I'm cloning my y-axis but I'm running into issues

Projects General Guidance
1

Hello all, I am having issues cloning my y-axis on my CNC shield. I'm using drv8825 drivers. I have four of them attached to my nema 23 stepper motors.

Obviously my shield is plugged into my Android Uno. Which is running grbl latest version as of today.

My CNC machine is a 1500x1500 mm workbee. With tingle tension rods

When I'm running my G-Code using Universal G-Code sender everything is fine for the first 5 minutes of the cut.

A few minutes in I realize that the y-axis is making some grinding noises. And my cut shifts varying three to five inches, it's never the same, and it may complete the cut in the correct spot or not.

Anyways the reason I'm here, is to seek some advice on how to problem solve this situation.

(((( The situation is; that my cloned y-axis stepper motors are running at different speeds one of them is opening up as much as two inches over the course of a 10-minute cut why could this be? ))))

The last cut that I ran where my y-axis scrambled. I decided to measure where the end of the y-axis is at the front of the machine to the face of the x-axis. The left front corner distance was 15 in. And the right front corner was 13 and 1/2 in.

As you can see there is an inch and a half difference in both of these distances. So I checked the square of the frame on my machine. It is exactly 89 in going from axis to axis so it is perfectly square.

This leads me to believe that one of my drivers is performing faster than the other.

So I calibrated my potentiometer and got my y1-axis and A-axis potentiometers set at exactly at 0.988 v.

(Should they be higher or should they be lower?) I know that they can max out at 2 volts but I'm not sure based on vlogs and other documents where they should be, I know that it states that if they trip the motors, turn them down and if the motors don't move then you should turn them up. I am assuming though that since there are two drivers and the axis that is being cloned, using jumpers on my y-axis, should be calibrated the same using the potentiometers so that the stepper motors perform congruently.

I know for a fact that my G-Code is good. It is not a code issue it is a machine or controller issue.

Possibly there is some information that you can pass along or some elucidation because I am out of options and at a loss. . .

Thank you in advance fellow champions!

2

Nema17 only tell about the mechanical mounting of the stepper.
The firdt to check is the G-cofe sent.
What is the load to the CNC?
Have You trief runming tjst G-code eithoutvlosd, toll up in the air?

3

I made a typo to the original post. My motors are Nema 23 steppers. I am not sure if this changes anything.

4

I haven't. I was hoping someone could give me some ideas before I run the machine again. I don't want to burn out my machine.

You're suggesting that I run the g code with the spindle up in the air?

I was also thinking it could be RFI and I guess if I run the code without my spindle running, I could find out if it is RFI from the water cooled spindle or not.

Anything else you would suggest?

Railroader:
Nema17 only tell about the mechanical mounting of the stepper.
The firdt to check is the G-cofe sent.
What is the load to the CNC?
Have You trief runming tjst G-code eithoutvlosd, toll up in the air?

5

NEMA 17 and 23 are measurements of the mounting holes.

They are not a description of power.

The "grinding" noise is a condition of too much resistance over motor force. It is also known as missing steps.

It is a result from....
Too fast acceleration
Too little motor force into too much resistance.

Your optimum stepper power will be by using 32 times more voltage than the nameplate. You should run at least 10 times higher voltage that nameplate.
However the cheap steppers have really hive nameplate voltages often as high as 5 volts or more.
That means you are limited to the voltage the driver can handle.

And in krrpig with this thread...
Your Android UNO shootd bre cohtcshb
Fine

6

Try running the machine without a cutter and without a part.

Try moving each axis all the way from end to End by hand to test for smooth movement and no resistance.

If you have leadscrews spin the screws by hand.

Check the GG-cofe on your Android UNO to make sure you acceleration ramp is not too ...
eithoutvlosd

If the ramp speed is too fast a stepper can miss steps.

The pulse has to be smooooothe......
No delay. The pulses to the motor must be uninterrupted for smooth wavefofms.

The ramp speed has to be set properly

And make sure your tingle tension rods ade working.

7

dave-in-nj:
Try running the machine without a cutter and without a part.

Try moving each axis all the way from end to End by hand to test for smooth movement and no resistance.

If you have leadscrews spin the screws by hand.

Check the GG-cofe on your Android UNO to make sure you acceleration ramp is not too ...
eithoutvlosd

If the ramp speed is too fast a stepper can miss steps.

The pulse has to be smooooothe......
No delay. The pulses to the motor must be uninterrupted for smooth wavefofms.

The ramp speed has to be set properly

And make sure your tingle tension rods ade working.

Thank you for the information.

What about the six pins underneath my drivers. Someone told me to put two jumpers on the last two pins. I believe this makes my steps 1/8 step. I am not running precision cuts. I make signs and hobby stuff.

Should I change my steps?

Could this be why I am allegedly missing steps?

Could you possibly explain that process to me in more detail that would be extremely helpful! I believe I was misguided in my previous setup.

Thank you

8

themckeels:
. I believe this makes my steps 1/8 step. I am

Should I change my steps?

Could this be why I am allegedly missing steps?

Yes. Use full step mode for maximum power to axle. Multistep increases precision but looses power.

9

A stepper motor has multiple motor coils. The electronics shift them off and on and off and on.
In the case of a bi-polar, they reverse the power into the coil and then switch it back to positive.

the smooooth pulsing makes the motor have power.
try running down hill, then, every 5th step, stop for 1 second.
you will fall on your face. YOU missed a step !

Steppers require a proper sequence of pulses and cannot tolerate missed pulses.
so, need to check your G-code.
I assume you are ramping up the stepper and ramping it down. this is in software and most of the libraries take care of that for you.

a note about microstepping. a full step is magnet to magnet, holds very strong. a half step is having a magnet halfway between two others. sort of a balance. as you move into quarter steps or more, you have much less force to hold the armature in place. microstepping does not improve resolution as the step location is not reliable.
micrcostepping does improve the acceleration at lower speeds. at higher speeds it gets in the way and causes an artificially low 'top end' for RPM
we live in the start/stop world, it does offer some benefits, but is not magic.

most NEMA32 steppers are double stack (like two pancakces) single stack and triple stack are a bit less common.
also, in the NEMA32 you will often find nameplate voltages in the low 1 volt range. these are great as you can use 24 volt power supply and get much more power out of them than the exact same motor with say, 5 volt nameplate.
Voltage is limited to the ability of the driver, most are limited to around 24 volts.
you can get some nice power drivers rated for 5-10 amp and 70 volts or so.

Theoretically, at 0 speed, the power of the stepper is at maximum. at high speed, it is weak, but number of pulses per second add up to the power provided.

steppers have a maximum speed based on many things. as long as you do not get to that point, all would be good.

Check the movement of each axis. if you have a lead screw, spin it by hand and feel for any resistance that is unexpected.
It is not uncommon on a new build to have a tight spot that needs to be adjusted.

the 'grinding' sounds like the motor is missing steps.

as was noted, try a dry run. no parts, just run the table through the paces and let it run the paths of the g-code.
If that works, then increment the work load till you can find the problem. too deep of a cut ????
too fast of acceleration on a direction change ?

10

Thank you very much for your detailed response. It is extremely helpful.

Dave nj - how do I lower my ramp speed?

Is that something I can change in universal g code sender under $$ commands? If so, what $ command is it?

I am going to run my machine without a spindle and see how it runs. I'll let you know the progress.

11

themckeels:
Is that something I can change in universal g code sender under $$ commands? If so, what $ command is it?

I am going to run my machine without a spindle and see how it runs. I'll let you know the progress.

I´m pretty shure that there are G-code parameters handling both speed, acceleration, deceleration etc. Take a dive in a G-code manual.

Yes, run Your CNC around, X and Y, large distances, without doing any cutting job. Tell how it works out!
You can also create patterns having lots of starts and stops during a large distance. Such a test outgh to tell if there are acceleration problems at this level.
Of course, everything is harder when a cutter is cutting but that will be next checking.

12

so far, we are just guessing
you said 'clone' that would imply the exact same motor, the exact same driver with the exact same settings using the exact same leadscrew, gear ratio, linear guides, et-al.

what you did not say is the voltage of the power supply
the setting on the drivers
the make/model/voltage/amps/resistance of the motors

the drivers have a voltage rating you do NOT want to exceed
the power supply have a current output that you DO want to have more of. often power supplies do not output what the label says.

then there is your g-code. I am not familair with your software, but usually there is a maximum ramp speed you set in the software. then you can change that in your G-code as the cut progresses.

if you know you can run for 5 minutes before the problems, then copy your code, cut out the first 5 minutes and run from that point.

if your power supply is starting to fail after about 5 minutes, then you are just not delivering the proper power. ditto the heat sinks on the drivers. Running a stepper on high power, you should have discomfort when putting your hand on the motor. it should feel so hot that you would not want to leave your hand there for 30 seconds. That is trying to push the maximum out of the stepper, few people outside of the CNC and industry worlds push them that hard.

After 5 minutes, your driver or your power supply might get hot and start to reduce output.
these are normal bugs you find as you shake out all the problems.

it would not be uncommon to find you are running along on one axis say at 500 steps fixed from a side, with the other axis moving at 1,500=2,000 and you somehow swapped values and both axes try to move for a moment. this would jamb the cutter into the work. hence the desire to 'cut air' and watch the path.
many software packages allow you to cut pixels. run the path on the computer screen to watch the tool path.

also, you can find videos of using a marker on paper. a simple sleeve with a spring and a marker. you draw the path on paper without too much lifting and dropping of the pen.

13

Follow up:

Hello everyone,

I found the problem/problems.

What I did to find the problem:

  1. checked all my stepper motor wires to find any loose wires. Found TWO loose wires one of which was on my Y axis.

  2. checked the power on my potentiometers on my stepper drivers. After doing some research I found the the power is x2 and my drivers should and they should not exceed 1.5 v, that means that the potentiometers should be set at 0.75v. so I changed them all to .75v

  3. removed all jumpers for micro stepping. I was at 8 micro steps but I changed my drivers to half steps. It was a little rough at one step. Too much vibration, but at half steps. I didn't loose power (I achieved this by placing a jumper on mo1 on my cnc shield underneath my driver.)

  4. I installed a small fan blowing right at my drivers. This is a small 5 volt fan that I bought for under $5 bucks. (Before the fan after ten minutes of operation the drivers would be hot to the touch. After the fan and power adjustments after running for an hour, there was no heat on my drivers. It is amazing how far a little wind goes.)

  5. once I made all my wiring and power adjustments I moved to the program. Checked my settings on Arduino using Universal G-Code sender under $$ setting.

  6. calibrated my steps per inch. $100-102. My settings have me about 50 steps

  7. also changed my acceleration on my motors can't recall which $ setting this is but you can locate it if you look at the Grbl handbook online. My acceleration was set to 150. Which was too high because I was missing steps. So I reduced it to the factory setting 50. (I ran the machine and it did start a little slower but I could see the difference in smoothness, I could probably increase this but I am okay with the slower acceleration if this means the machine will run g code without missing steps.)

  8. changed my post processor in VCARVE from xcarve to GRBL in.. this didn't seem to change much, but reviews say that this is the best option for my current setup.

  9. the RFI issue that I was having my hard limit switches being triggered without and contact to the limit switch. To solve this I pulled the spindle power cord out of the drag chain because it was running parallel to my stepper motor Cables and my limit switch cables. Bought a pack of ten fertile filters to reduce RFI. Placed one on every cable on my cnc machine. Probably overkill, but precautionary. And I put a hook in the ceiling and now have the spindle power cord suspended from the ceiling where it is not touching any of my other cables. This seemed to have solved my hard limit switch problem.

There is still one issue I am having where my cut depts are not uniform. Someone did tell me to check the table my cnc is on and make sure it is perfectly level at all corners of the machine. I believe this may be the issue because my floor is far from level.

Thank you again for your advice and assistance.

14

Well done finding those error sources!

What down forces do You think affects the milling bit? Is there a play in the mechanics in the z-direction (up/down)?
Could the material have "hard spots"?
Give some data of material, steel, aluminium etc., RPM, cutter dimension, cutting depth, cutting width, nr of flutes etc.

I'm playing with milling, turning, cutting and maybe I know something.

15

I am cutting 3/4 dry oak. Using two different bits 9/16 vbit. And 1/4 bullnose two flutes.

There isn't any play in the z axis. The wood that I was cutting seemed to be uniform.

My spindle maxes out at 24,000 rpm but I believe I am running at 16k.

The 3/4 oak was cut in 6 passes.

Thank you

Railroader:
Well done finding those error sources!

What down forces do You think affects the milling bit? Is there a play in the mechanics in the z-direction (up/down)?
Could the material have "hard spots"?
Give some data of material, steel, aluminium etc., RPM, cutter dimension, cutting depth, cutting width, nr of flutes etc.

I'm playing with milling, turning, cutting and maybe I know something.

16

Okey.
I'm not an old experienced machinist, have a few years of experience from metall milling, turning, cutting.
I don't get the full picture anyway. Oak is a quite hard material. Have You tried the same setup, the same G-code in a more soft material?
Try and lower the RPM, cutting depth, horisontal cutting speed etc. and check what difference that makes.
Of coarse, if You can, You could try and debug the stepping made and se if there is any jitter in it but I think that materials and machine parameters are the reason before electronics.

17

themckeels:
everything is fine for the first 5 minutes of the cut.

A few minutes in I realize that the y-axis is making some grinding noises. And my cut shifts varying three to five inches, it's never the same, and it may complete the cut in the correct spot or not.

The 'grinding' noise sounds like one or more motors are losing steps. This could be binding, causing the normal drive current to be insufficient or the drive current is set too low for the motors. The driver potentiometers set the drive current.
If you turn off stepper power you can feel for binding by moving the axis by hand.