read’ VI and channels this to the chart as fast as the loop will run. I am not sure whether I should be plotting the data wrt the CAN time stamp information as opposed to just letting the chart plot the data against some unknown (and uncontrolled) time reference. Perhaps this is the problem??

I have tried the VI with a ‘Wait’ function (to set the pace to approx 1-2mS) but the timing was equally erratic.

I am relatively inexperienced in the use of Labview (compared to some of you guys!) …. but would welcome any comments or suggestions you have as to where I may be going wrong.

Thanks & Regards,

Paul Towle.