So I was able to make some good progress. I was able to get the spindle direction flipped around so it works correctly. I still have the problem with the spindle not running at commanded speed, or at least there is a discrepancy as to what is commanded and what the PDB display is reporting. I would assume that the PDB display pulls the info off the spindle encoder, so it's probably more accurate. It was apparently doing this before, because when I reverted back to sagreen's files, the RPM behavior was identical.
500=410 =18% Low
1000=1140 =11.4% High
1500=1806 = 20% High
2000=2401 = 20% High
3000=3525 = 17.5% High
4000=4700 = 17.5% High
5000=5865 = 17.3% High
6000=6230 = 4% High
So if you throw out the highest and the lowest #, and the highest number, the bulk of the error appears to be pretty consistent 17-20% It appears there is some information over in one of the older threads that is pertinent to scaling the speed. I am going to go over and read through that a little more.
Also commanding RPM in the RPM window is still broken for me, It seems to work fine from the MDI, so not a dealbreaker.
I will give 2.1.4 a try tonight. Thanks guys.
I was able to fix the commander speed in the novakon.hal I think it is so strange that it does
Not work for you.
I will check spindle rpm tonight. See what the machine does. When setting it all up
I set the spindle speed to 6000 rpm on the control and adjusted the blue pot on the novkon board
Until it was reading 10volts. I would swear it was around 480 when commanding 500.
Tinkered with this thing so much I can not remember all the crap I have adjusted on the drives.
Was running mach4 with a hicon. But until they do something different with the trajectory planner
I have abbandonded it. Also like with linuxcnc and pathpilot motion stops on a accidental probe hit
Wow, so what day. After doing a bunch of research, it appeared that lincurve would be just the ticket to sort out the speed errors I was seeing. So I started working on it by re-enabling lincurve in the novakon.hal. That didn't go so well. So long story short, with the change to Linux Mint in 2.x PP no longer compiles the real time libs as kernel modules, instead they run in user space. So I found a thread on the LinuxCNC forum, and ended up compiling everything from scratch. I copied the new lincurve over, and bam, we are in business. My RPM error is down in the single digits across the board now, except for right at 6000, I will need to tweak that guy some more. Anyhow, I have attached a copy of my files, and a complete set of 2.x compatible rt libraries in case anyone should ever have a need.
Thanks for all the help along the way!!
Jason