Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: S Stewart on May 08, 2006, 01:52:09 AM
-
Everytime I run a program the feedrate tries to go back to zero between each line of code, thought this might have something to do with the 'exact stop' setting so went into the configuration setup and changed it to constant velocity and also added a G64 into the start of the code to see if this would solve the problem, unfortunately the feedrate still goes back to zero between each line of code so not sure where I've gone wrong.
-
Is this on a lathe or a mill,
Graham.
-
Also please give us a small sample of your program and you XML file :)
Thanks
Brian
-
We are running a router and producing the G-code out of Edgecam
Here is a sample of the some of the code that we are running and I have attached the xml file:
G90G40G49
G64
T01(6mm slot cutter)
G54M06
S4000M03
G0X122.216Y46.985
G43Z5.0H01
G1X198.331Z-3.0F2000.0
X199.679Y46.278
G17G2X203.314Y44.229Z-3.0R69.124
G1X204.167Y44.284
G2X211.032Y44.752Z-3.0R424.022
X216.836Y45.151Z-3.0R531.251
X229.29Y46.014Z-3.0R1400.554
G1X233.022Y46.268
X235.882Y46.457
X239.406Y46.76
X239.863Y46.793
X242.701Y46.985
X258.439
Y45.057
X254.286Y44.746
X250.342Y44.476
X246.658Y44.215
G3X239.192Z-3.0Y43.734R337.065
G1X238.755Y43.697
X236.112Y43.465
X224.053Y42.658
G3X217.045Z-3.0Y42.158R751.665
X211.234Z-3.0Y41.759R535.447
X204.372Z-3.0Y41.291R422.046
G1X202.573Y41.175
X200.562Y42.37
G3X195.796Z-3.0Y44.9R95.655
G1X193.336Y46.061
X192.248Y46.546
X191.212Y46.985
X182.711
G2X185.701Y45.942Z-3.0R93.446
G1X187.915Y45.101
X188.952Y44.683
X189.982Y44.248
G2X193.251Y42.788Z-3.0R63.297
X196.892Y40.964Z-3.0R65.726
X200.474Y38.937Z-3.0R65.031
Any ideas would be greatly appreciated
thanks
shelley
-
try to retune your motors... They look like thay are not set right....
-
Have spent alot of time trying to retune the motors and trying to run it at different speeds, but that hasn't fixed the problem..
We are able to get it run alot smoother by pulling the acceleration and velocity speeds in the motor tuning right back, but in doing this the machine is forced to run so slowly that the cutter is burning the wood. We are using Vexta drives and stepper motors and for the life of us can't get them to run efficiently and smoothly.
The feedrate itself, is still all over the show and during short arc movements the machine shakes and vibrates really violently. As it is trying to decelerate down to zero then accelerate back up to max feedrate in a short space of time.
We have been doing some feedrate tests and we can get it to work really well and smoothly on jog moves but as soon as you run g-code the feedrate accelerates and decelerates at the end of everyline went through and triple checked that constant velocity is set, but this seems to make no difference.
Almost feels as though I have ticked and unticked everything to exhaust all other possibilities and not sure what else I can do ???
-
Hi:
How abouyt the driver test run? Is it a clean waveform?? This sounds like *********ty pulse reponce. How does Drivertest.exe respond..
Thanks
Art
-
HAve run the driver test numerous times and keep getting the same result, so don't think it has anything to do with that.
The driver test run is showing:
Pulse per second in 25kh mode =23600 - 24800 or 22300 - 23200
CPU Speed= 1917
Apic timing constant= 4142
Interrupt used= 00
The waveform is pretty much flat with the occasional spike but no spike is over 5mm on the graph
Thanks
Shelley
-
Have also found that constant velocity holds when you are cutting arcs and circles but the feedrate fluctuates when cutting straight lines or point to point code so when you have a series of short lines the feedrate accelerates up to max feedrate and decelerates to zero for every line causing the extreme shakes and shudders of the machine
Thanks
Shelley
-
Shelly:
The only way to get away from that is to have a greater acceleration, Im told yours is 4? Thats pretty slow for a mm machine.. Hmm..
Try this as a test.
G1X5
G1X10
G1X20
G1X30
Do those 4 lines execute with no slowdown inbetween?
Thanks
Art
-
do you have backlash on?
-
Hi
We ran the machine with both backlash on and off to see if there was any difference but unfortunately no change, so we have left it turned off.
We have the acceleration set at 500mm per/sec and this is about where we can safely run the machine without the shudders getting too violent, so if we take the acceleration up much more than that we end up with extreme jolts inbetween each line of code.
We ran a test just doing g1 moves as suggested, and the lines ran smoothly with no slowdown inbetween each line on both the x and y axis thought we would try g2 moves
e.g. G2x50y100
G2x100y200
G2x200y300
G2x300y50
and as soon as both axis moved simultaneously the shudders and acceleration and deceleration reappeared and the feedrate fluctuated from 0 up to max feedrate between each 'g2' move.
Thanks
Shelley
-
Hi,
it sounds a bit like lack of power to the motors, how big is your power supply?
Graham.
-
Hi:
This can be the Gcode, if its full of very short vectors ( very short) it can stall the CV modes and make things jerky. It doesnt sound like CV is acting up since the straight line test runs. Make sure CV feedrate is not turned on on the settings tab, also make sure Angular descrimination on the confif logic is left turned off. Either of those may do things such as this..
Art
-
Hi
Went through the setting and made sure that the cv feedrate and angular discrimination wasn't turned on, and it was already turned off so that wiped out that possibilty.
We went through and changed the machine over to run in imperial instead of metric and it definitely made the maching alot smoother, but it isn't feasible for us to run in imperial so the only way we can get around the problem is by having the motor tuning in imperial and the visual diplay and programs in metric it seems to have made things run a little smoother but by no means great the feedrate is still jumping all over the show and accelerating and decelerating while doing straight line moves.
Are you able to give us a piece of test code that we can run as want to make sure that our post processor out of Edgecam is working properly and spitting out the right kind of code that we need.
Any help would be great
Thanks
Shelley
-
Hi:
You can test it with just importing a bmp or jpog, that generates alot of codes..
I tried your xml, but the file must be bad because the motors arent even turned on in it.
Send me your Mach3Mill.xml file. I tried running your code snipet below and it runs here without slowing, just a
bit of variabnce from 119 to abnout 120IPM.
I need your xml file to see why yours woudl stop. Waht feedrate are you trying to run that snippet at?
Art