3020 China mill ?s and issues.

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
ive decided to stay with ugs and see if I can work out the issue. I'm sure candle and easel are good programs, they are just not working for me, so after an initial boil down of where the problem begins , it seems to go offline when it wants to start the actual cut cycle. It will move to its start point, raise and lower z axis, then maybe one circle rotation, then go offline. ( square pocket with circle cut start)
I can jog fine in all directions, set zeros, return to zeros, all with no disconnection.
After it disconnects the spindle will run continuosly till I remove the 24 v power. If I plug in 24v without unplugging the usb , the spindle will start again and run continuously. Once I remove both 24v and usb, and plug back in , I can now reconnect with the board thru the computer.
I had a similar issue with the grbl and if I dismissed the first tool change it would run till it found an error. So my ? Is, could this be an issue with ugs that I'm missing
 
As of now I have not. I did remove a few zip ties , as all of the wires from the motors are bunched together very tightly. And with such a small machine, it appears no shielding.
My limited understanding of shielded wire would entail wrapping alum. foil around the wires so,, today I will go thru the grounds and read up on shielding and report back.
Maybe in the long scheme of things I can learn something and maybe others also, with this paticular issue. Thanks for the direction.
 
try twisting the wires to the steppers and to the endstops - if this does nor help then go for shielding.
if this was good for LAN then is must be as well for our steppers.
 
Yes, other options than aluminum foil, starting with Peter's suggestion. I don't actually know how much or whether that would help with stepper motors, since you don't have a ground per se, but it could definitely be helpful for any limit switches.

Make sure the power supply is not right up next to the electronics - cheap power supplies can be noisy. You may also want to look into getting some ferrite chokes / beads to put on the power lines (maybe also the signal lines) - I'm out of my depth here on exactly how to specify what to use, but if it were me, I'd pick up a cheap "kit" from Amazon or eBay or such and start experimenting.

Here's a quick example - absolutely no endorsement implied, nor any experience with the vendor; just the first one that came up in an Amazon search: https://www.amazon.com/Taigoehua-Su...9Y2xpY2tSZWRpcmVjdCZkb05vdExvZ0NsaWNrPXRydWU=
 
I am almost sure that your problem is in the grbl configuration.

Has your machine limit switches?
Are they set ON in your configuration file?
What happen with the soft limits? Are they set ON too?
Again, sending a $$ comand to the board will tell you all the status.

You can read all the info about this following this link: gnea/grbl

Regards!
 
Well gentlemen I am glad to report I have successfully ran my first cnc program. After reading up on shielding practices, I separated and twisted all the wires. It was literally a bird nest. So I ran the program and it did make the beginning cut, then shut down. But was further thru the code than any other time. Then I noticed the error report was all based on Java errors. While I have no idea what they meant, I went ahead and updated Java. Then she worked all good.
On the same note I wonder why since I had downloaded Java, just before starting per instructions
But two questions
1. Does the term noisy mean the same as emi interference Or is it a literal audible noise
2. The program ran fine except the Z axis zero was incorrect. On about the 4-5th pass was when we got to the wood. I have tried to set zero the same as you would manually , but it will start about 1/8 inch higher.
 
Great to hear of the success! I have not had enough hands-on experience to know what is going on with the Z height ... I would guess that is set in the UGS or bCNC or similar "g-code sender" program.
 
Well gentlemen I am glad to report I have successfully ran my first cnc program. After reading up on shielding practices, I separated and twisted all the wires. It was literally a bird nest. So I ran the program and it did make the beginning cut, then shut down. But was further thru the code than any other time. Then I noticed the error report was all based on Java errors. While I have no idea what they meant, I went ahead and updated Java. Then she worked all good.
On the same note I wonder why since I had downloaded Java, just before starting per instructions
But two questions
1. Does the term noisy mean the same as emi interference Or is it a literal audible noise
2. The program ran fine except the Z axis zero was incorrect. On about the 4-5th pass was when we got to the wood. I have tried to set zero the same as you would manually , but it will start about 1/8 inch higher.

Electrical engineers consider EMI interference as noise. You won't hear it but you can see and measure it with an oscilloscope

The Z axis starting location is set in your G-code. You'll have to look for it.
 
After fiddling with different settings we are pretty much back where we started. I been successful with 2 runs of a circle at .20 deep. The program ran full, no stops, no errors. But everything else I have tried ends early. All the error reports are related to Java runtime environment. Granted there are others also but Java is always the last one up to the disconnection of the board from ugs. Any fixes seem to be diving head first into javas code on a hacker/forensic level. Which I am not capable of, or care to learn.
Also when setting my zero which matches in fusion and ugs screens, but when program starts, it will go opposite tracking than what it shows. So say zero is my left front corner, when the program starts it will move from zero, all the way to the opposite corner then place itself for start cut.
 
I suggested you to try Grbl Panel. This software is wrote in C++ and is very good and run flawlessly. You can find the link in the post #13 of this thread.
I would like that you give it a try because I am almost sure that this will contribute to the solution of all the Java problems. No Java at all.
Other point to consider is that Grbl likes the zero machine position in the right/back corner. Also, take in account that Grbl do not support canned cycles.
I can post a GCode program that surely work with Grbl so you can do a test with something that you know is working. Just tell me if this may helps.
 
South Houston is my local, but I don't know anyone who could help, but I am usually a diy'er to my bones so, off we go.
Ok have tried grbl. As of now I cannot connect the jog controls Error states undefined feed rate. Gonna try to fix here in a few. Also noted was, my file would load and run the machine, but without being able to place zero yet , it quickly overran the axis limit. But we know we are connected ... Somewhere.??
 
my understanding is that in the grbl-settings a change from $23=0 to $23=3 would effect a change from 'home top right' to 'home bottom left' - give it a try!
 
a follow-up to my last post -
- have you seen
the first minute is all about installing the newest version of Java, but after that the whole Grbl-thing is explained in detail, like
- setting connection port and Baud rate
- activate homing
- set home position
- set start position
- how to change directions on one or more axes,
- activate endstops
- activate a depth gauge (on pin A5 if you have a woodpecker board)
and so on.

You should read it up on Github too, and print it out for reference, BUT I do recommend to watch the video and make notes -->
I myself am still in the process of building up my machine, a 3018 pro max - by the way one of the best builds of these small routers, being built entirely from extremly stout profiles (2040,4040) and 10mm plates and 12mm rods with no plastic at all. At the time being I was printing cases for the 6 endstops and the new handheld/offline controller with the 1.8" LCD screen, but had to stop because my printer has an issue with a blown stepper driver....
1596880292443.png
 
South Houston is my local, but I don't know anyone who could help, but I am usually a diy'er to my bones so, off we go.
Ok have tried grbl. As of now I cannot connect the jog controls Error states undefined feed rate. Gonna try to fix here in a few. Also noted was, my file would load and run the machine, but without being able to place zero yet , it quickly overran the axis limit. But we know we are connected ... Somewhere.??

You said you have tried grbl - did you mean Grbl Panel, as Leonardo suggested? If not, I really like his suggestion as a way to eliminate the java machine as a possible suspect. But if these results were referring to Grbl Panel, it sounds like it is missing some parameters, but otherwise showing some promise. Keeping my fingers crossed ...
 
After some more research tiny accomplishments were made. Although reading up on the grbl settings for $100-$122 are quite puzzling Some state the math being a high number such as 444, and some suggest low numbers like 40 for 100-101-102. Then for the 110-111-112 & 120-121-122 are experiment till happy.
What I have is an 8mm leadscrew with 2 mm lead. 200 steps per rev. With 8 micro steps Or for one revolution I move 3.62 mm
Under the G00 x? Command all my axis move sufficiently. But when I try to move with the jog it is very slow on the z axis---- but the x and y-- it stated undefined feed rate error.
So any help on numbers for a stock setup would be great. Limit switches are on the way soon to help grbl in that area.
 
To be sure of the pitch of your lead screws I would take a reference point and then move the axis (motor) by hand one revolution, then take a measure to know exactly what was the displacement.
If your stepper motor is 200 steps/rev. and you have the driver set at 8 micro steps and you have a displacement (pitch) of 2 mm per rev. you should set the $100 and $101 parameters at 800 (200 * 8 / 2).
It might be possible that your machine has different lead screw for the Z axis so you may want to check the Z axis pitch as well. Then the counts are the same.

Regarding the $ 110, $111 and $112 they are talking about how fast your motors can rotate without loosing steps and yes, this is a trial and error procedure to tune your machine at the best performance. This parameters also depend on variable conditions as how much is the weight of the parts on your table and how much is the weight of the tool head for the Z axis. Actually you are telling the controller how fast it can send the pulses to the motor (frequency). If your machine has NEMA 17 motors you may want to start low, perhaps around 500 or so. In the case of the NEMA 23 you could try 1000.

As for the $120, $121 and $122, these parameters set the acceleration. This is the ramp at the start of the movement. If you set this parameter too low the machine will not reach the maximum speed in short movements because it will be still accelerating at the time of arriving to the end of the movement. Again a trial and error procedure. Heavy load and small motors call for lower acceleration and speed rates. Anyway I would start setting these parameters at 100.

Having limits switches is a must because the controller needs to know where are the zeros to start counting correctly. Also you need to unlock and make a home cycle at the start so the limit switches should be there to be able to do that. ($21=1 (Hard limits enable, boolean)).

I also like to have the Soft limits enable ($20=1) so the software can check that the work is not exceeding the working area. These parameters are set with the $130, $131 and $132 (axis maximum travel). Anyway this is not strictly necessary.

Just as a reference, my machine has 13 mm lead screws with 5 mm pitch for X and Y axes and 8 mm lead screw with a pitch of 2 mm for the Z axis. 200 steps NEMA 23 motor and 24V power supply. The driver is set at 1/16 micro-steps. For this mechanical configuration I set the grbl parameters as follow:

$100 = 640.000 (X-axis travel resolution, step/mm)
$101 = 640.000 (Y-axis travel resolution, step/mm)
$102 = 1600.000 (Z-axis travel resolution, step/mm)
$110 = 2500.000 (X-axis maximum rate, mm/min)
$111 = 2500.000 (Y-axis maximum rate, mm/min)
$112 = 1000.000 (Z-axis maximum rate, mm/min)
$120 = 250.000 (X-axis acceleration, mm/sec^2)
$121 = 250.000 (Y-axis acceleration, mm/sec^2)
$122 = 180.000 (Z-axis acceleration, mm/sec^2)

You can take a look at my machine following this link: CNC controller

More on the machine here: CNC controller

Hope all this help a bit more!
 
Last edited:
I am able to get the machine under GRBL , to move under the g0 commands as expected
The jog does work for the z very slowly. It shows the correct amount of movement for the jog steps, just crazy slow. but the x,y, keep reporting errors of undefined feed rate.
In the grbl panel I have changed the values numerous times, and while they work with g commands, at a good rate, anything related the the jog gives undefined feed rate.
In the video you mentioned,he uses ugs as the command center for grbl. And uses the jog buttons on ugs to test. If I use UGS I have an expected jog movement. But not using the grbl program alone , again, undefined and slooow.
As of right now this seems to be my only issue. It will run a gcode program thru grbl alone at the rates set by fusion but as far as GRBL jog no dice.
I did find one site related to this issue, again his problems were solved thu the UGS platform by upping the steps from 1 to 1000. I tried this with grbl, and the z did move slowwly, trying for 1000 steps so no change.
Here's my setting right now. I will re enter the numbers you suggest and see if it helps. I'll know in a fewimage.jpg
 
Naiveambition. the setting of a machine like this is quite simple but we need to know the correct mechanical specs.

Again, if your stepper motor is 200 steps/rev. and you have the driver set at 8 micro steps and you have a displacement (pitch) of 2 mm per rev. you should set the $100 and $101 parameters at 800 (200 * 8 / 2). So $100=440 and $101=440 are wrong values.
Regarding the Z axis, what is the pitch of this lead screw? What is the driver's micro-step setting for this axis?
You have the $102 parameter set to 55, this seems to be wrong too.
Also $112=50 is way too slow, thus the problems you are having with this axis.
 
Back
Top