Machsupport Forum
Mach Discussion => Mach Screens => Topic started by: HeadSmess on March 18, 2013, 08:37:36 AM
-
allo! its about my screen, so im gunna put it here with "screens".
i made a little turbine impeller the other day, ripped out the gcode on deskproto still under trial rights.
so, the image i used is a left handed turbine. it spins clockwise to "suck".
wasnt really an issue, its more just fiddling, but!
it wasnt until id finished that i realised that it was now right handed!
the image is still left handed. but the toolpath display is right handed when i load the code onto mach3.
i obviously have axis rotation enabled in configurations.
now, obviously theres no Y offsets or i would have made a mess, not a mirror image.
so, if the hand of the turbines an issue, i can either rewrite the code by flipping the image, or by flipping the direction control in the settings(saves writing new code,huh?). im wary of doing this as then my table rotates in reverse to the toolpath display!
i currently have it set up so it rotates with the display, thankyou very much! :p but this appears to be incorrect :(
now, flipping images and rotation directions also affects whether im climb milling or not. new gcode for a flipped image is different, obviously(strategies...), though now i know i can mirror things with the click of one setting if i dont mind what way it cuts! i vaguely recall when i first ran the code thinking "so much for choosing climb..." but if the tables rotating the wrong way of course it wouldnt climb anymore!
what i want to know is...
how do i change the direction the display rotates in mach 3? without affecting parts produced or by perpetually living with it being round backwards? is it even possible?
changing stepper direction doesnt affect the toolpath displayed.
not that ive noticed, at least. didnt try the old close and reopen trick...
OOOOOOORRRRR
maybe its just deskproto being in trial mode thats doing it?
anyways. its not critical, i could live with it being backwards if i have no choice, but id rather have a choice, yes? and its going to bug me until i know for sure! :-[
-
No way for me to test here but are you sure your axis are following standard convention? If your code is correct the display will be correct no matter if the machine is moveing the right direction or not. At least in XYZ I know this to be true. Direction is all in perspective. Looking form one end a CW movement would be CCW viewed from the other end which I know you know, just wondering if it is following standard convention. Also, if the center of the diameter is 0 in Z a negetive move would of course be going deeper than the center as would a negative Y or X (depending on orentation) if their 0 is center of the diameter.
Brett
-
my axes are all conventional(i think!), "follow jog" does exactly that. follows the jogs. positive (tool)moves are RIGHT, AWAY, UP and CW, whilst negatives are LEFT, TOWARDS, DOWN, AND CCW. the display and my table/tool correspond with each other. :D and what comes off the machine resembles the display exactly. if i swap the rotary to go opposite to the display, ill get what i wanted, but not whats displayed.
perspective? looking at the table with it on my left so im looking down at the slots/chuck/front face/ rotating part... inclining my display so it is also being viewed from the same direction.
all other codes ive tried in 3 axis setups have worked with no flipping of dimensions.
only other 4 axis ive run is gear/spline wizard, and for straight splines or spur gears, direction is irrelevant. but i can go have a peek and see if it flips too... ::)
file attached is LH but mach3 displays and produces RH... or, both my copies do at least. (do the old curled fingers and poking thumb like for testing threads). both RH and LH copies of g-code rotate the table the same way, positive moves to 360 then a backtrack to 0.
it could be a setting on deskproto, it could just be the trial version, it could just be im going crazy and should give up...iunno!
-
Did I show you this link already?
http://www.machsupport.com/forum/index.php/topic,10308.0.html
The toolpath will display what the code tells it to. However if you have an axis reversed it can flip, mirror, or make opposite hand parts. Draw out a word parallel to the x axis to cut and then reverse your x and cut the part. It will be perfect............ if your viewing it through a mirror. Now reverse Y and cut another. It will now be mirrored and up side down. The toolpath with the exact same code will display no different if you have your axis reversed or not. Mach isnt smart enough to figure out which way is + and which is -, we have to do that. Its just smart enough to go where we tell it to. If we tell it turn right when it should turn left........... you know what that does.
Brett
-
Mach displays in conventional axis configuration . IF your toolpath display is correct then your G code is correct.
IF any of your axis's machines in reverse/mirror then "it" is reversed to conventional setup.
Just a thought. (;-)TP
-
talking to the machinist at my new (old) job, he says tables are usually on the right hand end of the table. mines on the left.
so if i mount it that end, keep axis settings the same...it will spin opposite to the display. switch direction and it will rotate with the display again. this could be it?
no. it cant be, because it will still produce the part to correspond with the display... and i dont want it on the right anyway as then i have to remount it on the left to cut any gears!
im just going to say its a bug in deskproto.
display corresponds to part produced, display is a mirror image of gcode generated, therefore it cant be mach3 thats playing silly buggers as it will only display what gcode tells it to...
guess i have a go of meshcam and see what happens?
i shall be back with a response...
and no, i didnt see that right hand rule link before.
definitely how my xyz is set.
but...it says positive A axis is counter clockwise looking from X+.
this would mean that it should rotate with the top towards the front? ie, by following curled pinky and ring finger from knuckle to nail?
if it does, then my display WILL rotate in reverse to the real rotation. as i have positive rotating from nail to knuckle...
that PDF needs a neat lil circular arrow for A!
-
hard freaking logic. i think i have it.
i think of my table in the 3 axes.
for the tool to move positive, as is convention, the table moves negative, basically. as the tool is fixed on a spindle that doesnt move, yes? if i had a gantry router, the tool would move... whatever. i think of the tool moving... zero being bottom left corner, a box of
xoyo
x10yo
x10y10
x0y10
x0y0
would be the tool moves right?
then away from me, left, and back to me and zero.
incremental?
x10
y10
x-10
y-10
the table will do the opposite.
the table will do the opposite.
i run a tool around the circumfrence of a circle.
to move it positively, the TOOL, the table goes backwards.
the table goes backwards.
the table goes backwards.
so i SHOULD HAVE THE DAMN DISPLAY SET UP IN REVERSE TO THE WAY THE ROTARY TABLE SPINS!!!!!!
i have to look at the table as if im a camera.
a TOOL!
its not rotating at all.
I AM.
sorry i even started this thread guys.... jeeeeeeesus wept... but i get it now. oh boy do i get it :)
how many dimensions did einstein come up with?
-
Conventional direction is best described via the Right Hand Rule.
Doesn't matter from where you look or which end is the chuck.
-
BUT you do have to remember what type of machine it is, the rules is for tool tip movement not always table or rotary movement.
For example knee mills are backwards to gantry mills and on rotarys you have to consider which way the tool tip moves NOT the chuck.
Just a thought, (;-) TP
-
................ on rotarys you have to consider which way the tool tip moves NOT the chuck.
Not going to argue about.......but I disagree.
Rotary axis direction IS the way the chuck moves.
-
To each his own(;-), Why would the rotary axis be any different than any other axis (;-) It is based on perceived tool tip travel not just the direction of table/chuck motion.
Your mileage may vary, (;-) TP
-
Why would the rotary axis be any different than any other axis
Because it is a ROTARY axis, not linear.
If you want to see conventional rotary axis direction demonstrated, take a look at Mach3
If you want any axis to move in a non conventional direction on your machine, then go for it. No one is going to take away your birthday. ;)
I'm going to shut up now. ;D
-
This is interesting. Good rule illustration Greg. :)
"A" set up with the thumb facing X+, the Y and the A are identical as far as + and - directions go. (relative to the tool)
"A" set up with the thumb facing Y+, the X and the A are opposite, ie a pos. A move will be the tool going to the left instead of right.
Considering a standard knee mill. Would a gantry router be similar ?
Does mach know the difference ?
-
It is all about the perceived movement of the tooltip NOT the table or gantry or rotary axis. (;-)
Rule of thumb from long ago.
IF the "table" moves then table will move opposite of tool tip
IF the "spindle" moves then the table movement will be the same as tool tip
Rotaries are counted as part of the table.
Just a thought, (;-) TP
-
It is all about the perceived movement of the tooltip NOT the table or gantry or rotary axis. (;-)
Rule of thumb from long ago.
Rotaries are counted as part of the table.
Then perhaps you had better tell Art and Brian. All these years they have had it wrong in Mach and did not realise it. ::) ;D
-
You do realize we are describing the same thing right? Only you are describing tool tip motion and I am describing Table motion to get proper tool tip motion as defined in MACh3.
Just a thought, (;-) TP
-
You do realize we are describing the same thing right? Only you are describing tool tip motion and I am describing Table motion to get proper tool tip motion as defined in MACh3.
No I am not talking about the table at all. Has nothing to do with the table. Different beast all together.
Talking about conventional direction of rotation of a rotary axis. As described by the right hand rule as shown above.
Rotary positions are described by angle, not cartesian coordinates.
You can have YOUR rotary axis set up however YOU want. Just pointing out what is conventional and followed by Mach and the CAM software that I have used.
So I will wish you well and leave you with it.
-
It is all about the perceived movement of the tooltip NOT the table or gantry or rotary axis. (;-)
Rule of thumb from long ago.
Rotaries are counted as part of the table.
Then perhaps you had better tell Art and Brian. All these years they have had it wrong in Mach and did not realise it. ::) ;D
im inclined to like this one.... cus to me, that gear/spline wizard looks wrong if displayed rotations are opposite to actual table rotations.
probably the thing that started this whole mess!
oh, fanuc, fanuc....
-
I wanted to leave this alone but am concerned that misinformation can be taken by forum readers to be correct. As demonstrated by the post above.
So I devised as simple a demonstration as I can of CONVENTIONAL rotary axis direction (right hand rule) as displayed in Mach3
Rotary axis positions are measured and described by ANGLE. Not to be confused with X Y positions described by cartesian coordinates.
The right Hand Rule describes the direction of the chuck or job movement. NOT the apparent tool movement.
Two pictures of simple profile toolpaths of lettering on a cylinder, displayed my Mach. Showing that Mach follows convention with rotary axis direction.
I have tried to make them from the same perspective. Included is the 3D compass for reference.
The first is toolpath set to cut correctly with CONVENTIONAL direction of the CHUCK. Displayed in Mach3. You should be able to see the visual display is correct.
The second is toolpath set up to cut correctly with rotary axis direction set up for APPARENT tool tip movement. Opposite to chuck movement. As per Terry's argument.
As I hope you can see Mach displays rotary toolpaths as per CONVENTIONAL rotary axis direction. And has done for as long as I have been using it.
I can not remember any user over those years who has cried "Mach has it wrong"
-
I assume that you are refering to Baxis in you example as it is inline with Y SO your chuck is moving CCw?
OK now show the point of origin of your code is it lower left? IF so then B positive would be to the right which would be CW tool motion or CCW chuck motion. Facing the chuck. YES?NO?
NOW there IS no real standard for the placement of the chuck on rotarys I know just as many that run it on the left as on the right as A axis. the exception being the B where MOST run it outboard to clear the column (mills).
NOW as to physical Chuck rotatation IF I swap from left to right I must rev the direction of the chuck to maintain toolpath orientation.
Just a thought, (;-) TP
-
I assume that you are refering to Baxis in you example as it is inline with Y SO your chuck is moving CCw?
Terry that is a B axis toolpath. To keep the explanation simple, I should have made an A axis example.
My mistake. Here is my fix.
Same basic example. Showing the same conventional direction, but this is an A axis toolpath.
Two pictures of simple profile toolpaths of lettering on a cylinder, displayed my Mach. Showing that Mach follows convention with rotary axis direction.
I have tried to make them from the same perspective. Included is the 3D compass for reference.
The first is toolpath set to cut correctly with CONVENTIONAL direction of the CHUCK. Displayed in Mach3. You should be able to see the visual display is correct.
The second is toolpath set up to cut correctly with rotary axis direction set up for APPARENT tool tip movement. Opposite to chuck movement. As per Terry's argument.
-
............SO your chuck is moving CCw?
IF so then B positive would be to the right which would be CW tool motion or CCW chuck motion. Facing the chuck. YES?NO?
NOW there IS no real standard for the placement of the chuck on rotarys I know just as many that run it on the left as on the right as A axis. the exception being the B where MOST run it outboard to clear the column (mills).
NOW as to physical Chuck rotatation IF I swap from left to right I must rev the direction of the chuck to maintain toolpath orientation.
This is a perfect example of why I keep banging on about the Right Hand Rule.
CW and CCW, "which end is the chuck??" "Facing the chuck. YES?NO???" "Is the sun over your left shoulder??"
The explanation possibilities are endless, as are the opportunities for misunderstanding those type of explanations.
The Right Hand rule simply explains any and all situations. Stand in front of your machine and do it.
Make a fist with your right hand, thumb extended and pointing in the positive direction of the axis you are interested in. Your fingers curl around in the direction of positive rotation.
-
I think you guys are basically saying the same thing ... sort of :)... partially :-\... but not really. :P
This was a bit confusing (to this beginner) but forcing myself to allow the Right Hand Rule to trump everything else makes it much more simple to understand.
My biggest hangup was thinking that the axis directions (in relation to the tool-tip vs. work surface) were the same with rotary as they are with their associated linear axis. This is only true with the A axis and it's linear cousin Y. The B axis is opposite of it's linear cousin X.
I say the RH Rule trumps all.
It works in all phases of the moon too. ;)
Thanks for a great discussion,
Russ :)
-
My biggest hangup was thinking that the axis directions (in relation to the tool-tip vs. work surface) were the same with rotary as they are with their associated linear axis.
Yeah you just need to forget that idea for rotary axis.
Just to throw a curly one. ...... What happens if you are cutting below centre of rotation???
Just stay with the right hand rule. :D
-
...... What happens if you are cutting below centre of rotation???
That is only possible when Jupiter is aligned with Mars. :)
Really though, does Wrapper allow a negative radius ? or can Wrapper even do it ?
Have you done it ? Can you explain how ... briefly ?
Thanks Greg,
Russ
-
OK folks, I'm going to bed now ... with a headache. :-\
I loaded this Wrapper thingy, really neat.
Simulating here on my laptop.
Used the Write Wizard , produced the attached with 0,0,0,0 where shown.
Noticed in the Wrapper extrema that ALL A moves are Negative.
Then I jogged the A axis and sure enough, the actual rotation of the part is opposite of the RH Rule.
But, it loads, looks and runs just fine and as expected/designed.
I guess all of the switching is done in the sw.
Just a bit confusing again.
Can you verify with yours that the jogging is opposite of the RH Rule ? Or, do I have something wrong here ?
Thanks,
Russ
-
Then I jogged the A axis and sure enough, the actual rotation of the part is opposite of the RH Rule.
Can you verify with yours that the jogging is opposite of the RH Rule ? Or, do I have something wrong here ?
I am not sure what you are getting at here Russ.
If you jog in a positive direction, ie, the DRO value is INCREASING, then the chuck, or job, will be turning according to the RH rule.
Your fingers curl around in the direction of positive rotation.
I think where users come unstuck, is expecting it to look, feel and act like a linear axis. It is not, it is a rotary.
Looking at your pic of your example job in Mach, you can see that it is correct. I assume you used CNCwrapper.
The fact that it comes out correct in Mach is because both Mach and CNCwrapper both follow convention for rotary direction. So they match.
This demonstrates the main reason why I think it is worth going with convention. Compatibility with various CAM software, controllers, etc. It is just industry standard.
However as I said, if users want to go their own way, then go for it. You are the user, its up to you.
-
Hi Greg, and thanks so much for responding.
Yep, using the Wrapper.
All set up in Mach (simulating) as conventional, nothing I can change there .... afaik.
Where you say " ..Mach and Wrapper both follow convention" is where i agree and that makes it work out properly.
But, this is where I'm at.
A is mounted in line with the x axis on a knee mill. So Y is converted to A.
The Wrapper converts all of the Y+ moves to A- (but in degrees of course).
So, although Mach and Wrapper follow convention in their respective internal conversions, the actual rotation of the A chuck is opposite of the RH rule.
If I set a hotkey for A+, and a Hotkey for A -, when hit they are the opposite of the rh rule (looking at the chuck) but perfect if looking at the relative movement of the tool tip.
It's neat the way it works, but the rh rule appears to indicate the direction of the tool in relation to the work surface, not the actual rotation of the cylindrical axis.
Either way, I see how it works and now know how to implement it.
I could see your apprehension in getting involved with this discussion, but truly appreciate your involvement.
You are a rotary wizard.
Thanks Greg.
Respectfully,
Russ
-
Russ
Still not sure what you mean here
If I set a hotkey for A+, and a Hotkey for A -, when hit they are the opposite of the rh rule (looking at the chuck) but perfect if looking at the relative movement of the tool tip.
If you press the A+ hotkey and the A axis DRO does not go positive (increase in value) then you have the A+ hotkey misassigned
A+ means jog in a positive direction. DRO increases.
I don't use the Mach standard 1024 screenset. So I had to load it up to check it out.
When on the tab fly out, which has on screen jog buttons, I press "4+" button and the A axis DRO increases as expected.
Does yours not do this?
-
When on the tab fly out, which has on screen jog buttons, I press "4+" button and the A axis DRO increases as expected.
Does yours not do this?
Yes, it does. But at the same time, the physical rotation of the A axis is running opposite of the RH rule.
Simply ... if you grab your chuck with your right hand from underneath with your thumb to the right(x+), then jog A +, the chuck rotates opposite of the rh rule.
Indicating that the tool is moving toward you, opposite as it would with a Y+ move, which would in turn be A +.
it all works, but the finger tips are pointing to the direction of the tool movement, as opposed to the table/surface movement.
It all works as yo describe, it's just hard to explicitly explain.
Thanks Greg,
Russ
-
Yes, it does. But at the same time, the physical rotation of the A axis is running opposite of the RH rule.
Well if you want the rotary to follow convention, always my preference, then you need to reverse the axis direction so that it DOES rotate according to the RH rule.
This is done by toggling the "reverse" check box in the homing menu. But you know that already.
Greolt
-
When on the tab fly out, which has on screen jog buttons, I press "4+" button and the A axis DRO increases as expected.
Does yours not do this?
Yes, it does. But at the same time, the physical rotation of the A axis is running opposite of the RH rule.
Greg, I'm running Mach3 unattached to a machine so reversing the outputs seems to have no effect.
Only looking at the Toolpath display. Stock screen. I'm picturing this on my knee mill in my mind.
I assume I am just seeing the convention built into Mach.
Are you saying that if you hit 4+ on your standard flyout (no screenset mods) the upper surface of the job moves towards you ?
Mine doesn't. How and where would I change this ?
4+ the A DRO counts UP+
4+ the UPPER SURFACE of the job is moving AWAY from the operator. Opposite of the RH rule.
UNLESS ... A+ is actually referring to the direction of the tool in relation to the surface as in a linear axis like Terry says, THEN it all adds up and is conventional as it is.
All of the jobs I've produced for the Wrapper have the origin at the lower left corner and all Wrapper moves are A-. They all look good running in simulation, just as planned .......... I think ::)
Thanks,
Russ
-
From the above post........to clarify.
Are you saying that if you hit 4+ on your standard flyout (no screenset mods) the upper surface of the job moves towards you ?.....while looking at the Mach3 Toolpath display on the screen ?
-
OK, back to the beginning, considering Brett's RULE post and other research and comments, I've come to the conclusion that the WORK never moves ... the TOOL always moves, regardless of the axis "type". ALL direction references are referring to the TOOL moving.
This is how Mach3 displays it, perfectly. :)
Like you said Greg, it can be configured to work either way, but I am inclined to kindly agree to disagree with your perception of conventional.
That's my story ..... and I'm stickin' to it. ;)
Thank you Greg, and others, for a great discussion.
Respectfully,
Russ
:)
-
You're too gracious Russ.
I should have heeded my own advise and left it alone some half dozen posts ago. :D
Greolt