[whiteriver]

Been having problems with my router and trying to do anything remotly
complicated. 2d seems ok. but 3d cutting just looses steps all over
the place.
Talked to Ron at Flashcut. Had them send me out a new updated board
and a 201F chip. The old one was 5 years old and a 201C chip. That
didn't fix the problem. Didn't seem to make a difference.
Any hard science on what to set the Step Pulse Poliarty, Step Pulse
Width or Enable Signal Poliarty in regards to Flashcut to Gecko?
I ran a 67,000 line 2d engraving and only lost .00025 to .03 in the 3
axis's. I then ran a 16,000 line 3d engraving and lost, well gained up
to a 1" in the Z and .02 to .03 in the x and y. I ran it 4 times with
different settings in the Signals tab and slowed the (start/stop
Feedrate inches per min) to 5 from 15. No help.
Have Nema 42 motors. Gecko 210 drives running in full micro stepping
mode. +5v common. X is a rack with a timing belt 1.535 screw revs/"
Y and Z are ball screws, 2 screw revs/"
Any ideas. I'm pulling my hair out. Need to get it up and running.
Some times it looses steps. Other times It seems to gain them.

Anyone else using Flashcut to Gecko? Any problems? Any sucess? What
are your settings? What works? What doesn't.

Donny

[HomeCNC]

If you are gaining steps you can eliminate the IPM speed as a factor. If you push the machine to fast you will loose steps never gain them.

Are you sure that you have your math correct for calculating how many steps it takes to move 1 inch? Have you placed a dial indicator on your machine and told the controller to move 1 inch and looked at the indicator to see if you did move 1 inch? Do this on all three axies.

[whiteriver]

Yep its dialed in just right. It messes up mainly in 3d or 2d with alot of changes. Mainly the x and z axis. Its also random. Just a 2d profile cut it seems to do ok. Anything more and the mess seems to compound.

Donny

[balsaman]

You can gain steps if the accel/decel is not setup correctly. The motors tries to decel. and momentum of the motor/axis pushes it over steps...

Eric

[whiteriver]

I have slowed it down so slow it never can get up to speed. I have sped it up so fast it looks instant. Both ways it still screws up about the same amount. Never exact. Always random.

Donny

[ECRC]

I'm having the same random problem. I have changed everything in the setup tabs to no avail. I've also noticed that I get much faster rapids with mach 2(demo)

If I don't get it going soon I may change to Mach2.

Doug

[steveald]

When I built my machine, I took my time with the mechanical assembly, but rushed through the wiring just to see it run. I have been experiencing the same sort of problem as you, my course of action is to rewire, using shielded wire, keeping it as short as I can, etc... I'll let you know how it works out.
If the problem only occurs in 3D, maybe you need a larger power supply to handle all 3 motors under load, or perhaps your X and Z cables are interfering with each other and need better shielding. I have been using a 10A 24V supply to run all axis, and will be changing to 24V for Z and a 75V 6A to run X and Y.
Anyway, the gist of my message is that if it only messes up in 3D and works fine in 2D I would look at your wiring rather than your mechanical setup.
Also, if you seem to be gaining steps in one direction, maybe you are just losing steps travelling in the opposite direction.
Good luck with it, I know how frustrating random and intermittent problems can be.

Steve

[gberscht]

Hello: I have Designed a built 3 router tables and use Flashcut systems with them. I was having the same trouble with the machines not running very smooth.

If you are using any of the input lines for things like homming switches etc. you must use a cat 5 cable to clear any noise problems. There is also a certain paired order you have to use,I can send it to anyone that needs it. This should clear up a lot of problems.

A way to test if this is your problem is run your machine with the cabling you are presently using for inputs and get a good feel for how smooth it is running. Then turn off the power and unplug the input cable and run it again. Always use the same gcode file and the same speeds. If there is a change in how your machine runs then you should change to a cat 5 cable. You will notice quit a change.

Something else that shoud be done is to run a ground wire from the signal generatior to the control box this may help with noise and will stop the machine from creaping when sitting idol

As for losing steps in 3d and not 2d , there can be a few reasons for this.
eg. size of motors equals speed , if you look at a 3d gcode file there a big dif ference than a 2d file, the distance between moves can be vey small especialy on the z axis.

There is constant direction change and if you push the speeds and the motors or drive system are to small then you will most likely lose steps. something that you may try if possiable is run a small 3d file then scale it up as large as you can and then run the file at the same speeds and see if it makes a difference.

Hope this helps,

Greg

[whiteriver]

Greg,
You posted:
If you are using any of the input lines for things like homming switches etc. you must use a cat 5 cable to clear any noise problems. There is also a certain paired order you have to use,I can send it to anyone that needs it. This should clear up a lot of problems.
Are you talking about running your input lines thru your Motor Signals Plug. I run my inputs and Motor Signals seperatly.
As for size of motors if anything my Nema 42's are toooo big. This Is a retrofit on a old thermwood router that was designed to have about 800 more pounds on the gantry than is up there now..
You also posted;
Something else that shoud be done is to run a ground wire from the signal generatior to the control box this may help with noise and will stop the machine from creaping when sitting idol
When you say control box do you mean the computer? I have the geckos set to stay at full power even when at idel so they wont move I hope.
My motors run real ruff when doing alot of direction changes or alot of line segments no mater what speed. They run smooth in straigt lines or arcs or circles. Its just when they speed up and slow down.
What are you Feedramp ratings,start stop and countuinous countoring set at and what size motors,voltages and amprages are you running.

Donny

[gberscht]

Donny:

The inputs out of the signal generator a molex plug on the back of th box. If you are using it try unpluging it and see if it makes a difference.

I have Flashcuts complete system, I run a ground wire from the signal generater to the control driver box.

My system is a 3.5amp micro steper drive with timing belt gearing.
424 oz motors.
start and stop is at 25 ramping 2000 and continuous contouring 30in/min.

what are your settings.

how well can you cut a spline?


Greg

[steveald]

Hi Greg,

I am running Flashcut as well, and having problems in one axis. Can you tell me the method you use for wiring the twisted pairs to the home switches please? Also, if I turn on the driver box before turning on the signal generator, my motors go all over the place. I think I read somewhere that the signal generator needs to be on first, but sometimes I forget. Is it normal for the driver box to behave like that?

Thanks,

Steve

[whiteriver]

I always turned on my drives and let them warm up first then turned on the SG. No problem with the motors moving. Everything stayed tight.
I finnaly got my lost step problem fixed after a year of fighting with it off and on. I got rid of Flashcut and installed Mach2. My machine is dead on even after running for 6 hours and 500,000 lines of code at 150/ipm. Runs smoother and quiter and 4.75 times faster than with Flashcut.
I turned my 8000lb paper weight into a $50,000 Router by switching from Flashcut to Mach2.

Donny

[steveald]

Hi Donny,

I'm glad you have things working, although I'm disappointed you had to get rid of Flashcut. I really like the interface. Which version of the software were you using? Did you try 1.x or just 2.x? Maybe it's a bug in the new(er) software. I have heard that they have had constant contouring problems in the past. My motors go postal without the SG turned on, so I have that problem to fix. When I get that looked after, if I'm still having problems, I'll try Mach2. Flashcut was a lot of money to just put in a box

Thanks for sharing your experience,

Steve

[stanleyr]

Any hard science on what to set the Step Pulse Poliarty, Step Pulse
Width or Enable Signal Poliarty in regards to Flashcut to Gecko?

Anyone else using Flashcut to Gecko? Any problems? Any sucess? What
are your settings? What works? What doesn't.

Donny

I know this is more than a year later. If others are having a problem with FlashCut this may help.

Step pulse polarity and continuous contouring (for 3D) will have a big effect on lost steps. I found using the high polarity was a problem with the G210s. Low polarity was a vast improvement.

Set the continuous contouring down to a very low value along with your stop/start rate. Start both of them at 5 and work up in increments of 5 until you see you are skipping steps. Back off to the previous value or somewhere between the current value and last value.

A G210 needs a minimum of 4us of pulse width. You can set FC to 6us of pulse width.

If you use the dynamic enable feature, (I think) the disable input wire will ground out at the G210 and it will shut off the power to the stepper motor - causing lost steps.

Other items will be the quality of the stepper motor, the voltage you run the motors at, and the amperage. The G210's can handle a lot of voltage if your motors can. You can read about this on Gecko's web site in the 'Support' page: The Step Motor Basics White Paper: Power Supplies (page 2) http://www.geckodrives.com/ycom/docu...r_supplies.pdf. The literature will indicate running the G210 voltage up to 25 times the motor winding voltage (BE SURE YOU UNDERSTAND WHAT THAT MEANS). I am NOT talking about a 24V motor times 25. Rather, it is a formula based on the resistance of the winding and the rated amperage. Anyway, you can try running at 20 (for safety) times the winding voltage to see if that improves your torque too.

Don't know if this would have helped before you went to Mach2, but it is good that you are happy with your system now.

[cncadmin]

I just got my system today, and plan on using it with gecko steppers and servos drives.

[stanleyr]

Just ran my system FC with Gecko and open loop steppers @ 10uSteps this weekend after spending some time with the motor/driver settings (voltage and amperage). Ended up with 60,000 lines of contouring code and the motors did extremely well - within .0005 at the end of the day.

I also wanted to tell you another way of setting "Continuous Contouring" that is slightly more scientific than just bumping it up from the start/stop feed rate value:

Run some code with predictable contouring in it and listen to the motors. Lower the Continous Contouring (CC) value until the motors start to hesitate and chop their way through the contour. Change the CC value to a higher amount in small increments until you hear your motors blend together without hesitation through the contouring movement. As long as the motors are blending smoothly and not missing steps, that CC value should be good.

My CC values are set at 3.85 for X and Y and 2.75 for Z. My machine is not a barn burner (25 IPM for example tops at 10uS), but is suitable for my needs.

[cncadmin]

What is the fastest IPM that can be run for a stepper set-up?

[stanleyr]

That depends on many things such as the mass of the axis movement; the thread pitch of the driving screw; the gear reduction; the software's capabilities of generating pulses to command the motor drivers; the minimum pulse width necessary for the motor driver, etc.

I have a 10 turns/inch lead screw, with a 2.5 to 1 reduction. So my machine is really set up for fine movements. My stepper must turn the equivalent of 5000 steps or 50,000 micro steps at the 10uStep setting just to go 1 inch! . So you can see that in order to go 20ipm on my machine for 1 axis, the software must generate enough pulses to produce 1,000,000 uSteps in a minute or 16,666 uSteps in one second. That's (flame2) !

Back to your question though: If you think of it in RPM, my stepper is actually turning 500 RPM when the axis is at 20ipm. You can do your own math as to what your axis speed will be. But my motors are Nema 34 motors and are perfectly happy at that speed for rapid movement. This is on a Wabeco benchtop mill with a 80lb Kurt vise on the xy table. I generally cut at 5 to 15ipm. After many hours of running, the x and y motors peak out at 125 DegF (65 DegF rise) and the z motor reaches 145 DegF (85 DegF rise), measured at the case. This is well below the (flame2) point!

I have run the motors at faster speeds than this by reducing the micro step setting at the driver - say from 10uStp to 5uStp or 2uStp. I was able to (rapid) move near 50ipm at 2uStp, which I believe comes out to 1250 RPM, 500,000 pulses in a minute or 8333 pulses/second :rainfro: . I don't need to run this fast (saves the ACME screw and nut for more important things); and although my machine ran reliably at that speed, it does increase the chance of resonance and lost steps for my particular machine ; so why bother?

I found through trial and error that my setup behaves much better when I use the 10uStp setting; higher motor driver regulated voltage of 50-55VDC; and incrementally increasing amperage settings - keeping a watchful eye on the motor's case temperatures.