New DIY Router Cutting Issues

[charlie1126]

The spindle info is below. The power to the drive is 36V 13.8Amps. I put the dip switches to the max amperage on the drives.

Frequency:	400 Hz	MPN:	UNBSU007
Speed:	8000-24000 R/min	Model:	1.5KW
Runout off:	less than 0.005mm	Country/Region of Manufacture:	China
Input Voltage:	220±15%	Inpute Voltage:	220V
Output Voltage:	-208-240VAC	Outpute Frequency:	0-400 Hz
Input Frequency:	48-63 Hz	Diameter:	65mm x Length:188mm
Output Frequency:	0-400 Hz	Power:	1.5 KW
Output Phase:	3 phase	Voltage:	220V~250V
UPC:	283961317893	EAN:	0283961317893

[Biggs427]

What's the inductance of the motors? This value + voltage to the drive will tell how fast your stepper can spin.

What kind of linear motion do you have? From what I see in your GRBL config it looks like 5mm pitch ballscrew.

15mm/sec2 is really low and if I happen to be right, 1200mm/min should be 250rpm shouldn't be too hard to obtain. Unless your steppers inductance is really high.

Have you tested to decouple the stepper from the linear motion and spin it by hand? Maybe it binds somewhere.

[skrubol]

Ok. I just used meshcam to convert the gcode and for like a minute it looked like it was doing everything right. Then it had a mind of its own. X and y motors were really hot. Not sure if that's normal. Maybe the drivers

The spindle info is below. The power to the drive is 36V 13.8Amps. I put the dip switches to the max amperage on the drives.

You might be running the motors too hard. Try backing off on the current a bit. Too much current will overheat the motors and reduce their power as well. Do you have specs for the steppers? What is the max current for the TB6600?

[charlie1126]

I am using a linear kit that I bought on ebay. This is what is included.

2 x SBR20-300mm Linear guide

2 x SBR20-600mm Linear guide

2 x SBR20-1000mm Linear guide

12 x SBR20UU Linear blocks

1 x SFU / RM 1605-350mm- C7 ballscrew with nuts and end machined

1 x SFU / RM 1605-650mm - C7 ballscrew with nuts and end machined

1 x SFU / RM 1605-1050mm- C7 ballscrew with nuts and end machined

3 x BK / BF 12 support for ballscrew

3 x 6.35mm*10mm Couplings

[charlie1126]

I have tested the motors by spinning them by hand and they spin fine. No binds.

Shaft diameter 8 mm; Shaft length 20 mm; Body Length 76mm; full length 97mm; Step Angle 1.8° Step Accuracy 5% Holding Torque 1.9N·m (269oz·in) Rated Current/phase 3.0 A Phase Resistance 1.2 ohms Rated Voltage (≠driving voltage) 3.6 V Inductance 3.7 mH ±20%

4amps is the max current.

What's the inductance of the motors? This value + voltage to the drive will tell how fast your stepper can spin.

What kind of linear motion do you have? From what I see in your GRBL config it looks like 5mm pitch ballscrew.

15mm/sec2 is really low and if I happen to be right, 1200mm/min should be 250rpm shouldn't be too hard to obtain. Unless your steppers inductance is really high.

Have you tested to decouple the stepper from the linear motion and spin it by hand? Maybe it binds somewhere.

[skrubol]

Definitely back off on the current setting then. Feeding 4A to a 3A stepper isn't doing it any good. That's 78% more power than it's designed for (4/3^2).

[charlie1126]

Ok I will. I thought I read somewhere that someone said that the TB6600 dont give out anywhere near where they say they can so thats why I changed it to the max.

What do you think I should set my acceleration to? Thanks

[highoctanecnc]

It looks to me like your motors are stalling. I tested your code in UCCNC controller and it previewed without a problem. So, that means the code is good, at least if you're using a controller that interprets the code the way UCCNC does.

Does your controller have an on-screen preview? If so, take a look at the preview. If it looks right, this is likely not a controller problem and would point more toward the stalling possibility.

If it is stalling due to lack of motor torque (for one of several possible reasons), running the mandala.txt file at a slower speed will give you an indication of that. Although, there should not have been much load on the machine at such a small per-pass depth and 55 IPM.

Before you look into all the possible reasons it could be stalling, you could rule out any G-Code interpretation problems by running a simple test square. I have attached a test file to cut a 4" square with the bit on the center line (so it will come out a little smaller than 4"). Try cutting this and see how it does. Then report the results back here. If it didn't cut well, we can proceed to diagnose the stalling.

By the way, did you hear the motors stall at any point while it was cutting?

Also, your control signals could be noisy (electrical interference). Or, your drivers could be susceptible to noise.

Are your motors wired bipolar series, bipolar parallel, or unipolar?

[highoctanecnc]

Oops. I read this thread from end to the beginning. I was posting based on the stuff you sent in on the 15th. But hopefully there's something in there you can use that will help you figure this out.

[charlie1126]

Thanks a lot. I will run that test tomorrow. I never hear the motors stall. Sometimes it works perfect then sometimes it will work perfect for a couple of minutes then decide to go way off course.

I am not sure how the motors are wired. I had to switch them around a couple times to get everything running in the right direction. I think I will re wire them tomorrow. I didn't think there was more than one way to wire them. Well 2 actually. One cw and one ccw.

I also don't think I'm using the correct spindle wire. Might be getting a ton of noise.

I will report back tomorrow night as I'm out of town tonight. . Thanks again.

It looks to me like your motors are stalling. I tested your code in UCCNC controller and it previewed without a problem. So, that means the code is good, at least if you're using a controller that interprets the code the way UCCNC does.

Does your controller have an on-screen preview? If so, take a look at the preview. If it looks right, this is likely not a controller problem and would point more toward the stalling possibility.

If it is stalling due to lack of motor torque (for one of several possible reasons), running the mandala.txt file at a slower speed will give you an indication of that. Although, there should not have been much load on the machine at such a small per-pass depth and 55 IPM.

Before you look into all the possible reasons it could be stalling, you could rule out any G-Code interpretation problems by running a simple test square. I have attached a test file to cut a 4" square with the bit on the center line (so it will come out a little smaller than 4"). Try cutting this and see how it does. Then report the results back here. If it didn't cut well, we can proceed to diagnose the stalling.

By the way, did you hear the motors stall at any point while it was cutting?

Also, your control signals could be noisy (electrical interference). Or, your drivers could be susceptible to noise.

Are your motors wired bipolar series, bipolar parallel, or unipolar?

[louieatienza]

Not for nothing, but that's some serious tool stick-out. You'd get better results using the shortest tool that gets the job done.

[charlie1126]

So I switched over from using Arduino with Grbl to using a PC with Mach 3 and breakout board. After some research and messing around I have it working properly. I cut a circle, square and other basic shapes and it was fine. Ran for a few minutes at a time cutting out the shapes with no issues.

Then last night I tried something a little more complex and it runs for like 2 minutes and when it gets to a certain spot it just runs right off path in a straight line. See image. I did this 3 times and it goes off path in the same exact spot. I just ran the code through some online simulators while here at work and they all show the exact design i want.

What would cause it to do this at the same exact location everytime? I will do it again when I get home to see what line in the code it is up to to see if it has anything to do with the code but if in a simulator it shows up right im guessing the code should be good.

The only thing is that in the Mach3 preview screen it did not show the preview. All the good ones showed it.

I also think that one of the simulators I have at home showed some type of collision but still showed the proper design path.

I will test again tonight and see what I can figure out.

[charlie1126]

forgot pic and code.

[LeeWay]

If you mean the diagonal line, then it looks like the code is sending it home. Below safe Z too, sooo, I suspect something in the gcode. Just check out the line number when it does it and post up those few lines around it.

[skrubol]

Do you have soft limits configured correctly for your machine, and did you get a limits warning when starting the code? If it's sending it waay off into space that might be why the preview didn't show, if the diagonal line was much larger than the rest of the design.
Mach has the option to single step through the code. You could turn that on when you know that line is getting near and be able to find the offending line.

[SteveWill]

Do you have a G28 in code, remove it.

[charlie1126]

The home position is right at the bottom left corner of design. That line went past it. But I checked the code and the line where it does that and the lines before and after are fine.

On another note one of my drivers blew today. Also my x axis was going whacky. I manually would bring it left and it would go left then if I clicked left again it would go right and back and forth. I shielded my spindle wires and that seemed to fix that.

I'm going to order new drivers now. Are the sainsmarts good?

Do you have a G28 in code, remove it.

[charlie1126]

I have not set any limits yet. I have switches just haven't hooked them up

Do you have soft limits configured correctly for your machine, and did you get a limits warning when starting the code? If it's sending it waay off into space that might be why the preview didn't show, if the diagonal line was much larger than the rest of the design.
Mach has the option to single step through the code. You could turn that on when you know that line is getting near and be able to find the offending line.

[Biggs427]

I'm going to order new drivers now. Are the sainsmarts good?

I don't know about the the Sainsmart but the DM556T from Stepper Online does a good job for me. I'm running it on my Z and X axis and it's smoother than the TB6600 it replaces. The price is also quite good.
https://www.omc-stepperonline.com/st...or-dm556t.html

Edit: Also, make certain that all your cables are shielded and properly grounded.

[LeeWay]

If you get unrequested motion when jogging, that can be a sign that you have loose motor or drive connections.
That situation will also kill drivers if the problem is intermittent.