Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: inspiredcabinetry on December 23, 2015, 12:33:54 PM
-
Hi
Im using mach 3 on industrial cnc machine 4x8 table.
I have this re occuring constant problem where the right side of machine cuts correct and left side of machine cuts 1/32 over.
I have thought since the begining it was the machine and not the mach software but i have exhausted all possibilities with the machine.
I am curious , is it possible there is some setting in mach3 that would cause this to happen?
For more info i have done the following to my industrial cnc machine.
The tech supports there have been helpful even offering to pick up machine.
1) changed x driver
2) changed x cable
3) checked connections to motor.
4) installed " smooth stepper " device (supposed to make machine run better.)
5) changed parallel cables
6) changed circuit board to where printer cable gets attached.
7) calibrated mach 3 many times to see if changes were made. Using "axis calibrator"
8) realigned top racks for pinion to ride in
9) when that didnt fix it i requested entirely new racks and mounted them myself but problem still there.
10) changed computer 3 times even getting manufacturer suggested one.
I have done all these and more trying to resolve this issue. Company was willing to pick up machine but they said what i was thinking " we changed every part on the machine "
Any insight would be greatly appreciated whether its on machine or in mach 3 ....there were some other minor adjustments made in mach such as motor tuning.
This problem has occured since day one of machine (although not noticed ) and as of today its a year with same problem.
Joseph marrone inspiredcabinetry .
Much much appreciated
-
Does the problem reoccur front to back ? DOes it repeat the same effect every time in the exact same spot ?
Are you talking about errors on a feature size lets say a circle and a circle cut on the left side is different than the one on the right side ? IS the circle perfectly round or out of round ?
Did you use the exact same Gcode sequence on each test and just move the work offset over from left to right.
IF the front left different from the back left ?
Is X different than Y ??
Did you check for pinion runout ? Did you check for gear clearance in the rack or IF the rack runs exactly parallel with the gear all teh way across the Y
I can go on but I think you get the idea ??
-
No it does not occur front to back (my y axis) only left to right (my x axis)
I should also mention , because i know they are set different per machine, my x axis (where the problem is ) is driven by 1 motor along 48". The y axis is driven by 2 motors but theres no problem on that axis .
Yes it is always the correct size on the left side of the table (my x 0-24" ) and always incorrect on right side of table (my x 24-48) however i can recalibrate the x axis so it cuts correct on right side but then left side would come out bigger.
Yes ive used the same gcode. Problem has occured for over a year . Im mostly running cabinets parts on my machine so its sometimes more critical then others. Same gcode one gets generated by " cabinet vision" other gets generated by aspire vcarve pro.
Oh ....i see..... Draw one box and cut it then offset 24" ....i will have to give that one a try although i imagine the same result will occur but i will try that one .
No its the same result the full 96" off the machine same result no matter where i cut along the y axis.
Next question a little tough to answer . Not sure what " pinion run out " means but i did disconnect the x motor and push the x axis manually. There were minute discrepentcies in play between the rack and pinion system ....i painfully adjusted that so the tension was equal along the whole run ...... I literally had to use a sheet of paper to get it just right . I honestly feel and felt that it cant possibly be this critical but did it anyway. Adter all that the problem was still identical.im not sure about the " parallel " question either unless i answered that above . But i did also shim the thompson rail out again about the thickness of 2 sheets of paper in 2 spots......i was convinced at the time this would solve it but it didnt. That box still cuts exactly 1/32" different then other side.
It just kind of bugs me that the problem is so consisttant and exact ly wrong. Its not like the right side varies in how much its iff by . Especially when i made some of the changes above ,.....you would think that making some of those adjustments its kind of impossible NOT to get a different result. Which is y i turned to the mach forums. Hoping there is some setting for 1/2 the table (which i know how it sounds just writing it but i had to give it a shot)
Let me know what you think and im already thankful for your response
-joe
-
Oh regarding the circles. Earlier in the year the techs at machine place told me to leave 1/16" play in gear that rides the rack .....i guess thats called the pinion? But anyway with that slop in there i was having problems with my circles a.k.a hinge boreouts. But after i snugged that back up so there was no play the hinges boreouts came out perfect again.
-
If you cut a 2" square on the X zero side then cut a series of 2" squares across the table to the X+ side of the table, how would the squares measure up? Also check were the 2" cut outs come to, are they out relative to the Gcode. Is the problem cummulative or in one position. Cummulative would point to wear in drive or incorrect step set up. One position would point to broken tooth/gear/chain link................?
Good luck
-
OK Mach3 software has NO idea if you are on the left or right side of the table IF you use the same Gcode. So that tells me you have a mechanical problem on the gantry drive mechanism not being the exact same on both sides. Is the gear rack full length across the gantry or 2 piece. If 2 piece where is the split ? . The amount of gear clearance WILL change the pinion pitch circle diameter of teh gear and rack so it could change the exact motion from the left to the right side if there is MORE clearance on one side compared to the other.
Check in the formula section of config and Make SURE there are no values in the formula lines.
-
Just thinking out loud...
If the X axis rack is in two parts is the pitch of each part 'exactly' the same ??
Alternately if the rack is in one part then is the pitch 'exactly' the same along the whole of its length ??
Not sure how you would check it (without removing the rack) but there must be a way.
Tweakie.
-
Man i am soooo happy i posted .
You guys are giving me great info and im glad we are all on the page with the thought processes here.
Ive asked colleagues around me but they are not familiar so this has been a great help!
@tweakie .
Yes x axis is in 2 parts . My racks are not pitched they are straight up and down although ive been told the pitched ones are better because more teeth grab instead of just one . Its something i might have to look into .
I did ask for entirely new racks from manufacturer to replace the existing ones to rule out that. Although not that u r mentioning pitch i will have to go back and triple check they are sitting flat .
Thanks!
@br549
yeah i kind of figured it had nothing to do with mach i was just running out of options. Plus there are soooooo many micro adjustments inside mach that i thought there might be a small possibility .
Yes the split in the racks is directly down the center. Its approx 2 26" racks with a seam in the center. I changed both those racks to brand new from the manufacturer and attached them both atop the gantry. I kept motor disconnected and manually pushed spindle back and farth readjusting until it was smooth and even the whole way.
Is the " pinion " the gear with all the teeth that interlock the rack? Cause that has not been changed and these responses are leading to that being an issue. Although maybe not. Ill adress that in next response .
Also what is or where is " formula section in config" u know what .....illl have to reread when im in the shop.
Thanks!
@ bfgstew
Funny u say that . I did think along the same lines and ran a test cutting 4" squares instead of 2" and i did notice very very slight dimension changes from Square to square (less then 1/64 or around) note : this test was pre smooth stepper. But when measuring 7 squares together they would measure the correct dimension on one side of the machine and that same 1/32 off on the other side. The pinion(gear) that runs along the rack may have a damaged or worn tooth. I have visually checked it but since since the variation is so slight maybe its not visible with the eye and i need a better way to check that.
My high school math escapes me . The diameter of the pinion gear is approx 6" what would be the outside perimeter?. Or is that the cicumfrance ? Is that pie.r2? The reason why i ask is if there is one ormore damaged tooth on that wheel wouldnt the error in dimension still occur on both sides of the table ? Unless its somehow happening 1 x on 1s and 2x's on the other.
After all the changes and part replacements ive done above this is one small piece ive never replaced.
This seems like the next logical part to replace and see if it remedies it.
Thanks to all!
Ill check back in in a 2 or so weeks if this remedied the problem.
-joe
-
IF it were a pinion gear problem then you would see the error everywhere . The rack on the right hand side is pointing as the Problem As teh left hand side cuts correctly. I would seriously look at gear clearance all the way across that rack. IF you can then switch racks from side to side and see IF the problem follows the switch. I think IF you look closely at the tooth clearance on that rack then the problem may show itself.
What you are looking for is the gear pitch diameter that is what sets up the motion. It is the point where the gear tooth is making constant contact with the rack tooth.
-
Just something easy to do using the calibrate feature:.
1. Calibrate movement of one of the racks. Set the steps per unit based on that.
2. Now move the gantry to the other rack. Calibrate that rack for the same travel .
The steps per unit and distance should be the same as in 1.
3. Now move midway in rack / step 1. and use calibrate across the spit between the two racks.
Move in one direct and take out the backlash.
Now reverse the direction of movement opposite of what you did above and do the same.
------------------------
Make note of the each step. You will find any difference between one rack and the other, across the joint line,
and opposite direction. Key here is to be able to measure the travel distance accurately.
FWIW,
RICH
-
Update. More info toward all responses.
I was hoping to post a picture here ...... I imagine it can be done but i dont see the option atm.
Anyway this is the test i ran.
I cut a series of rectangles from top y+ to bottom y-
Rectangles are all 3" in y direction.
1rect)44 x 3
This rect is the Correct size
3)14" x 3"
The Left 2 rects were correct the right rect was bigger
8)4" x 3"
1&2 correct 3rd small 4th correct 5&6 small 7th correct 8 large
2) 20" x 3"
Left side correct right side small
I really think this points to a broken or worn tooth on the gears 3 gears interlock each other mathematically its the only explaination i can figure for why the box sizes vary so when i cut 8 of them.
For arguement sake lets say 1 wheel had a brroken tooth and as it turned the broken tooth hit some boxes 0 times some boxes once and some boxes 2 times. It would make sense .....no?
I just emailed the manufacturer for pricing. If u guys thinkt can be something else before i purchase the parts and spend my time please weigh on .
Thanks so much!
-
I was hoping to post a picture here ...... I imagine it can be done but i dont see the option atm.
Click on the Reply tab (rather than using Quick Reply) then use Additional Options... to attach the pictures (size images to about 800 pix. wide).
Tweakie.
-
Thanks!
I see it.
It wont let me post pics from ipad .
Ill have it transfer, change file type and upload later.
Its really just pictures to show what im looking at but sometimes pictures really do speak a thousand words.
-
Changed pinion.
Problem still persists.
1 side cuts 1/32 bigger then other.
-
So you now have no bad gears after changing the pinion.
Great!
So here is what I would do:
A. I would manually calculate the steps per unit including what a step per unit is in linear movement. Make note of the calculated value.
The calculated value is theoretically correct if all was mechanically aligned properly. Any actual movement should be close to the calculated
value. You will need to assume something as a standard when using the calibration feature so lets assume the left side is adjusted correctly.
If the steps per unit is correct you should be able to machine accurately.
When using the calibration feature of Mach make sure you always remove any backlash and move in the same direct for what follows.
1. Use the calibrate feature for the left side. Note the steps per unit value. And set the steps per unit for the left side.
2. " " right . Note the steps per unit value. DO NOT set the steps per unit!
3. Now calibrate across the spit of the rack, 1/2 of the pinion diameter rotation for each side of the split. Make note of the steps per unit.
B. Compare the steps per unit in terms of distance values of 1 &2.
1. Adjust / tweak the right side rack. Make a calibration check and see if the steps per unit are closer to the left side.
You are making a small mechanical adjustment to the right side to see if you can duplicate the left side rack.
Ideally the steps / per unit should be the same for the left and right side, BUT, a small difference may be of no concern if
the difference is small.
With the right side rack adjusted correctly you can then check what affect the spit does and if it is acceptable.
Note that that a rack may nay not linear and you can check portions of them to see the amount of increase / decrease / or variance.
Know what you have before you adjust and such that you have a basis as you adjust.
Remember that one rotation of the pinion would be a good linear reference to use.
There are easier ways to do this but unfortunately you do not have the equipment to do it.
RICH
-
The rack is in 3 pieces?
Is the pinion spring loaded into the rack,or fixed location.
My guess is one of two things.
One, you have a bad section of rack.
Or, if the pinion position if fixed, perhaps the pinion height is not parallel to the rack, getting looser on the right side?
-
At gerry
2questions below
The rack is in 2 pieces.
Its along the x axis driven by 1 motor.
Pinion is at a fixed location(not spring, belt , chain or pulley )
I thought it was the rack but replaced it with 2 brand new racks with the same result.
Pinion is fixed and parallel throughout..
When motor is removed and i push the spindle assembly back and forth along the x axis it used to get tight at the far right.
I modified the rack slighly so it requires even pressure throughout the x axis.
1)Also note then when i change direction there is Zero play in the pinion,which in my understanding, this rules out backlash as being an issue?
2)is there supposed to be an extremely slight angle to the racks? Like 1 degree or less?
-
At rich.
Im going to follow your response step by step explaining as i go just to make sure i am understanding correctly but before i do that i do want to mention i dont see the fix.
Woukdnt Calibrating 3 different locations on the x axis just change the entire x axis all 3 times or is it possible to actually calculate each section of rack?
Because if i can calibrate each section that will solve all my problems.
So....
Yes i ruled out a bad pinion so that is good.
A) calculate steps i did cut a series of rectangles for this problem.
I dont remember exact numbers but for explaination ourposes this should suffice.
(Replace the following word box with" rectangle")
10 boxes 4" wide
4 boxes 10" wide
3 boxes 14" wide
2 boxes 20" wide
1 box 40" wide
From this above information i get the following numbers as answers to your next suggestions.
using the 3 box test from above
A1) 14"
A2)14 1/32" (i did nothing to calibrateas you suggest)
A3) 14" although im not following the math you are asking for here? "1/2 pinion diameter"? If diameter is 6" you want me to travel 3" on wither side? Or 1/2 the circumference? In other words have the pinion make a half of turn on one side and then a half a turn on the other side of rack? I think thats what u r suggesting but not entirely sure.
B) 1/32or slightly greater
B1) how? How do i adjust or tweak a solid 1 piece rack that has groves into the steel? What small mechanical adjustments can i make? Are there things i can do like shim on an angle or something to that effect? This could also solve my problem but not sure how to go about doing this. Yeah i dont mind if steps per unit are different from one side to the other but can i set it that way? If so where? The only steps per unit i have found adjust for the entire axis unless i am mistaken.
I follow you regarding rotation and remainder of your response.
Last thing.......you mentioned easier ways but i dont have the tools.
Can you mention some of those tools ?
If they r inexpensive enough i have no problem spending sone money to resolve this.
Or if someone has the name of a local tech here in ny thatd be great. I dont mind paying i just want this to cut properly.
As always iappreciate your and everyone taking the time to help me try to resolve this.
It means alot!
-
First let me say that what you find by machining a piece of wood and measuring is just not the way I test.
Using measurements just eliminates any thing to do with possible code or actual cutting.
Simply put, steps / unit from pc = a linear distance of travel movement.
You calibrate the axis for only one section, say the left, and then test / check what the error is across the joint and also the right section.
Now you know comparative error amounts. For the joint, you position to the left of the joint by approx 180 degrees of the pinion travel, and then move another 180 deg of pinion travel, which puts you on the right side rack.
SO,
You can compare the two different racks separately for linearity of movement and after that you can check across the joint.
How would I do it with my equipment?
I place a 48" calibrated scale, which is accurate to within 0.001" at each 0.1" division and can measure to within .0002" in between the divisions.
An optical alignment scope with measuring capability ( optical micrometer ),is used to read the scale.
The axis is moved and readings of movement are taken and the linearity is plotted. I have even checked the reverse movement as sometimes the reverse movement is
different than the forward movement. So one easily and quickly profile the movement.
One can accomplish the same thing with long calipers or a known reference but it is cumbersome to do for long distances.
BTW,
If one has a nice 48" Starrett scale graduated in 0.010" and calibrated you can use that along with a cross haired optical magnifier.
Also make note of what Jerry said as the pinion contact on the rack may be different along the rack and would affect travel.
RICH
-
Yep, googled some of those tools you mentioned......some expensive stuff there.
I think its time i start researching a quality machinist and let a pro handle it.(it would cost far less then just one of those tools)
It seems i am in over my head.
I have dealt with manufacturer tech support and they were unable to diagnose and resolve the isdue, ive also dealt with a more local cnc shop and they too were unable to help but i will look for someone with the right tools, knowledge and know how to work on machine and check back here when it gets resolved.
Thanks again to all for your input and assistance!
-
Get a tape like this and a USB microscope for about $20 to read it.
http://www.amazon.com/Starrett-SM66W-Adhesive-Graduation-Interval/dp/B004HMDFRM/ref=pd_sim_469_2?ie=UTF8&dpID=51MC%2BHuT1TL&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&refRID=0BKZVMCS77YKKKKKBYQ9
Should allow you to fairly accurately measure long distances. Not as precise as a $600 precision rule, but close enough for a router.
-
Were you eve able to solve your problem, because I have the same situation. Thanks John