Only getting 1/5th the distance for axis travel.

So I have a new setup that I was able to quickly get up and running and all was great(zentoolworks 7x7 kit). I didn't notice until I tried to run my first gcode that it was off my 1/5 on all axis. So I went back to the config wizard and tested there. Sure enough, 100mm only gave me 20mm of travel, 50mm = 10mm, etc. Most of the config was copied from their wiki. Here is the hardware:

nema 17 stepper motors 1.8deg
1.25mm pitch leadscrews
cncgeeker.com's 3 axis board TB6560 has 1, 1/2, 1/8, 1/16 microstep resolution (set for 1/8)

I just can't seem to get it to be the correct distance. I even tried setting to microsteps on the driver to 1 because that was the complete opposite setting for dip switches, figured I might be reading them upside or something.

Here are my configuration screens.











Help would be greatly appreciated I'm racking my brain trying to get this to work.

Thanks
Antoni

If you are off by 1/5 try just changing the pulley setting to 1 to 5.

After using Stepconfig to create the proper setup I often just go into my configuration file and edit the scale to get the right movement.

Rick G

Rick G wrote:

If you are off by 1/5 try just changing the pulley setting to 1 to 5.

After using Stepconfig to create the proper setup I often just go into my configuration file and edit the scale to get the right movement.

Rick G

BigJohnT wrote:

Just like Rick I just change the scale number to be what I need.

Does this scale effect the hardware communication? I guess I'm just asking if I will have to slow down my total speed to avoid missing steps or if this is just internal to the software so to get 5x scale it sends 5x steps over 5x the time? Not 5x the steps in the same amount of time.


BigJohnT wrote:

Just like Rick I just change the scale number to be what I need. All the other numbers don't mean anything only the Axis SCALE: number is used. So it looks like yours needs to be 256 or there about. But that number does not sound right for 200 steps a rev at 1/8 microstepping and 1.25mm pitch for a direct connection.

You don't mention if your steppers are directly connected to your lead screw or via a reduction belt.

For testing I put a mark on my stepper shaft and make sure I know how many steps it takes to get one rev on the stepper... then go from there.

John

Sorry, my steppers are directly connected to the lead screw. How did you go about testing how many steps it takes to get one rev? I'm curious if you did it through software or some kind of manual wiring?

Thanks for your help.

Well,

I've done some more testing with this. I tried issuing movements in axis and it was not correct. I then decided to start rulling out my pc hardware. So I used a different pc. This time I used windows 7 and mach3. I actually ended up with the same results??? The scale was still off.

So I decided to test the other side of it; the driver hardware. I wrote up a little program on an arduino I had lying around. It pins out directly to the driver controller. Wrote up a little serial interface where I can set settings like steps, frequency, and hold time. Then enable and run an axis. Turns out....It's dead on. I set the driver board for 1:1 microsteps and it worked fine. 200 steps moved it around 360deg, 400 steps for 1:2, 1600 for 1:8 etc. I then hooked up my dial indicator and stepped it 1280 when it was micro stepped for 1:8. It moved 1mm as expected.

(actually if anyone is interested, I'm thinking about open sourcing this [simple] code out on google source for other people's testing)

I must be doing some wrong with the software config. I just don't know what it is. I'm even using the same parallel cable. I'm gonna go through it again I guess.

I just start with a setting that I think is correct, then set up a dial indicator on the axis.
I go to mdi and issue a command for say a .10 inch move and note the amount the machine moves. Like John said make this a slow move so you can see what is going on.
Take the results scale =xxx and movement was .xxx" do the math and enter the correct scale, then test again.

Rick G

So the only way I could get the conf to make the correct amount of movement is with getting the axis scale to 6400steps/mm. I had to turn the vel and accel way down to .5m/s and 50 respectively. The movements are not smooth at all. It's vibrating/pulsating fairly well. There seems to be more error introduced as well. If I let it bounce back and forth around one revolution it seems to fluctuate between .003" and .010" I also bumped my jitter setting to 50k. When I had it set for 1280steps/mm the scale would be off but the movements were smooth and accurate. I'll have to compare with my uC tester because perhaps the error is being introduced by the hardware.

I understand that I can correct the steps/mm axis scale to fine tune everything. However it's still 5x off. That seems like a huge number to indicate that something is not right and it's obviously pc/setup related.

i would check your Latencey timing , also try reversing the logic of the step and direction lines , i'm sure it's timing related etc , hope it's not a Laptop of course .

try slowing down the velocity and acceleration , it could well be a combination of one or both .

i would also reduce your microsteps to 1/2 or 1/4 to start with , easier to find the problem and then increase later , although i do not see any benefit running greater than 1/2 step , with your table setup , as your driving the shaft directly .



Dave