584,830 active members*
5,783 visitors online*
Register for free
Login
Page 2 of 2 12
Results 21 to 30 of 30
  1. #21
    Join Date
    May 2006
    Posts
    4043

    Re: 4th axis feedrates

    Hi Mark,

    Say running the GCode with your TP settings I'm getting different times.

    With Facet Angle 0.5 degrees - 66 Seconds
    With Facet Angle 5.0 degrees - 42 Seconds

    I can understand my 0.5 degrees being faster than your 2 minutes as my PC and USB might be faster.

    But I don't understand the 5 degrees being slower than your 30 seconds. If no PC or USB limitation our times should be the same.

    There are some extra rapid moves at the beginning and end. Are you including those in your measurements?

    Also please post your Init file so I have the exact same Rapid speeds.
    TK
    http://dynomotion.com

  2. #22
    Join Date
    May 2012
    Posts
    537

    Re: 4th axis feedrates

    Hi Tom,

    Yes, I eliminated the rapids and m code time to give a more direct comparison of the code we are concerned about. I just took an approximate time from the 2 videos of 4th axis rotating. Just had a second look, seems like its closer to 34 seconds for the 5 degree video.

    Yes, your right again. Im using a fairly low end Windows 7 computer on my machine. All I had at the time, and stuck with it as its been getting the job done. Looks like a better computer would have helped here.

    See attached init file. I have a lot of things going on in my init file. Not sure if that can be slowing down things as well.

    Mark

  3. #23
    Join Date
    May 2012
    Posts
    537

    Re: 4th axis feedrates

    Tom,

    This problem I was having with the small facet angle slowing things down, What is the slowest part of this problem? Is this slowdown from my computer doing the trajectory planner calculations, or the communication between my computer and klfop, or something to do with Kflop?

    I've ran into this problem again with normal 3 axis surfacing. Very minor slowdown this time, but same thing as before, if you increase facet angle it runs much faster. I do seem to get a better finish with a smaller facet angle. Would a faster computer allow me to use smaller angle and no slowdowns? Thanks again.

    Mark

  4. #24
    Join Date
    May 2006
    Posts
    4043

    Re: 4th axis feedrates

    Hi Mark,

    I would expect the bottleneck to be the USB Communication. You can use SimpleFormsCS.exe to check the speed. More KFLOP User Threads running steals CPU time from the Communication. What the Threads are doing doesn't matter. My Computer (older Intel i7) shows:

    No User Threads running in KFLOP
    PC->KFLOP N=100000 Int32, Time=1.880 sec, 472KBytes/sec
    KFLOP->PC N=100000 Int32, Time=1.060 sec, 838KBytes/sec

    One User Thread running in KFLOP
    PC->KFLOP N=100000 Int32, Time=2.258 sec, 393KBytes/sec
    KFLOP->PC N=100000 Int32, Time=1.938 sec, 458KBytes/sec

    You might also try more KLP Low Pass Filtering to improve smoothness.

    HTH
    TK
    http://dynomotion.com

  5. #25
    Join Date
    May 2012
    Posts
    537

    Re: 4th axis feedrates

    Hi Tom,

    Cool, neat that we can easily check this. Looks like im way behind you on the pc-kflop times. This is with my init file running. With no threads running its basically the same, but the kflop-pc time went down to about 1 second.

    Come to think of it, it takes forever for this computer to write larger files to USB drives too, so maybe this is affecting my Kflop performance the same way. Im not even sure its USB 2.0 or if its older. This was a pretty low end computer even when it was new, and its pretty old now too. Maybe I should upgrade? What do you think?

    Mark

  6. #26
    Join Date
    May 2012
    Posts
    537

    Re: 4th axis feedrates

    Just did a USB speed test on this computer and it actually did better then my dell I7 in my office? My office computer was writing at 6.8mb/sec and the mills computer was around 8mb/sec. This may also be limited by the USB drive I was running the test on. But certainly much faster then its talking to kflop.

    Could it still be the computer slowing things down or something else? Also tried testing right after rebooting computer/kflop and no threads running at all. Same result, around 12.5 seconds. What would you consider a normal range Tom? Kinda wish i could try it with results closer to yours.

  7. #27
    Join Date
    May 2006
    Posts
    4043

    Re: 4th axis feedrates

    Hi Mark,

    Strange. I would think even older slow computers should be able to handle USB 2.0 rates easily.

    You might try updating the KFLOP USB Driver to the latest in V4.35b.

    You might try a different USB port.

    You might try that KFLOP on your Office Computer.
    TK
    http://dynomotion.com

  8. #28
    Join Date
    May 2012
    Posts
    537

    Re: 4th axis feedrates

    Hi Tom,

    Little update. Tried updating to 4.35b but no change. Tried updating the usb driver to the latest with 4.35 but it said current driver is already up to date. Didnt really want to tear into my office computer now but tried a laptop (I3 processor) and the time dropped to 6.3 seconds. So looks like it is related to computer somehow.

    I have another nearly identical dell I7 that need a HD, I think ill try and set this up for my mill computer after Christmas. Cant wait to try it. Im assuming if I get better results with this simpleformcs test, I should be able to run a smaller facet angle without slowdowns right?

  9. #29
    Join Date
    May 2006
    Posts
    4043

    Re: 4th axis feedrates

    Hi Mark,

    Im assuming if I get better results with this simpleformcs test, I should be able to run a smaller facet angle without slowdowns right?
    That should be correct. Although it isn't clear why you need such a small facet angle.
    TK
    http://dynomotion.com

  10. #30
    Join Date
    May 2012
    Posts
    537

    Re: 4th axis feedrates

    Quote Originally Posted by TomKerekes View Post
    Hi Mark,

    That should be correct. Although it isn't clear why you need such a small facet angle.
    I guess im a bit of a perfectionist. When it was set smaller I got those huge slowdowns on my 4th axis, so I left facet angle much higher the past little while. I was doing some 3d surfacing the other day and I wanted a really good finish on and noticed a bit of a funny pattern, so dropped back to a lower value and surface finish was noticeably improved. But a couple of parts in the program slowed down, I guess due to the transfer being too slow. Looking back at all this, I've noticed this problem several times before. Sometimes while doing small helix entrys, it would feed at programmed feed for a few passes, then run slower for the rest of the helix. Nearly certain this is the same problem.

    So right now, I have to choose a balance of surface finish and performance. And im hoping I can have best of both worlds with a faster transfer rates. Plus Ill have a fully functional backup computer for the machine right ready to plug in and go. This is something I wanted to have anyway, never know when you might need it.

Page 2 of 2 12

Similar Threads

  1. Nakamura C-Axis Feedrates...
    By MaCroB in forum Uncategorised MetalWorking Machines
    Replies: 3
    Last Post: 05-11-2019, 05:52 AM
  2. Rotational axis (a) feedrates REALLY SLOW!
    By rcsimpson in forum BobCad-Cam
    Replies: 7
    Last Post: 12-02-2011, 06:27 AM
  3. More Q's! Feedrates, etc
    By binfordw in forum Benchtop Machines
    Replies: 16
    Last Post: 03-29-2010, 01:21 AM
  4. G0's and Feedrates???
    By Moondog in forum Machines running Mach Software
    Replies: 5
    Last Post: 04-18-2007, 09:26 AM
  5. X1 Feedrates etc...
    By itsme in forum Benchtop Machines
    Replies: 2
    Last Post: 08-20-2006, 01:37 AM

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •