Thread: Getting tired of not getting it right

Hi Bob,
You have made an honest machine with cost consideration and machinery that you have available. You are not doing yourself justice by comparing to others that have purchased much more expensive materials and have CNC'd machines to build them or contract that work out.

Let's start on the steps first.
As you say pitch is 1.8143mm pitch.
Your motors are 1.8 degree or 200 full steps per revolution.
According to the sheet you are 1/8 microstepping through the Kelling driver.
So calculation is 200 times 8 then divide by 1.8142 = 881.93 or there abouts.
So this is telling you that 882 pulses (microsteps) are needed to move the screw 1mm - more on that later.

Speed is controlled my motor tuning. Don't expect a 1.8mm pitch screw to run you machine fast. Stepper motors produce their greatest torque at standstill and the torque curve drops off rapidly as the motor spin faster. So you need a really fast motor to deliver speed and steppers can't do this with enough torque to move your axes. This is the reason why we go for bigger pitch screws as this translates to higher machine speeds. The trade off is resolution but lets look at the figures so that we can put this in perspective.
Your screw at 1.83mm and on steppers with 1/8 microstepping will deliver a theoretical resolution of 0.0011mm. I doubt any home built machine can deliver that resolution so it is just a theoretical value.
Compare this to a typical ballscrew with a pitch of 5mm and using the same stepper and microsteps and you get a resolution of 0.003 which again is theoretical for a home made machine. What has happened is that you have a much faster machine without effective loss of resolution. This is achieved by bringing back the speed within the motor torque range.
Another issue is the limitation of your computer and parallel port to deliver the number of steps required to move the motors at speed. Taking your example of 882 pulses per mm then to move an axis of your machine at 1m/min or 16.67mm/sec the computer needs to send down 14,700 pulses per second. Multiply that by three axes and you have 44,100 pulses per second. I have just checked Mach3 and the maximum setting for parallel port at 25htz on 882 steps is 1704mm/min so this should work fine on an old clunker computer.

If you set up the motor pins as per your attached sheet that should be fine. If a motor goes in the wrong direction ignore it and get back here to get a fix. As long as the motor travels in both directions all is well.

For initial feeds and speeds I would set the motors up with 500mm/min and acceleration at 150mm/sec/sec. Also set you step pulse to 2us on the same screen. Don't be tempted for more speed and less acceleration. If the steppers lock up at 500 then slowly drop by 50's the speed until you get reliable results. If all sounds well at 500 then increase speed or acceleration but not both by 50's until it stalls - a horrible grating noise and no movement. You are not damaging your motors when it stalls but don't sit there pushing the button as they are not going to move.

Bob this is more for general information as well so excuse the long post. I hope I haven't confused you by waffling on too much.

Hang in there Bob,
I cant help as I'm just a newbie myself.
I'm sure a lot of people have had the same feeling of not understanding the complexity of these machines, myself included. I just sat in the workshop and tried different combinations and settings, and did alot of reading on the forums. When things do fall into place its a EUREKA moment and you realize that it really isn't so complex. I personally feel its a good thing to trouble shoot our machines so we get an understanding of how and why they work.
GoodLuck

Thanks for that triden, I did a lot of reading too, it is just that whatever words I read seemed to sort of float about until they fell into a huge muddy lake. However with people like Rod on the BB I am sure that all my troubles will dissipate. I was just feeling a bit (a lot) discouraged as I had been trying different settings for a while and things just seemed to be getting worse.

Rod.
Thanks you for your clear, explicit and lucid instructions. Those numbers when entered made the machine a LOT better behaved and I smiled. Then I smiled a bit more.

When I used the numbers you suggested I put the instruction G00 Y100 X0 Z0 into the MDI and when I measured the actual movement of the head I found that it was only about 5 mm out. So, to find the exact number that should be entered, should it be: 881.9314298(the EXACT number) /95 (the amount of movement measured in mm) * 100 (the movement in mm desired) to get an exact and accurate movement of the router head.

In the original sheet I attached I showed that I had made the Z axis a 1/16 th step. I have now changed this so that it is the same as the other two. I can no longer even remember the rationale behind my using this smaller step.

I am sure that when I have time to trial different accelerations and velocities I will get a lot more out of the machine.

I still haven't actually cut anything that anyone would recognise as a machined item, but I now have the feeling that it is within my grasp.

Thanks again

Bob

Good to hear you are close.

Yes your calculation will work but there is a better way to get the exact figure now you are close.
Go to the "Settings" screen.

Above the "Reset" button is the axis calibration so hit the "Set Steps per Unit" button.
Select axis - say X axis as a start.
Click "OK"

Mark where your X axis is (lay a rule down on the table with the cutter at zero or use a vernier gauge)

Enter 100 in the box and click "OK"

Your machine will move 100mm

Measure how far it actually moved and enter that into the new box that appeared on screen.

Mach3 will calculate the correct steps per mm and all you have to do is accept the figure.

Do this for the other axes.

The reason I have taken you to this screen is that on the motor tuning screen it will only accept counting number and not decimal places. Once you have run the above test it accepts decimal places - yeah a bit of a quirk of Mach3 but it is an easy sequence to follow and allows for the thread being out a bit rather than accept the theoretical calculation.

Hi Rod
Thanks again for all your help so far.
I have been away for a couple of days so I was unable to get back to you on the outcomes of your suggestions for the final setup.
I did as you said, and Mach3 set up the 3 axis with three different sets of figures even though they should theoretically all be the same.
The X axis is 923.6 + 10 more decimal places, the Y axis is 905.5+ and the Z axis is 917.2+.
I measured the axis' a couple of times on each test and I got a variation of up to 9.1 on the total counts. I'm not sure if this is enough to make a large difference but as I was measuring over a distance of 150mm I would have expected the differences to average out.

I even ran the Roadrunner program (without actually turning on the router) and when I multiplied the scaling by 25 on the X and Y axis, the router seemed to follow a path that agreed with the 193*160 that the toolpath said I should be getting.

The motors seemed to run very hot. Too hot to actually keep holding onto the motors for more a few seconds. Is this normal? All up, the Roadrunner took about 60 minutes to complete the cut. This seemed to me to be very slow. Is this also normal for this program?

You also said to ignore (for the time being) motors that seemed to go in the wrong direction. Well,my X axis doesn't seem to me to be going in the right direction, but I am not sure that this is not all a matter of perspective. When I press on the left arrow, the X axis moves to the right and also the converse when I press right arrow. However, if I was at the other end of the table then the setup would be correct, so is this something that should be changed or not?

Bob

Hi Bob,
To set up you axes in the right direction the conventional way is to do this.
Stand at the front of your machine. The back left corner is Zero, Zero. This will orientate your machine with any drawings you do on your computer.

So X axis moves in a postive direction towards you and a negative direction away from you. Y axis moves in a positive direction to the right and negative to the left. This is the one that usually tricks everybody - Z axis moves in a negative direction going down and positive direction going up.

Now if you stand at the back left corner of your machine you are looking at the same orientation as you have drawn your part on screen. This is important as you need to mount your material the same way you drew it on the screen. Don't worry I have confused myself at times and mounted the material longways instead of across the table.
If any of your axes are travelling the wrong way the easy way to fix it is to go to Mach3 "ports and pins" and "motor outputs" and click on "Dir LowActive" and then the apply button. It is the cheats method but until you get all your problems sorted it will fix it for now. You might have to go to "config", "save settings", exit and restart Mach3 for it to work.

Stepper motors are designed to run hot. I remember seeing that 80 to 100 degrees is normal operating temperature. If they are not smelling or worse sending out smoke then they are probably OK. Not my strong point the electronics so perhaps somebody else can chime in with better information.

As for the variation in steps it is a bit odd but maybe OK given you are not using precision screws. For the moment let's not get too hung up on the finer details so we can get your machine cutting jobs,.You can fine tune down the track.

Did you tune your motors? If you have done that then perhaps this is as fast as your machine can cut. As I said a fine pitch screw will not deliver speed.
If you haven't tuned them then time to do so. If you need help with that I did give a few tips in an earlier post but get back if I have been a bit vague.
I just had a thought (dangerous) but the Gcode would have been running in imperial feed speeds (inches per minute) on a metric machine so that is probably why it was running slow. I just checked and it is running at 60 so no wonder it took so long.

To change that in Mach3 load the gcode file. Hit the "Edit Gcode" button next to the "Cycle Start" button and the file will open in "Notepad". On the second line it has F60.0000. Just change this to F500, save and exit Notepad.
Should run a lot quicker this time around. Again lets not get hung up on this imperial/metric thing as it will not be a problem when you do your own jobs.

Sorry about all the waffling

Hi Rod
You nailed it with the imperial/metric setup. When I changed the F60 to F500 it completed the roadrunner job in just over 12 minutes. A speed increase of over 8 times what I was getting.
I also have found what I think is another error in the roadrunner program. The last line says to set y to 0.2000 and z to 0.0000. I think they should be the other way around.

Now, as to the overall setup of the X and Y axis. If you look at the first photo I uploaded, you can see where I would normally be sitting at the end of the table and in front of the screen. I have been calling the cross axis the X axis and the long axis (going away from where I am sitting) the Y axis. After reading various things I am not sure that this is correct. This is because everyones definition of "in front of" will depend on their own point of view.

So, in order to cut the roadrunner, I had positioned the router so that its Y0 X0 point was at the right hand corner that was closest to my screen. (Close to that green thing in the picture.)

I will now change all the wiring around on the C10 so that it agrees with everybody else's world view
(Edit. Silly me. Of course I don't need to change the wires over, just tell Mach that they are using each others pins.)

As to your supposed waffling - keep it up and do more of it! I usually pick up extra important information in what you are calling waffling. I am sure that other beginners reading this will also get lots of important tips and knowledge

Bob

Hi Bob,
You are getting closer.

Note I said conventional setup for X,Y and Z. You can change it around to almost anything - you are allowed to.

What you do need to do is to get your mind orientated between drawing and machining. You need to be able to sit at the computer and design a job and then go to the machine and see the same orientation you drew on the computer. If you don't you start cutting at right angles to what you intended. I've done it once or twice but fortunately it was only sheet material and not a prize jewelery box. I think there are good reasons why the convention setup of back left corner is Zero, Zero.

I think most position themsleves at the front of the machine at the end of X positive direction. The reason for this is that you can mount and view the job cutting much easier becasue you are not looking through the gantry beam as you would anywhere else around the table. The problem with this is that the drawing on the Mach3 screen is now out of orientation to the table - in reverse so just be aware of this. If you prefer to position yourself in the same orientation then setup on the left side of the machine - Y at zero and anywhere along the X axis.

Well, although I am generally speaking an unconventional person I will stay with the group on this one and keep the X axis and Y axis in the NORMAL place

However, this then leads to the problem of the directional buttons on the MDO not being properly orientated. I notice that there is a sub directory on the Mach software that is /mach3/bitmaps/millBitmaps that also contains the arrow icons. Unless there is an easier way to do it, I could just rename the icons so that they pointed the way that I wanted them to. ie rename JogButtonY-.jpg to JogButtonY+.jpg. They are in fact jpeg files, not bitmaps.

Hi Bob,
Did you mean MPG?
In any case it is probably better to get used to the buttons as they are. The arrows point to plus and minus and that is how I deal with them.
In any case if you move to another position around the table your modifed buttons will be wrong so best to train your mind to work with it.
Perhaps the keyboard shortcuts are easier for you to use and you don't have to press tab to bring up the MPG.
They can be assigned in "Config" then "System Hotkeys"

Hi Rod
Yes, of course I meant MPG.

OK, I won't change the buttons. I was just saying that I could if I really wanted to.

OK, Next problem (two of them actually)

First problem:
I have set up a relay on the number one pin of the C10 to turn the router on and off by software. I have tried many different combinations of port and pins to get this to work. I can get the router to turn on permanently, but it refuses to let me turn it on with an M3 and off with an M5.


Second problem:

I ran the following code with the router off and the z set at Z52 to make sure that the router didn't foul the board at any point and that the length of the program was appropriate. After doing this several times I thought "OK, I'll give it a go. My first actual program run". So, I reset Z to 0.5 and zeroed the other axis as well, I plugged in the router externally and started the program; I wish I hadn't.The first thing it did was drill the hole, and then without withdrawing the router completely from the hole it started traveling towards the next position. I hit Esc and stopped it but not before it managed to scroll a 40 mm long * 10 mm deep channel into the base.
The program is intended to drill a series of holes 20 mm deep at 50 mm spacing for a length of 500 mm. At the end of that, I was going to return to 0,0,0 and add 50 to the Y axis until the whole base was covered in neat holes so that I could use these holes to hold the workpieces in position using pins and wedges. That was the theory at any rate.

Do you have any idea why this program would not work as it should have done and as it did do before I turned on the router?

Line of holes.tap

G21 (Millimetres)

G00 X0 Y0
Z0
G00 X0 Y0

Z-20
G00 X0 Y0
Z0.5

G00 X-50 Y0
Z-20
G00 X-50 Y0
Z0.5

G00 X-100 Y0
Z-20
G00 X-100 Y0
Z0.5

G00 X-150 Y0
Z-20
G00 X-150 Y0
Z0.5

G00 X-200 Y0
Z-20
G00 X-200 Y0
Z0.5

G00 X-250 Y0
Z-20
G00 X-250 Y0
Z0.5

G00 X-300 Y0
Z-20
G00 X-300 Y0
Z0.5

G00 X-350 Y0
Z-20
G00 X-350 Y0
Z0.5

G00 X-400 Y0
Z-20
G00 X-400 Y0
Z0.5

G00 X-450 Y0
Z-20
G00 X-450 Y0
Z0.5

G00 X-500 Y0
Z-20
G00 X-500 Y0
Z0.5

Bob

I can’t explain why the relay doesn’t work. I would use a different output pin than Pin 1 though. Windows when it fires up looks at pin 1 and this can trigger the relay and start your router when Windows starts. Gives you a bit of a scare.

First I would give more clearance in the safe Z height. 0.5mm is a little close for comfort unless you are doing 3D which is a long running sequence and you take risks with a lower safe Z eight..
10 and even 20mm is better if you have clamps to navigate over.
Also you have not set a feed rate in the drilling cycle so it is drilling at your rapid speed. On each of your Z-20 moves make it Z-20 F300 and see how that goes.


The obvious question is did you set Mach3 DRO for Z axis after setting up the router height. How are you setting up the cutter to Zero?

If you have set up the cutter to the table properly and set Mach Z axis DRO to match then you are loosing steps but this is strange as you are loosing steps on the upwards of Z axis. If you lost steps on the drilling itself then the router would be higher and not caused the dig in problem.
You could be getting Electro Magnetic Interference noise from the router causing the missed steps.
Try running the code again with the router up in the air but with the router turned on this time. See if the Z axis is in the same position when you finish the job.

Don't worry about your table top it will get a few more scars before you get through the learning curve. A bit of filler and light sanding will cover it up. While you sort out your current problems chuck a spoilboard on the table and do your runs with a lower Z depth.

Hi Rod
Well, I tried it again and it worked properly. Yes, I am pretty certain that I had set up the Z axis zero properly. I have been MPGing to the thickness of a sheet of paper and then zeroing both the reference all home and the actual X, Y and Zs. The 0.5 mm just cleared all over, so at least the bed must be flat. However, I will in future give it about 10 mm clearance.

I didn't know that I could set the feed rate like that. That is easily done and good to know. I do seem to have a problem with excess movement in the head and this is giving me an oval hole. This is probably caused by plunging too quickly into the board and putting excess pressure on the bit so this is timely advice.

I also realise that it is possible to set the G codes to automatically add -50 to the axis and to perform a series of actions x number of times, but I don't know how to do that yet. I did look at the rowofholespattern.tap that comes with Mach3, but I couldn't get it to work when I tried using those codes with my measurements.

One other thing you may be able to help me with is, how do I use a smaller bit (say a 6mm bit) to make a larger hole (say 10mm dia) and peck it rather than just using a straight drilling action? Again, I read the rowofholespattern.tap for some clues but again failed to understand.

I don't know why, but it is a lot easier to understand when you are telling me how to do it.

Bob

Hi Bob,
You can't use a drill bit to do a miling job. What you are asking is to pocket a hole 10mm diam with a 6mm bit. Either use a router bit of milling upcut bit and use a pocketing sequence. You need a tool to be able to cut on the side so that rules out a drill bit.
Sounds like you need to invest in a CAD/CAM package.