Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: mturneruk on April 17, 2014, 01:43:49 AM
-
Hi Guys
Any help much appreciated.
Background....
I have built an Openbuilds OX, 3 axis CNC Mill.
I have installed homeswitches on the x and y axis and added the home switches in ports and pins.
I have setup my table size limits in "Homing/Limits" and i have autozero switched on.
Motors Nema23 - Unknown brand.
TB6600 drivers
HyCNC 125-4L USB Board (Like a smooth stepper).
I used the "Set steps per unit" auto calc thing to set steps per unit for my motors - to move 100mm.
44.52 was the steps per value that made it move the correct distance. Seems low? - but distance moved fairly accurate...
I made velocity 3000, and acceleration 300 - To get a reasonable travel speed.. Is there something else i have to do to config the motors ?
The machine will home, i.e. it will travel until it hits the switches and then reverse until switch untriggered, however when i do this the machine will not zero.
Everytime i press Refall home, the machine homes but the machine and work co-ordiantes seem to decrease by about 3mm (Into the nagative).
So if I am homed - then i zero the work co-ordinates, then home again, the work co-ordinates and machine co-ordinates are decreasing (into negatives) by about 3mm every time i press ref all home.
Anyone got any ideas ?
Cheers
Martin
-
If you turn off the Auto Zero then Mach3 will set the DRO's to zero, when on it will set the DRO's to the values in the home off entries in the home/soft limits dialog.
See section 5.6 in the Mach3Mill_1.84.pdf manual
http://www.machsupport.com/wp-content/uploads/2013/02/Mach3Mill_1.84.pdf
-
Hi there... Thanks for your reply, however this is not working.
I have Auto Zero enabled and i have home switches.
I press "Ref all home" the machine finds the home switches ok, and reverses a few mm, but then does NOT set the machine co-ordinates to zero or the value of "home off".
It doesent appear to set them to anything.
It seems that the only way i can get machine co-ords to zero is to turn everything off and start again.
Any advice much appreciated.
Cheers
Martin
-
Hi Martin,
Using Version R3.043.057
Auto Zero un-checked (red X), DRO will not zero OR set to the Home Off value.
With Auto Zero checked, the DRO will set to the value set in the Home Off dialog, 0.000 by default.
Russ
:)
-
Might be an issue with the HyCNC 125-4L plugin.
Have you contacted them ?
Russ
-
If you are definitely looking at machine coords (and not work offset) then I suspect it is as Russ mentioned, the external controller/plugin that is at fault.
Hood
-
I have a simular setup, shut down mach3 software, unplug the usb cable, turn off the power to the motors, move all your axis to their zero point
power up the motors, plugin the usb cable and run mach3.
highlight / enable machine cord's button and press the zero buttons, all the values will zero.
not sure if this is the correct way to do it but it works
-
ignore that, got the wrong end of the stick
-
Thanks guys.
But now i've got multiple problems. They are as follows.
1. Motors sound like they are randomly skipping beats as the axis moves up and down.
I assume this means that the motors are missing steps and that probably led to no 2.
2. I tried to cut my first part today - (A dust guard attachment to attach my hoover to), and the result is wildly off what was it was supposed to be. I mean like between 3-5mm errors across the piece.
Circles not circular etc. A complete waste of time...
Just to say - I have lost count of how many times i have set calibration across all axis with "set steps" and an accurate ruler, and it seems reasonably accurate in Jog Mode.
When you run gCode the head seems to wonder. It certainly never returns to 0,0,0 in the work co-ords.
3. Machine Co-ordinates won't zero as mentioned previously.
Could it be the HyCNC 125 USB board that is causing the inaccuracies and skipping, or is it more likely to be the stepper drivers?
Is it a set-up thing. I have spent all week on it and have got nowhere. Trying all kinds of settings.
I have these tb6600 drivers, and i have tried several mirco step options - 1/16 and 1/8th...
I have also adjusted the ampage pot up and down which made no difference.
http://www.mib-instruments.com/servlet/the-3453/TB6600-Stepper-Motor-Driver/Detail
Also these motors....
http://www.cnccheap.com/index.php?route=product/product&path=73_76&product_id=54
Seriously considering sending all of these electronics back. I can't spend much more time messing around with Chinese stuff to end up with no result.
Thanks for any help you can give me to get started....
Best Regards
Martin
-
Piccyman - Yes thats what i have been doing to get a machine zero. Surely there must a problem.
-
Hi Martin,
Sorry to hear that you are having problems getting everything to run correctly.
Just to be sure - you have set motor directions so that the origin (X zero – Y zero) is with the cutter at the bottom left hand corner of the work ??
It’s open to debate but this is what I would suggest…
Initially, I would recommend setting up your motor tuning steps per unit using 1/8 micro-stepping (there is little or no advantage in going to 1/16).
Set your motor current so that the motors will run warm to the touch but not so hot that you cannot hold your hand on them with comfort.
Check your wiring runs…
Inputs (limit / home switches, e-stop etc.) should be connected using shielded wiring, with the shield connected to a single point earth at the controller end. All other earth connections, including the machine frame, should connect to this single point earth. (If your wiring looks like a rats nest then that is asking for problems).
Mains wiring should be routed separately to the signal wiring (Inputs).
Perhaps posting a picture of what you have at present (layout of the wiring) may help with suggestions.
Hope this is of assistance,
Tweakie.
-
Hello Martin,
I am using the 60khz version of the same card and the same Stepper Motor Driver, i had the skipping problem too, i set the motor driver on the nema 23's to 1/16 and 10000mm/m and acceleration to 125
I also switched to vista, that made a difference, for some reason a clean install of vista made a big difference, try doing a clean install of your OS and see if that helps.
My problem that Hood is looking at is, after moving away from the homing switch after homing the A axis won't stop moving away :(
John
-
Hi Guys.
Thanks for your replies.
I have not got shielded wiring. Do you really think that's the problem?
I made the decision about that after reading that it doesen't make much difference in a home environment and speaking to robocutters.co.uk where i got my kit from.
I couldent get the shielded wires in my carriers.....
I haven't got my A Axis driver yet, so i am not expecting the machine to be highly accurate, as the A lags behind slightly the Y.
However not 5mm out which is my experience so far!
Piccyman: I tried setting up a slaved A axis in Mach3. Then my Y homing switch did not stop the carriage at all.
Even though the light lit up on the diagnostics screen in Mach3, the machine did not stop when homing and pressing the switch.
I tested this a dozen times. The x still worked. So i think this is going to be another problem...
I was not planning on having any switches on the A Axis as i thought that would just follow the Y Axis. Is that not the case ?
Does your machine reset the machine co-ordinates when homed?
Running on Windows 7 - 64 Bit. Not keen on a fresh install to be honest, this machine is my main pc and has so much stuff on it. I honestly don't think thats the problem as its a top of the line PC with loads of power and graphics.
Tweakie: I tried the x axis on 1/8 and that made no difference to the skipping.
Yesterday i reinstalled mach3 and noticed that when i first fired up the machine that there was no skipping. Brilliant i thought. Fixed.
After 10 minutes the skipping seemed to return getting worse as time went on.
What would be the dogs in terms of alternatives to these electronics?
Reading the smoothstepper forum, it seems as some of those guys have similar problems as well - So i'm not sure that would necessarily help.
I have invested over £1000 in this machine, and i just can't accept a crap result in the parts that come out.
Thanks again for your help
Martin
-
PS. Who is Hood ?
-
Pics attached...
Motors skipping straight away this morning..
Cheers
Martin
-
More Pics
-
PS. My steps per seem really low compared to others i have read......
What makes my machine measure correctly in the auto set-up "Set Steps per unit" in mach3 is....
Steps Per : 44.31168831
Velocity : 1500
Acceleration : 300
Does that seem wrong?
Cheers
Martin
-
What gears are you using on the y axis?
-
Hi Martin,
Again, just my opinion…
Despite what you may have been told you do need to replace your inputs, switch wiring with shielded cables.
Looking at your pics you also need to change your ‘PSU to drivers’ wiring with a more substantial size of wire and connect ‘point to point’ not ‘daisy chained’.
Your Velocity and Acceleration settings look fine.
Tweakie.
-
I have
Y Axis 36 tooth pulley @ 2mm pitch, the steps/mm are 53.8354 vel: 10000 acc 125
X axis 20 tooth pulley @ 3mm pitch, 53.7268 gel: 20000 acc 125
-
Hi Piccyman,
Those Velocity and Acceleration settings are fine also although I think you may run into problems with the axes mass if fitting a Porter-Cable style of router spindle.
In my view, the object of the exercise is not to have the fastest machine on the planet but rather the most accurate and reliable.
You may want to look again at your ‘steps per unit’ as according to my calculations…
20 tooth pulley, 3mm pitch belt, 200 steps per rev stepper, 1/16 micro-stepping equates to 53.3333… steps per mm.
36 tooth pulley, 2mm pitch belt, 200 steps per rev stepper, 1/16 micro-stepping equates to 44.4444… steps per mm.
This, of course, assumes that your belt is accurate (has not stretched etc.).
Tweakie.
-
I have just checked the 2mm pitch on the y axis pulley is 30t
so both steps / mm should be around 53.333
http://calculator.josefprusa.cz this is handy
-
Thanks Tweakie.
I have a bit of shielded cable. I will try your suggestions on one axis only and see if it that makes a difference.
I have spent an age wiring all this - this week. Not to keen on doing it all again to find no difference.
If i can get a smooth running motor, then I will be well chuffed. You really think it will work?
Best Regards
Martin.
-
its not the shielded cable, mine doesn't reset all axis to zero after auto homing, but i wouldn't expect it too
After you auto home, move x and y axis to say Z 50 and X 50, then press zero x and zero y buttons , you can do the same with z axis
after all 3 are zeroed move the axis's to a new location, then press the goto zero, i am not sure you really need the axis set to zero after rehoming, as far as i know you would normally move all 3 axis to the work zero point and then press zero x, zero y button and z zero button, this is now your working zero, and goto zero will take to back to your work starting point.
i might be wrong
-
Well i have tried disabling all my other axis and switches and just running the x axis.
I also pulled all the boards out of my box away from the power supply and kept all the wiring separate.
I also used thicker wire for the power supply to the driver.
That made no difference.
Piccyman - have you got your motors running 100 smooth now, with no skips, notches or bumps ?
Just for information I don't think its the motors either because the NEMA 17 for my z is skipping as well as my NEMA23's on x & Y.
Any other suggestions ?
New drivers, motors, USB board or all of it?
Cheers
Martin
-
I wound down the current on the 23's it made them smoother, there is a pot on the driver board, with current markers, you can set it with a voltmeter, mine are now .8v
-
Thanks Piccyman - I did try that, but i will try it again tonight. I think I mainly tried it higher....
What accuracy do you get?
Any skips jumps and bumps?
Thanks for your help
Martin
-
no skipping, but it made noises like it was struggling, if you push you x axis by hand with the power off, it makes the same noise, i wound the current down and switched to clean vista, that sorted it.
i have built mine differntly to yours, the motor drivers and breakout card are mounted on the x axis.
I am not sure where all the noise comes from if it does at all, however bunching all your cables up is not a good idea noise wise.
maybe someone can enlighten me
-
Nope that didn't work for me... In fact at a lower current it got worse until the motor just hummed and didn't move.
Just noticed my power supply has a variable voltage. Will try tweaking that next. What voltage are you running at ?
Are you using VSlot extrusion on your machine?
Martin
-
Just to say - These skips just sound like a little blip. Not a graunching jump. However if you watch the cutting head,
you can see a slight jiggle as it happens sometimes. I am sure that's causing the in-accuracies.
Cheers Mate
Martin
-
what happens if you disconnect the belt and just turn the motor, does it still blip? also what voltage does your power supply kick out?
-
Also check the pitch and quantity of teeth on your pulleys, it would seem that the steps / mm are very critical, i changed mine by -.5 and -.6mm on the other, and i got really weird results when drawing out a block of 4x4 25mm circles
-
yeah - still get the blipping with the belt off. Tried that...
However I have just ran a 1500 line gcode (In mid air) over 7min 30secs and it did return fairly accurately to zero afterwards.
Maybe i should ignore the blips.
I am wondering if I had my multipass depth to high. I was trying to cut 6mm mdf in passes of 1.5mm with a 1/8 cutter.
Maybe i was asking to much and it just coulden't cope with it. Somewhere in the program the cutter definitely moved from where it was supposed to be.
I have since cut down my Z axis length and made the gantry more rigid by adjusting the wheels...
Will run some more tests this week.
PS. What do you use yours for?
Cheers
Martin
-
Also check the pitch and quantity of teeth on your pulleys, it would seem that the steps / mm are very critical, i changed mine by -.5 and -.6mm
Where, how did you change that ?
In mach 3 or got different belts pulleys etc
-
motor tuning
-
Hello martin, 2 things i have found this morning that makes a difference to the smooth running
Turn off watchdog on relay 1, its done by going to Config > config plugins > hycnc_4l, this made a huge difference
Set the voltage correctly on the motor driver to match the current required for the nema motors, have you looked at this yet?
-
Hey Piccyman....
So my Watchdog checkbox is off already...
Do you mean set the Amps to match the ampage of the motors on the driver board - Mine are about right I think.
My Motors say 3Amps per phase.. I assume that means 3A on the pots on the boards.
So ran some code tonight to cut out letters of kids names as a test.
Machine running much more smoothly tonight when running Gcode, (but seemed only to skip when jogging.)
However the machine kept stopping half way through the code, as it was in the middle of cutting out a letter. It did it 3 times.
I have just run the same code at least 3 times without the cutter running and it did not stop once. That surely must be down to the router causing interference of some kind.
So i have just ordered shielded cable. I have heard it and been told it, so I thought i would get it and do it, then that is ruled out.
I think i have also decided to forget the Machine co-ordinates being zero. I can't see that it matters to much as i think you said.
I will use the switches on the X & Y as limit switches rather than homing switches, and forget homing. I will just set work zero at the corner of my job.
Hopefully that will get around the A Axis problem of homing not working when you have a slaved axis.
My Z axis shortening seems to of made the machine more solid anyway.
Cheers
Martin
-
I think its down to a process hogging interrupts or processor time on your pc, like anti virus, it doesn't matter how powerful your machine a lot of software is written so poorly and with no regards for other software running
As for the machine homing, just home it after you have run the job and have finished
My A axis homes well now
-
You shouldn't need to be guessing things.
your (Nema23) motors rated currents are 3A so that is what you should set your drives to. From what I can see on the pictures you linked to you should do this by monitoring Vref as you turn the pot - is this what your documentation says?. The formula appears to be I=(vref*1/3)/0.151 from what I can see so for 3A you should be looking for a vref of 1.359V if my maths is correct.
You haven't given info about your Nema17 so I can't comment.
From the info you posted about your N23 motors, they ideally need 43V however your drivers have a max of 36V so that is what your power supply should be - is it?
Your power supply should be able to source at least 2/3 of all your motor requirements added together. i.e. your two N23s will need 6A*2/3 + 2/3 of whatever your N17 needs - can it?
From your pics it looks like you are only driving your gantry from one side - is that correct?. If so then you're going to suffer from racking so you can't expect much in terms of accurate cuts.
As has been said earlier (and I agree) your homing issues are more than likely down to your external motion controler/plugin.
When you're using an external controller, what's running on Windows is not particularly of any concern as it would be if you were using the parallel port.
-
Thanks Stirling.
Power supply is 24V / 20A.
Gantry is waiting A Axis driver. Only have 3 at the moment. 4th in transit.
Will Check the voltage using the forumla you supplied and see if its somewhere near. Where do i put my meter to measure the Vref ?
Thanks for your help
Martin
-
there are 2 points just under the pot on the board
-
Power supply is 24V / 20A.
Gantry is waiting A Axis driver. Only have 3 at the moment. 4th in transit.
Will Check the voltage using the forumla you supplied and see if its somewhere near. Where do i put my meter to measure the Vref ?
Hmmmmm. Your power supply is not ideal - problem is you're already running your motors at little more than half their potential. Not enough V, but more Amps than you could shake a stick at.
It's not MY formula - don't be blaming me when it all lights up ;D - it's right there on YOUR board. If not then the piccy you linked to is different. Check the manual.
there are 2 points just under the pot on the board
Hmmmmmm. are you sure they're the points you need to be measuring? Again from the piccy - there's two terminals on top of the board marked VRef and Gnd. You may be right - they may be the same thing - You guys have the boards and manual. Just checking.
-
Ok - If all else fails RTFM!
Your both absolutely right. The formula you gave is on the manual. (1 piece of paper.)
I have now accurately set the Ampage fairly accurately to 3Amps on all pots. (Which seems to equal 1.36 V) across the two terminals.
e.g. (1.36 * 0.333) / 0.151 = 2.999A
As I think i mentioned my programs have been crashing 1/2 way through. I initially thought this was interference, but tonight i have successfully run
programs in mid air with the cutter running, and then cutting mdf on a very slow feedrate.
I suspect the crashes occur when the motors are under more load. E.g. When the feedrate or cut depth is higher.
I did have a feedrate of 1500, and now i have dropped it to 1000.
Looking at the manual, the drivers say they have TSD (Thermal shutdown), Over Current Detection (ISD) and built in Under Voltage Detection (UVLO).
I suspect that one of these is kicking in at higher feed rates. I expect Over Current Detection - the boards are not getting hot.
Do you think a higher voltage power supply would solve this ?
What would you recommend in terms of volts required - Just reread your post.
Where the hell am i going to get a 36V 12A power supply from?
Thanks for your very useful help...
Martin
-
one in china...
http://www.ebay.co.uk/itm/Universal-12V-24V-36V-AC-8-3A-10A-15A-16-7A-30A-LED-Strip-Switching-Power-Supply-/400606068732?pt=UK_BOI_Electrical_Test_Measurement_Equipment_ET&var=&hash=item5d45fb7ffc
Forget that...This one is ideal, and in the UK.
http://www.ebay.co.uk/itm/600W-High-Power-CNC-Stepper-Servo-Motor-Driver-Switching-Power-Supply-36V-PSU-/360893661695?pt=UK_BOI_Electrical_Test_Measurement_Equipment_ET&hash=item5406f041ff
And an even Cheaper one in the UK...
http://www.ebay.co.uk/itm/400W-PSU-36V-11A-CNC-Power-Supply-Stepper-Motor-/111312987299?pt=UK_BOI_Industrial_Automation_Control_ET&hash=item19eac564a3
-
OK - So I now have screened cables all-round and I am still get skips, blips and judders at random points when jogging.
Sometimes it doesen't do it immediately, but then soon starts. Sometimes its really bad. I can't find a pattern...
When i am running a program, it seems fine.
I have ordered a 36V power supply, so hopefully that will hope with the program stopping 1/2 way through
-
My post has been deleted as it fell on deaf ears :-[
Tweakie.