So I'm finally getting around to trying to learn how to use SONAR again and have found this problem.
I bought SONAR8 and this Axiom 61 controller (and old grey one not the red one) in probably 2008 (?) and upgraded to X1 a year ago. 

After doing the ACT learning I then discover inside my project that the rotary knobs (B1 to B8) I have assigned to panning wont pan to 0 (center) - the closest to centered they will get is 1%L and 1%R, but not to 0. I have to use the mouse to carefully get it to not be panned left or right 1%.
It's like there is no 0 detent... it goes from 1%L to 1%R with no in-between.

I'm sure this must be an old resolved issue here, but I had no results from the search function.

Thanks for the help!

I have a Carillon 61 controller and it does exactly the same thing. I tried a lot of fixes/workarounds but no joy. Does it really matter? I doubt it. I certainly couldn't hear any difference between 1%L and 1%R.
Funny thing is, when I hooked up the controller to a previous mix that had the pans set at 0, they would go to 1%L or 1%R as soon as I touched the controller even though I hadn't moved them one way or the other. Strange.
When I removed the controller and restarted the project, the pans would be back to 0. Strange and stranger.
Could it be that I only paid 50 bucks for a cheap Carillon controller?
When I plug in the Carillon 61, its is detected by Sonar as an Axiom controller. Strange, stranger and strangest.
The Carillon comes with no driver software, it says in the instructions that none are required for Windows, and I could find none on the web. I assume that Sonar gets its driver from Windows and knows what it is doing when it detects it.
I will leave it up to better minds than mine to discuss this.

My AlphaTrack does the same but I can double click the pots and doing so sets them to 0. Can you double click the pots on your controller?

Hi gswitz, yes that does work. Still, when I move the controller again, they won't return to 0, I have to double click them every time I want to return them to 0.
I guess this is a bug of sorts, but as I said, I'm not particularly worried by this behaviour as I can't tell the difference between 1%L or R, and 0.
For me, I'm not going to pursue this, there are other issues that I would rather use my time on. I guess I will keep my eye on this thread to see if there are any others that consider this a problem, but am not going to try and solve it.