Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Davek0974 on April 14, 2019, 12:02:05 PM
-
Hi all,
i want to add basic turning to my mill conversion by building a servo driven spindle to mount on the table.
Is Mach3 Turn just a screen-set or totally different build of Mach?
Does Mill understand constant-surface-speed commands?
Thanks
dave
-
Hi all,
i want to add basic turning to my mill conversion by building a servo driven spindle to mount on the table.
Is Mach3 Turn just a screen-set or totally different build of Mach?
Does Mill understand constant-surface-speed commands?
Thanks
dave
MACH3 turn is a different profile turning MACH3 into a different "sub-product" with different screensets (*.lset - different to *.set for mill), different plug-ins, wizards and different set of macros.
YOU cannot use a MILL profile for turning.
But you can have turn (lathe) sreensets with more than 2 axes.
-
And also to add, CSS does not work in Mach3.
-
Hi,
at the current time Mach4 does not do Mill-Turn per se. It, like Mach3, has a multiaxis Lathe profile to allow milling
ops on a lathe and also A, B, C axes and Out-of-Band axes (up to 6) for turning ops on a mill. It accommodates CSS.
This is from Smurph:
You are welcome for the new stuff. It is my pleasure. Stuff I have in mind for the future is 3rd order planner, tie the multiple instances together for true multi-path capabilities (Screw machines, Mill/Turns, etc..), and kinematics.
So the intention is that Mach4 have multiple instances for genuine Mill-Turns. When is the next question.
Craig
-
And also to add, CSS does not work in Mach3.
https://www.machsupport.com/forum/index.php?topic=25647.0
Myth BUSTED?
-
And also to add, CSS does not work in Mach3.
https://www.machsupport.com/forum/index.php?topic=25647.0
Myth BUSTED?
No, that proved it was not a myth and was indeed the case that CSS does not work in Mach 3. Spindle speed will vary at the correct rate/speed but the feed per rev does not stay constant. Mach basically changes to feed per minute when you start a CSS move and that only changes when you again command a different feedrate.
The macro attached at the end may work, can't recall if I tested it or not but Mach3 itself can not do true CSS.
-
Hi,
at the current time Mach4 does not do Mill-Turn per se. It, like Mach3, has a multiaxis Lathe profile to allow milling
ops on a lathe and also A, B, C axes and Out-of-Band axes (up to 6) for turning ops on a mill. It accommodates CSS.
Thanks all,
i had an idea this would not be a walk in the park:)
I want to add a simple spindle to my mill so i have basic CNC turning ability, 4th axis/mill-turn would be nice but thats even more advanced i think.
The spindle will be servo driven, my mill spindle is vfd driven. Looks like i will be mounting the tools to the quill so mill Z-axis will be lathe X-axis and mill X-axis will be lathe Z-axis.
I use a CS-Labs IP/A controller. Would i need to use Mach turn or can this all be done from Mach mill?
I use Fusion360 for my CAD/CAM, is there a recommended post for Mach3 turning ?
If anyone cares to throw tips into the pot it would be appreciated;)
-
Hi,
if the turning spindle is free running in the manner of a lathe spindle then you would be advised to use a lathe
profile and consider the mill spindle as an add on to the lathe.
If however you want the turning spindle to be indexed then use a mill profile and consider the turning axis to be
a C axis, ie coordinated with you regular linear axes.
Craig
-
Hi,
if the turning spindle is free running in the manner of a lathe spindle then you would be advised to use a lathe
profile and consider the mill spindle as an add on to the lathe.
If however you want the turning spindle to be indexed then use a mill profile and consider the turning axis to be
a C axis, ie coordinated with you regular linear axes.
Craig
Thanks
I thought Mach3 couldn't spin an axis indefinitely?
I would prefer to use the mill profile as i have a custom screen and used to using it, i could easily add a new screen for turning jobs.
My C axis is used on my knee for mill tool-length compensation via custom M6 macros, can I use A or B instead?
Is there a post for Fusion that would work in this setup or do i need to look to get one altered to suit?
-
Hi,
I think Mach3 is like Mach4 in that it has one spindle only. Thus if you have a free running mill spindle the you can't
have a lathe spindle as well without getting creative.
I've never used it but I believe Mach3 has a 'swap axis' command such that you could have two spindles say, provided
you use only one at a time.
I have upgraded to Mach4 and am more familiar with those arrangements. Mach4 introduces Out-of-Band (OB) axes, up to six
of them in addition to X, Y, Z, A, B and C axes. One, and only one, OB axis may be a spindle with all the usual spindle commands
and interpretations, things like PMW, spindle PID or alternately free running step/direction. The other five remaining OB axes
can be jogged only. Thus if you wanted to turn your turning axis 60 degrees to mill a flat when making a nut say you would
jog that axis 60 degrees using an API command. It is also possible to jog continuously. Lets say you jogged the OB axis
for a move of 108,000,000 degrees or 300,000 revolutions. Its a bit of a cheat, its not a genuinely free rotating spindle
but surely you can complete your lathe turning op in 300,000 revs!
Note that the OB axes are NOT coordinated with X,Y,Z,A,B,C axes. Thus if you started a 108,000,000 degree jog as I explained
above you can now still execute coordinated moves while the jog is running. This gives you the means of 'creating' MillTurn
behavior.
Note also that Mach4 has the possibility, not yet implemented but hinted at by Smurph, that would allow two machines
to operate at once on the same PC, using multiple instances of Mach4. There is supposed to be a common interface
between the two instances that allow genuine MillTurn behavior. Looking forward to that, and incidentally a third order
motion planner.
Craig
-
And also to add, CSS does not work in Mach3.
https://www.machsupport.com/forum/index.php?topic=25647.0
Myth BUSTED?
No, that proved it was not a myth and was indeed the case that CSS does not work in Mach 3. Spindle speed will vary at the correct rate/speed but the feed per rev does not stay constant. Mach basically changes to feed per minute when you start a CSS move and that only changes when you again command a different feedrate.
The macro attached at the end may work, can't recall if I tested it or not but Mach3 itself can not do true CSS.
So MACH4 turn (Hobby or just Industrial version ONLY?) has the G96 and G97 for Constant Surface Speed turning.
One more reason to go MACH4.
-
Hi,
Mach4Hobby has CSS.
I suspect though that it suffers from the same fault that Mach3 does.
Mach3, according to the thread Hood linked to, makes a pretty fair approximation to constant surface speed
but cannot do Feed per Rev accurately. As was pointed out for Feed per Rev to be accurate requires that the spindle
be synchronized with the X axis and that is incompatible with buffered control system.
Mach4, despite it differences, is still a buffered control system and therefore probably not capable of accurate Feed per Rev
for exactly the same reason as Mach3. I have not tested it an Mach4 nor does my setup allow for it to be tested. I would
caution that if you require accurate Feed per Rev then Mach4 warrants investigation.....
Accurate Feed per Rev can be obtained by a C axis as opposed to a free running spindle. This is certainly capable of being
done in Mach4 and I can't imagine any reason that it could not be done in Mach3 as well.
Craig
-
Mach3 TURN does not even have G96 (CSS) and G97 (to turn off CSS and go back to speed per rev).
-
Ok, so i'll forget CSS completely.
Does Mach3 have an OB axis or is that solely Mach4?
I can't use SwapAxis because my main spindle is VFD and my turn spindle will be servo - the Mach manual states both axes must be similar in setup IIRC Plus I don't believe SwapAxis works with CS-Labs controllers as they do not use ports and pins very much in their setup.
-
Hi,
Does Mach3 have an OB axis or is that solely Mach4?
To my knowledge it is Mach4 only.
My understanding is that CSLabs Mach4 plugin is a bit buggy and contrary to CSLabs previous reputation they are
very slack about improving the plugin.
Craig
-
Ok, so sticking with Mach3 I am limited to a new setup using Mach Turn, no CSS and no milling?
It will still do what i want but i always like the bells and whistles :)
-
DazTheGas has a SOFTWARE (LUA SCRIPT) solution to "CSS turning" on a MACH4 MILL profile
https://www.machsupport.com/forum/index.php?topic=39086.msg263072#msg263072
Worth a look.
-
EASE-eTURN screenset (EURO25) claims to support CSS using MACH3.
https://www.practicalmachinist.com/vb/cnc-machining/ease-eturn-mach3-turn-screenset-328827/
Also worth a look.
-
If you want "basic turning" capability why not simply have a separate lathe profile for your milling machine, mounting the tool on the table and putting the work in the chuck? "Turn" is based on "Mill" with the mill laid on its left side then on its back. You could probably make your own screen. This could be interesting: http://www.vinland.com/CNC-Mill-Lathe.html
-
If you want "basic turning" capability why not simply have a separate lathe profile for your milling machine, mounting the tool on the table and putting the work in the chuck? "Turn" is based on "Mill" with the mill laid on its left side then on its back. You could probably make your own screen. This could be interesting: http://www.vinland.com/CNC-Mill-Lathe.html
I can't do mill spindle 'turning' as i need to pass 25mm bar through the spindle, mill spindles can't do that :)
I'll try a new mill profile, designate the mill X-axis as lathe Z-axis and the mill Z-axis as the lathe X-axis, spindle will be declared as a servo spindle drive is CS-Labs setup.
See how that goes......
-
Does Mach3Mill understand the T0101 lathe tool change command style?
Seems not as i get a "Slot number too large" error, but thought i'd check ;)
Can it be made to understand it?
-
Seems this could be useful, DoOEMButton(131) - toggle mill/turn mode
Not found much about it - anyone know exactly what it does?
It does load the turn tool table and also allows the T0505 tool change code to pass ok
Also not found a LED or way to indicate what mode Mach is in but that doesn't mean there's not one ;)
-
Does Mach3Mill understand the T0101 lathe tool change command style?
Seems not as i get a "Slot number too large" error, but thought i'd check ;)
Can it be made to understand it?
MACH3 TURN (Lath) uses the format TAABB to define the tool number and the offset where AA is the Tool number and BB is the offset number. That means that the max number of tools is 99 and similar for the offsets.
The so called ToolTable is actually the Tool Offset Table and the max number of offsets you can have are as above 99.
Mach3Mill only uses the format T### where ### is from 0 to 253. 254 tools max.
-
Mach3 TURN does not even have G96 (CSS) and G97 (to turn off CSS and go back to speed per rev).
It does, G96 is CSS, G97 turns it off G50 is max RPM. As mentioned however it does not work correctly, the spindle speed will vary but the axis feed gets locked at the current feed/rev and converted to feed/min until a new feedrate is commanded.
Hi,
Mach4Hobby has CSS.
I suspect though that it suffers from the same fault that Mach3 does.
Mach3, according to the thread Hood linked to, makes a pretty fair approximation to constant surface speed
but cannot do Feed per Rev accurately. As was pointed out for Feed per Rev to be accurate requires that the spindle
be synchronized with the X axis and that is incompatible with buffered control system.
Mach4, despite it differences, is still a buffered control system and therefore probably not capable of accurate Feed per Rev
for exactly the same reason as Mach3. I have not tested it an Mach4 nor does my setup allow for it to be tested. I would
caution that if you require accurate Feed per Rev then Mach4 warrants investigation.....
Accurate Feed per Rev can be obtained by a C axis as opposed to a free running spindle. This is certainly capable of being
done in Mach4 and I can't imagine any reason that it could not be done in Mach3 as well.
Craig
CSS worked when I was testing. If I recall there were inconsitencies in the feed when using metric but that got sorted. I have not messed with it for probably 4 or 5 years though so things may have changed.
If you want "basic turning" capability why not simply have a separate lathe profile for your milling machine, mounting the tool on the table and putting the work in the chuck? "Turn" is based on "Mill" with the mill laid on its left side then on its back. You could probably make your own screen. This could be interesting: http://www.vinland.com/CNC-Mill-Lathe.html
I can't do mill spindle 'turning' as i need to pass 25mm bar through the spindle, mill spindles can't do that :)
I'll try a new mill profile, designate the mill X-axis as lathe Z-axis and the mill Z-axis as the lathe X-axis, spindle will be declared as a servo spindle drive is CS-Labs setup.
See how that goes......
As was said, why not have a separate lathe profile and use that when wanting to Turn or are you requiring something in Mill that Turn doesn't have?
Simpson36 makes a 4th axis for a mill but if I recall he uses SwapAxis to shift spindle and axis and may also have some other electronics involved, it has been a while since we spoke.
-
It was only 3 years ago :D
https://www.machsupport.com/forum/index.php?topic=33175.msg230618
-
Hiya Hood, how's the plasma cutter running?
Here's my thing...
I have my mill converted, i now want to do basic lathe work but my manual lathe is not worth converting and i will never have enough work for a full CNC lathe although I did look long and hard about getting a small one like a Denford Mirac etc.
I need to pass a 1" alu bar through headstock so mill spindle turning will not work.
I looked at the InTurn by Simpson36 and its pretty comprehensive, in fact way too advanced for my mediocre needs, he now uses a hell of a lot of custom electronics to make it sing via Mach3/4.
I cannot use SwapAxis() as i use a CS-Labs IP/A controller and that does not use the ports&pins in Mach so SwapAxis does nothing.
I will need to use another Mach Profile to swap my Z axis for X and my X axis for Z, the tools will be mounted on a quill clamp so no mill-turn work (yet) - its a Bridgeport manual mill (or was) so i'll be using Z as X. The quill clamp uses the BT30 socket for stability.
I have tested G32 threading and that functioned ok in Mach3Mill - i currently have an encoder on the main spindle and will have one on the aux spindle, switching from one to the other can be done electronically then the lathe profile is selected. I can only have one spindle encoder with CS-Labs.
I really want to stick with Mill as i'm used to it, I have my own screen set that I can easily tweak, it all seems to work fine for mill at least.
Not sure yet if it will be gang-tooled or a simple QC tool post setup. I'll probably only need a few tools anyway.
All i need is something like Simpson36' early model, I have now got a 5C spin-index spindle, bearings and a 750W servo and drive plus, aluminium tooling plate are on order.
I really want to stick with Fusion360 for my CAD/CAM so no doubt it will need the post modifying to output mill tool-changes instead of lathe codes.
Sounds simple, probably not, should probably have just bought a mini-lathe but there you go ;)
-
Still not sure why you can't just use a Turn profile unless you are wanting to do milling operations and turning operations at the same time, it sounds like you don't though.
Ok your screenset, that is not an issuer, just turn it into a Turn screenset. Copy it to your desktop or similar and change the extension to .lset then paste into your Mach3 folder. You can now open a lathe profile and choose your newly re-extensioned screenset and you now have a lathe profile with your screenset.
It is infact the screeset that determines what a profile is. If you opened your mill profile and chose a lathe screenset then shut down and reopened it would now be a lathe profile.
-
Sounds good, i'll play with that thanks.
It did start out as a mill-turn project but on a Bridgeport thats just not going to happen i think as the only Z is the quill so the tools must be mounted on that and not on the head like other mill-turn projects are. I tend to get stuck in a groove and still think mill-turn although now its mill or turn and not both.
That would mean no post processor tweaks needed as well.
-
Ok, impressed so far, i now have my screen as a lathe, thanks.
Clearly i need to look at IJK stuff as (pictured) it's wanting to turn a circle rather than a radius ;)
-
Probably your tools are set as Front in the tool table, you can either set as rear there or change the "reversed arcs in front post." option see screenshot.
-
Spot on, thanks. :)
Makes much more sense now, code runs fine on my laptop at least, display looks good, need to alter my tool change macro as that was heavily tweaked to pass length offset to the knee.
-
Personally I would set the tools to rear post in the tool table and have the setting for showing the tool above. That way it would be more representative of the actual setup you have.
-
Yep, thanks
looks very good now.
-
Any encoder guru's here ? :)
Trying to find out if the "Encoder" output from my AC servo drives would be able to drive two inputs? This output is from a line-driver chip.
I'm using a CS-Labs IP/A controller and trying to avoid fitting a separate encoder to my 4th axis spindle - hoping to use the motor encoder and index pulse in parallel with both the main controller and the 'enc' module which is used for threading/tapping synchronised operations.
The spindle is driven at 1:1 ratio via a non-slip toothed belt.
The pid loop is closed within the motor/drive so i think the CS-Labs input is only used for homing and display actions.
Any thoughts?
-
I do that on the Chiron. Buffered encoder outputs from the servo drive goes to CSMIO/IP-A and Enc module.
Oh and as far as I am aware the loop is closed in the CSMIO, the drive itself will just be in velocity or torque mode.
-
Thanks Hood,
Will give it a go soon.
Yes, i goofed on the pid loop, you are correct of course, IIRC the drives are in velocity mode.
-
Ok, progress is made- I now have a working turn setup using my screenset, spent a few hours and reconfigured my mill X & Z axes so they now work as a lathe should, homing, soft-limits all ok.
I can now run turn code straight from Fusion and it all seems to run ok.
Only air-cutting as i'm still waiting on a part to finish my lathe spindle build.
I have an issue with threading however, still working on this, it will run the first pass of G32 but as soon as it reaches the end of the block it will e-stop with the message "E-Stop button pressed"
Cant figure out what is making it think a button was pressed or what settings are wrong yet.
The motor is still wired for the main mill motor/vfd but it does have an encoder and will happily rigid-tap.
This is with the CS-Labs IP/A controller.
Any ideas?
-
Have you tried a G76 to see if that exhibits the same problem?
You can easily do one via the simple threading wizard.
-
I'll give that a spin tomorrow, thanks
-
Have you tried a G76 to see if that exhibits the same problem?
You can easily do one via the simple threading wizard.
I had a look but having never used that screenset or a wizard before it didn't work out well :(
Could not make sense of the numbers it was generating, looked like inches maybe? But i'm 100% metric, it also had soft-limit error on x IIRC but could not see why.
-
It will likely be in inch when you first open but you just plug in your relevant metric numbers. Now the soft limits could be due to the fact that the wizard sends the axes to the tool change position you have chosen as a G53 which to me is a pain and I changed the code in the wizard to a G54, so either edit your code to G54 or choose a G53 toolchange position within your limits.
Below is a screenshot of some settings you could try. Note that there are default values you will want to change as well, they are available when you press the Settings button, once you have changed them you are unlikely to need to change again.
-
Thanks for that, got some minor advance now.
I took out the tool change and the g53 line.
it failed first time on speed limit exceeded so i took it down to S500 instead of S1000
Then it ran, it made one pass then instant e-stop as before.
-
Can't see how the threading code could cause a false E-Stop unless there is more vibration than normal that is activating a limit switch. Even then I would think the message would be limit rather than E-Stop. I personally would disable limits etc as a test but obviously that is not the best solution as you would then only have the Reset on screen to use in an emergency.
-
Doesn't appear to be a limit, no LEDS are shown in diagnostics etc, it just says e-stop pressed the install the second pass is called in either G32 or G76, even typing code manually in MDI has same effect - 1st pass perfect then as soon as 2nd pass is called - estop.
I have not got my servo spindle fitted yet - this is still running on the VFD spindle but it does have an encoder and will happily rigid-tap so i can't see why it won't thread cut.
Is it possible the machine is tuned too slowly? Again, I can't see why that would be an issue but?
-
Can't really see being tuned too slow would be the problem.
Not really got any idea of what to try unless hard wiring the E-Stop and changing to another input temporarily to see if the problem continues.
Not sure if the CSMIO log would give you any info or not.
-
Just had a play and it is speed related, both in G76 and G32.
If i slow down to S100 it worked perfectly all the way through in both styles.
Using the SPO knob i could wind it up to about 175 but any more and it failed with e-stop again.
It must be connected with one or more of my axis setups????
Speed or acceleration?
-
Ok, seems its code related?
The setting in Fusion shown in the attache picture is the one, turned off I can run the code at up to 1100 rpm, with it turned on its 100rpm
After 1100rpm i get speed errors which i guess is understandable.
I have no idea why G76 failed though - no connection with this unless it too has a "fade at end" option ??
-
Not sure what "Fade thread end " means unless it is chamfer. Try setting the chamfer to 0 in Mach and see.
I would however think that would make things worse as it would mean instant pullout rather than a progressive one, having said that I think the chamfer in Mach is degrees so if that doesn't help by setting to zero try something like 720 and see.
-
From what i can see, its supposed to withdraw the tool at a taper as it finishes the pass, so its the stock end not the front.
Without it the spindle does a few turns before retracting the tool so i guess a groove will be cut at run-out?
Will keep playing but at least its progress.
If i had the last parts for my spindle build i could get it fitted :(
Thanks for the help as always.
-
That is what Chamfer is, the distance of the start of the retract at the end of each pass.
The CSMIO and Mach however have an issue when threading, the commands from Mach to CSMIO are delayed at the end of each pass and you end up with an annular groove. It is Mach that is the problem as I did extensive tests in the early days and eventually found an old version of Mach that behaved properly, sadly there were things missing from that version which were lathe specific which I wanted so I had to go back to a newer version. As it was only on the wee Conect 121 lathe that I had the CSMIO on it wasn't an issue as it was just used for messing around. It was however the reason why I never put the CSMIO on the Churchill lathe as often threads with an annular groove were not acceptable for me.
With the SZGH control I have just recently fitted the difference is night and day, everything with it is instant, it just lacks some of the adaptability that Mach can give you.
-
Thanks Hood,
You can see it will form a groove even when air-cutting, i did find a few threads on the Autodesk forum about this and the result was to turn a relief groove, i used to to this anyway when thread cutting on the manual lathe so i guess it will work.
The delay could well be why the fade/chamfer fails.
I keep looking at other controls, even the CSL SimCNC but nothing has the stuff i like to play with like Mach3 does :( I want to customise my screens, have things like tool length comp taken to the knee, easy tool-change macros, macro code that speaks my language (vb) ;) and so many more. I'm just a Mach3 dinosaur i guess.
-
Here is a vid of the way the CSMIO/Mach3 normally work. I am altering the spindle override in the vid to show how the CSMIO can track but it does show the slow retract well.
https://www.youtube.com/watch?v=sKp343Ubf7U
Here is one of it working properly which I thought was a new plugin but in the end I discovered it was because I had used an old version of Mach, 022 if I recall.
https://www.youtube.com/watch?v=8c1sndp-554
-
Mach3 has a lot to offer but it has its down sides as well. Mach4 looked to be like it was going to be great, it probably will be and maybe even is, as I haven't tried it for years. The downside for me was that I would have to learn Lua and I wasn't willing to do that as I had enough trouble with learning VB.
I will at some point try out SimCNC on the Chiron but again that will mean learning a new language so it may not progress very far.
The SZGH on the lathe seems to be working well for me so far and I have been able to do all I want with ease and the programming language seems to be a kind of home brew but is very simple which suits me :D
One thing I really like with the SZGH is I can press Hold and it is instant, I can change screens, look at parameters or whatever while it is running and there are no issues.
As said it is not as adaptable as Mach3 but no way I would go back now on the Lathe, mill is a different thing and I am quite happy with the CSMIO and Mach on the Chiron but the instant feedhold and overrides would be nice :D
-
Same sort of reasons here, i like VB.
My main issue i think is that my mill is tweaked to hell and back plus now it's a lathe as well - i can't see a commercial controller hacking that somehow :)
Its a pity dev was stopped on Mach3, i can fully understand why but Mach4 is still way behind. SimCNC is promising but getting info on customising from CSL is slow to negligible. I do have the test license from them and i have it installed on my laptop but thats as far as i got because there was no way to get my tool-comp setup to run on it, maybe it has now, will have another look some time.
-
Hi Dave,
i can fully understand why but Mach4 is still way behind
Have you tried it recently? By my estimate its Mach3 which is behind.
I suspect you would be very disappointed with CSMIO Mach4 plugin though.....its pretty damn buggy and CSLabs have
done nothing about it for months and months.
Craig
-
I will admit to not having looked lately.
I have read plenty about the buggy CSL plugin, I have a feeling this is caused by them shifting all their staff over to developing SimCNC maybe?
-
I think the chamfer in Mach is degrees
That is correct Hood ( G76 MACH 3 threading cycle it's the "L" value ).
The “L” retract or chamfer angle is in "angle of revolution".....or a "rotational angle".
and not an actually a chamfer. L90 would retract over 1/4 of a revolution, L360 would
be 1 rev, 720, two revs and so on. Don't think you will find that info in any of the code descriptions.
In the threading wizard the defined length "includes" the the distance of the L retract distance.
I don't know if the retract distance is included in G76, but you posted that it is not in reply #49.
Too busy / lazy to check it all out.
I have not used Fusion 360 for some time thus do not know what changes have been done.
I know that with the correct post processor it did generate correct G76 code for Mach3.
Just to avoid any confusion the above is for Mach# ver .062 and not using an external motion controller.
Most of the times for threading I use G32 output since one can monitor and pinpoint problems in
threading. The code is generated using wizards and they work great.
BTW,
How the end of the threading pass is done can vary depending on what is required as you said.
From an engineering point of view the end condition infuences a stress intensification factor which
could could reduce the loading of the thread. FWIW ;)
RICH
-
Interesting, thanks.
I did try the "Cycle" option in Fusion, its supposed to output G76 canned cycle thread code, however the post processor fails and says its not supported and to use the G32 option :)
-
Dave,
Fanuc, Mach3, Mach4, etc all have Gcode but individual dialects so to say. So threading cycles can vary in
G# and function. KISS is becoming more important as time moves on for me.
Mach4 / Mach3 ........ Like you, I am also dated since Lua is a place to drink, eat, and be happy. No pun intended Graig. :D
RICH
-
I just got a reply on Autodesk support, it seems there is a problem with the post processor i attached to my query.
Is there a recommended one for Fusion and Mach3 Turn??
-
The lathe post processor I used in Fusion has the following info when opened:
Copyright (C) 2012-2016 by Autodesk, Inc.
All rights reserved.
Mach3 Lathe post processor configuration.
$Revision: 41301 18455ff67c8df5aa948736e62844bc67de5c8913 $
$Date: 2017-01-27 13:44:58 $
FORKID {506EED9A-A4C0-40D3-BC83-CEC4CE91AAEA}
*/
description = "Generic Mach3 Turning";
vendor = "Artsoft";
vendorUrl = "http://www.machsupport.com";
legal = "Copyright (C) 2012-2016 by Autodesk, Inc.";
certificationLevel = 2;
minimumRevision = 24000;
longDescription = "Generic turning post for Mach3.";
----------------------------------------------------------------------------
The following is the directory of where Mach3 Turning CPS files are located on my computer:
C:\Users\user\AppData\Local\Autodesk\webdeploy\production\890691bd91ea0296c1b94572fdc7dda4ba2e48f0\Applications\CAM360\Data\Posts
---------------------------------------------------------------------
Note that I did modify the CPS file to remove unwanted code that it generated and frankly you really need to know how to do and what to look for.
There is an "hour long video" on how to modify a post processor, but guessing only that video was on You Tube....so could be wrong.
Sorry, but, it's been over a year since I used Fusion.
RICH
-
Thanks, slightly different, this is the current one..
/**
Copyright (C) 2012-2018 by Autodesk, Inc.
All rights reserved.
Mach3 Lathe post processor configuration.
$Revision: 42207 5fa31bcb4a6a27a706165056de7dcda5f53f36d9 $
$Date: 2018-12-05 15:57:57 $
FORKID {506EED9A-A4C0-40D3-BC83-CEC4CE91AAEA}
*/
I will seek the video and have a look. It would have been handy if the Autodesk guy said where the problem was ;)
-
Ooh, exciting :D :D
Managed to successfully edit my post, it now outputs what i want, instead of G28 X.. Z.. it checks the post properties for "G53 home" values, if i enter one i then get G53 X.. Z..
Sounds simple but took a few hours to learn what does what and where to stick brackets etc ::)
The post had the G53 bits but was not used where i wanted it, just for a parts-catcher :)
-
Just got a reply from CS-Labs regarding G32...
N22 G32 Z-29.816 F1.5
N23 G32 X10. Z-30. F1.5
You can't use two G32 commands one after another. You have to divide them with G0 or G1 command.
The limitation is caused by Mach3 construction.
Mach3 wasn't originally adapted to the G32 function executed by an external controller. We made it possible but with the above limitation.
That explains a lot i think, i don't know if adding a G0 in the code will affect things or if its better to stick with no taper at the thread end???
-
I'm setting up my servo drive for the 4th axis, I have not fitted the ENC module yet - should i be able to run the motor without the ENC module??
The drive will lock up as expected when enabled but i cannot get any movement at all.
The manual is vague on servo spindles :( and I have no idea of what to set and where - ports & pins, motor tuning, CS plugin etc.
The 'net seems just as vague as well :(
Any chance someone with a CSMIO IP/A controller and a servo spindle could enlighten me as to what settings i need and where please????
If the post comes, i'll be able to fit the ENC module but i thought that was only for sync when threading??
-
you can run spindle without ENC module.
you have to set analog Output in CSMIO plugin config.
https://en.cs-lab.eu/wp-content/uploads/2016/06/user_manual_csmioipa_en.pdf
see chapter 10.6
-
btw witch servo Controller are you using?
-
Yeah done them, i'll go back out and re-check though.
Do you have to enable Spindle DAC as well as spindle axis? The manual shows both which seems a bit odd as i thought the DAC was for 0-10v VFD outputs?
Do ports & pins plus motor tuning have any bearing at all???
-
btw witch servo Controller are you using?
AASD Chinese from AliExpress
-
are you running this Controller in step/dir mode or 0-10V analog mode?
-
are you running this Controller in step/dir mode or 0-10V analog mode?
I just ordered the same unit as previous and they were all pre-set for 10-0-10 analogue control, i guess it could be set wrong?
If it is then that will be a problem as the manual is hopeless :(
-
think you mean this Manual:
https://www.machsupport.com/forum/index.php?action=dlattach;topic=33543.0;attach=44811
it's not hopeless, only hard to read, but the Controller is the chineese Standard servo Controller witch is sold under
maybe 150 different Labels.
back to the Problem.
for a quick test you can set the enable Input a connect a 1.5v battery to the analog Input (pin25/13 ) to have a look that it is
in analog mode.
as far as i know CSMIO can not be Setup to +/-10V spindle control (only for the moving axis), only 0-10V and
use a servo Controller Input for CCW Operation.Mach3 has only the possibilty to Setup one Output for CW and
an other Output for CCW spindle Operation. therefore you have to use two relay's one for CW (only one NO contact
needed) and the CCW relay (two NO contacts needed).
then the CW relay Switches only the SRV-ON
and CCW relay Switches SVR-ON and CCW Input.
for the Moment i see not in the Manual, how to get a CCW Input confiured, but will Keep on studying.
-
I have just been poking with a meter and the CSMIO is outputting about +8v at present so its trying to get the drive to spin.
If i click the backwards jog button in plugin config/pid tuning the voltage drops to negative so CW/CCW seems to be 10-0-10 as expected
I checked the 25 pin connector at the drive and the plug is receiving the voltage from the IP/A
The drive is configured for location mode and is pre-programmed for the correct motor according to the values shown in parameters.
I can't do much more - if its not moving with 8v signal it won't move with 1.5v surely?
Hopefully not a dead drive :(
-
for the enable (SVR-ON) you have
24V to pin and
0V switched to pin 6 ?
-
Hi
Yes, i used the same wiring connection layout as the other 4 drives.
I just pulled the control plug, manually enabled it via Pn003=1, the motor locked as expected, then applied 1.5v to pins 13 & 25.
Result was no movement at all.
Seems its drive related at least.
Is there any more i can check or is it time email the seller :(
-
i think it should bo in Speed mode not in Location mode.
compare Parameter Pn002 with the other drives.
-
Will check another drive, back in a minute....
-
It was in Location mode pn0002=2 and the other drives are all =1 so changed it to 1 but still dead.
-
what is the Controller Display showing after power up?
have you saved the Parameters via Fn001?
-
Hi
i cant recall what it was showing but if you twist the motor the numbers go up and down :)
Didn't know about para 1, will have to be later as have hospital visit to go on now ;)
-
not Parameter 001 auxilary function 001!
see 3.4.3 in Manual.
but if you turn the Motor by Hand it seams not to be enabled at all.
-
When its enabled, the motor is locked as expected, it just does not move when told.
Will check aux function 1
Ta
-
check pn0002 and pn0003 as well
-
Yeah i'll set those again in the morning, Pn002 was 1 for speed mode, pn003 was 1 for manual enable or 0 for external enable.
-
any success?
-
Hey Dave , your email is rejecting me . It says I am spamming you (;-)
(;-) TP
-
Hey Dave , your email is rejecting me . It says I am spamming you (;-)
(;-) TP
WTF :) Thats the last thing i want ;)
I did just get one from you and i replied.
Dave
-
any success?
Yes, at last, i reset those two parameters again and enabled it and it sprang into life :) - If had rebooted yesterday it would probably have done it then too.
So, moving on.....
Pid Tuning :(
Of course, CS Labs in their infinite wisdom have removed the Auto-Tune from the spindle servo page :(
I have been plugging numbers for an hour now and although it runs its not happy. It will overshoot then slowly crawl back to dead-band stop, or it will run ok at lower speeds (<1500) but at 3000 it will e-pid fault on max following error accumulator being massive (>100000)
I've had it jumping about so violently i had to e-stop :) and had it sit there slowly moving back and forth but never stopping.
Any ideas how to pid tune a spindle so it can run at 3000 and stop when told?
Getting there but no champagne yet :(
-
I set the Chirons spindle up as an axis, did an AutoTune then changed it to a spindle and entered the settings, I then tweaked from there.
-
Ooh, thats an idea :)
Thanks
-
I set the Chirons spindle up as an axis, did an AutoTune then changed it to a spindle and entered the settings, I then tweaked from there.
Ok, just tried that, she was not happy.
It started to tune, set some low values and then just started thrashing back and forth wildly, had to e-stop to stop it.
It seems i can get somewhere with P=3000, i=0, d=0, kff=0 but, it's not balanced in behaviour FWD/REV so FWD seems not too bad but REV it seems to continue after the signal is removed then stop instantly. I don't know if it makes a difference but the following error is massive too.
DAC offset is 0.4v which is inline with other drives.
A short vid showing 2 FWD runs and 2 REV runs
https://youtu.be/k3MQxbtKppQ
-
try to set Pn146 to 0(hope this will affect acceleration too).
It seems i can get somewhere with P=3000, i=0, d=0,
a PID with I Zero is bad you have to have a I part, D part i would say in this case is not
necessary.
a PID with I Zero will allways go back and forth (swing).
-
DAC offset is 0.4v which is inline with other drives
that seems to be very high DAC Offset is normaly adjusted by set Speed Zero and then adjust
Offset so much until Motor standstill.
-
In CS-Labs you just press the auto set zero and it stores the offset voltage.
-
have you tryed to set Pn146 to 0, this will Switch of the ramp Generator in the Driver (i think).
-
have you tryed to set Pn146 to 0, this will Switch of the ramp Generator in the Driver (i think).
I'll be doing that this morning, few jobs to get out of way then i can get back on it ;)
-
Seems a tiny bit better but not working yet
The voltage from the CSMIO DAC is off-centre or not linear?
With the drive showing input signal voltage, i tested various speeds and its ok up to 1800rpm or 6.2v and works ok FWD or REV
At 1900rpm it goes out of balance, FWD runs ok at 6.46v but reverse jumps to 9.98v and runs flat out, when you remove the command it runs on for a second or two then stops suddenly.
No point tuning if i cant get this sorted.
Any ideas?
-
Any ideas?
not realy, maybe use another free analog Output?
can not test this here, the IP/S does not Support spindle axis with analog out only step/dir.
-
Ok, changed DAC channel, no change at all so its not the controller.
Could it be encoder related?
Here's a short vid...
https://youtu.be/Sbmr4-Z9T8U
Shows what happens, what makes me think encoder is that it will run ok forwards and the following error stays high but stable, in reverse it will run away above a certain speed and the FE will keep rising.
Could a bad feedback be telling the controller to send more voltage hence the voltage being wrong giving me a false lead??
Slowly doing my head in now.
-
watched the Video several times, not sure what you mean.
it can be Encoder related, i think you are using Encoder Output from servo Controller?
how are Pn016 and Pn017 set?
-
Yes, signal goes Encoder -> Drive -> Controller
Will check those parameters now....
-
pn0016 & 17 set to 1
I have reached my conclusion this is encoder related.
Why?
Setting "open loop" in plugin config removes the encoder loop from the controller and i can request 0 to 3000rpm FWD and REV from Mach3 MDI input using the usual M3s3000 or M4s3000 and it works perfectly, it will also allow direct M4-M4 swaps and will slow down, stop, reverse and speed up ok.
So my conclusion so far is *********e encoder or bad signalling.
I'm digging the 'scope out to have a look-see whats coming out of the drive, not sure it will make much sense to me but i'm gussing to look out for a TTL level square-wave from ground to each A+ & A- with the two being inverted?
-
Two pics below, one is A channel the other is B channel, both crap in my view
Will now try and get signal as it comes from the drive socket but will take a while.
This was captured with motor turning at 100rpm
-
Its a crap cable, signal was clean at the drive socket, changed the cable and plug and things seem better :)
Still can't use any i or d values in the pid but the p is now up around 12000 which is similar to other axes.
Adding any i makes it over shoot and creep back.
Ah well its a bit of a success i guess, thanks for the help so far.
-
happy to hear that you have figured it out. :)
would be interested witch Setup you are planning for lathemill?
-
At the moment i just have a QC tool post mounted on the quill using the bracket i made for a high-speed spindle (no longer used), the lathe is just a 5C shaft mounted with some deep-groove bearings in some 1" alu tooling plate blocks which i can clamp down to the table when needed. It has a 750W AC servo at 1:1 on it so plenty of power for a small shaft hopefully.
I can't do mill-turn because the quill is holding the tooling and a Bridgeport has no Z axis apart from the quill so i can't use a gang tooling bar at the same time. It's a bit of a limited lathe as such but hopefully will do what i wanted and means i don't have to try and shoe-horn another machine in ;)
-
Getting there...
Mashed a few tips, learnt a few things ;)
Made this little test, concentric diameters, came out well but diameter was +0.3mm so my X zero setting must off by 0.15mm??
Need to get some tooling now.
Also need to figure out how to reset it after removing the spindle and tool mount??
I guess the the tool-table offsets will remain ok as long as i don't alter the tooling, so will it be a simple as resetting X on the master tool and away we go??
-
If you have accurate homing, which you should with an IP-A, then I would use Home position as the "master tool", that way all your tool offsets are from the X (Z on mill) home position. It means if you need to replace a tool which would have been the master tool then you do not need to set up every other tool again.
As long as your tool holder is repeatable then the tool tables offsets will be as before.
-
Makes sense, just ;)
Trouble i have (I think) is that the spindle is an attachment with no fixed abode - it will have to come off for milling etc so
I have stored and added a button to send the knee and Y axis to a preset location so thats two axes out of the way but X & Z still happen....
-
If you have accurate homing, which you should with an IP-A, then I would use Home position as the "master tool", that way all your tool offsets are from the X (Z on mill) home position. It means if you need to replace a tool which would have been the master tool then you do not need to set up every other tool again.
As long as your tool holder is repeatable then the tool tables offsets will be as before.
Can you elaborate a bit more on this if you get the time??
I do have index-homing but can't get my head round how this method works
:)
-
https://www.youtube.com/watch?v=Kla1sZ8IF6k
-
https://www.youtube.com/watch?v=Kla1sZ8IF6k
Thanks Hood,
So you have no "master" tool?
I'll test this tomorrow hopefully but does this work when the numbers are oddly large?....
I see your lathe homes pretty close to the work area, my mill homes bottom left so my "lathe" Z area will be about 300mm plus from home zero.
Its hard to figure without doing it but all offsets are relative to each other? You can touch off with any tool?
Provided of course the tool number is called first I guess?
-
The home position is the "master tool" all tools are offset from there so it means you can replace any tool and reset the tool offsets for it and not have to worry about it messing with any other tool. As all tools are set to their respective diameters from the home position you do not usually need to alter the X value for your G54 (or other) offset. The Z G54 will need to be set for every different part you do but that is easy enough as you just call any tool you like, move it to the end of the stock and then zero the Z work offset. Doing that will mean every other tool will also now have their Z Zero work offset position as zero when they are touching the end of the material.
-
The home position is the "master tool" all tools are offset from there so it means you can replace any tool and reset the tool offsets for it and not have to worry about it messing with any other tool. As all tools are set to their respective diameters from the home position you do not usually need to alter the X value for your G54 (or other) offset. The Z G54 will need to be set for every different part you do but that is easy enough as you just call any tool you like, move it to the end of the stock and then zero the Z work offset. Doing that will mean every other tool will also now have their Z Zero work offset position as zero when they are touching the end of the material.
Sounds excellent, will play with this.
Thanks
-
I see your lathe homes pretty close to the work area, my mill homes bottom left so my "lathe" Z area will be about 300mm plus from home zero.
Oh and that lathe was just a wee lathe I got to test out the IP-S when I first got one. I no longer have the lathe, I took the IP-S off it and put it on the plasma and gave the lathe to a friends son.
My real lathe is a wee bit bigger and probably has home positions similar to where they will be on the Bridgeport, the travel on it is approx 1000mm for Z and 420mm for X but I home approx midway on both axes.
Until recently I had the ESS on it but I have just put an SZGH control on it.
-
Nice, is that controller self-contained so no pc to fail etc??
Could it replace Mach3 when you have trickery going on with tool length being passed to a knee axis or is that stuff still the sole domain of Mach??
Hows the plasma running ?
-
The home position is the "master tool" all tools are offset from there so it means you can replace any tool and reset the tool offsets for it and not have to worry about it messing with any other tool. As all tools are set to their respective diameters from the home position you do not usually need to alter the X value for your G54 (or other) offset. The Z G54 will need to be set for every different part you do but that is easy enough as you just call any tool you like, move it to the end of the stock and then zero the Z work offset. Doing that will mean every other tool will also now have their Z Zero work offset position as zero when they are touching the end of the material.
It works, I couldn't figure it out but once you try it, it makes sense, thanks Hood.
Next - my servo spindle drive seems odd.
At up to 2990 rpm in CS-Labs Pid tuning, it runs nicely with a low following error, about 1500 IIRC, at full speed of 3000 rpm I see a runaway following error - just carries on rising and would eventually trip an e-Pid fault on max following error i guess.
Its not serious, its only 10rpm lost but seems odd??
-
FWIW,
I also do similar to Hood's reply #112, but, no switches used here, so home and any work offsets are
set manualy.
By similar I am meaning, all tools have the same base location from which they move to touch off.
Tool offsets are probed instead of touching off or machining, taking a measurements and inputing info.
Probing makes populating the tool table quick, accurate, repeatable, and somewhat automated. BUT,
a tool setup page to accomplish the above was done.
Once a tool table is populated all tools relate to the master tool ( could be an actual master tool or
a common base location) and all tools relate to each other. Thus you can use any tool to touch off
to the work or setup to replace a tool.
Have fun.......,
RICH
-
I have just put an SZGH control on it.
Nice, is that controller self-contained so no pc to fail etc??
Curious too Dave! Didn't see it clearly explained on their site.
Thanks Hood,
Russ
-
The home position is the "master tool" all tools are offset from there so it means you can replace any tool and reset the tool offsets for it and not have to worry about it messing with any other tool. As all tools are set to their respective diameters from the home position you do not usually need to alter the X value for your G54 (or other) offset. The Z G54 will need to be set for every different part you do but that is easy enough as you just call any tool you like, move it to the end of the stock and then zero the Z work offset. Doing that will mean every other tool will also now have their Z Zero work offset position as zero when they are touching the end of the material.
It works, I couldn't figure it out but once you try it, it makes sense, thanks Hood.
Next - my servo spindle drive seems odd.
At up to 2990 rpm in CS-Labs Pid tuning, it runs nicely with a low following error, about 1500 IIRC, at full speed of 3000 rpm I see a runaway following error - just carries on rising and would eventually trip an e-Pid fault on max following error i guess.
Its not serious, its only 10rpm lost but seems odd??
sounds like you have reached the 10V at 2990rpm.
-
This is probably not really the place we should be discussing other controls so I will keep things brief.
Nice, is that controller self-contained so no pc to fail etc??
Yes self contained but on the other hand if something fails it more than likely means a new controller
Could it replace Mach3 when you have trickery going on with tool length being passed to a knee axis or is that stuff still the sole domain of Mach??
That may be possible as you have 9 user macros and the tool change macro itself which you can fully edit. Also the in built PLC can be edited if you can figure out what is in it at the moment (no comments/names)
Hows the plasma running ?
Plasma is running great, don't know how I ever managed without one :D
-
The home position is the "master tool" all tools are offset from there so it means you can replace any tool and reset the tool offsets for it and not have to worry about it messing with any other tool. As all tools are set to their respective diameters from the home position you do not usually need to alter the X value for your G54 (or other) offset. The Z G54 will need to be set for every different part you do but that is easy enough as you just call any tool you like, move it to the end of the stock and then zero the Z work offset. Doing that will mean every other tool will also now have their Z Zero work offset position as zero when they are touching the end of the material.
It works, I couldn't figure it out but once you try it, it makes sense, thanks Hood.
Next - my servo spindle drive seems odd.
At up to 2990 rpm in CS-Labs Pid tuning, it runs nicely with a low following error, about 1500 IIRC, at full speed of 3000 rpm I see a runaway following error - just carries on rising and would eventually trip an e-Pid fault on max following error i guess.
Its not serious, its only 10rpm lost but seems odd??
sounds like you have reached the 10V at 2990rpm.
I'll check that, thanks, i'll also put a tacho on it and see where its at ;)
-
This is probably not really the place we should be discussing other controls so I will keep things brief.
Nice, is that controller self-contained so no pc to fail etc??
Yes self contained but on the other hand if something fails it more than likely means a new controller
Could it replace Mach3 when you have trickery going on with tool length being passed to a knee axis or is that stuff still the sole domain of Mach??
That may be possible as you have 9 user macros and the tool change macro itself which you can fully edit. Also the in built PLC can be edited if you can figure out what is in it at the moment (no comments/names)
Hows the plasma running ?
Plasma is running great, don't know how I ever managed without one :D
Oops, my bad forgot about that ;)
Sounds interesting though, and yes the plasma table is my most used machine ;)
-
FWIW,
I also do similar to Hood's reply #112, but, no switches used here, so home and any work offsets are
set manualy.
By similar I am meaning, all tools have the same base location from which they move to touch off.
Tool offsets are probed instead of touching off or machining, taking a measurements and inputing info.
Probing makes populating the tool table quick, accurate, repeatable, and somewhat automated. BUT,
a tool setup page to accomplish the above was done.
Once a tool table is populated all tools relate to the master tool ( could be an actual master tool or
a common base location) and all tools relate to each other. Thus you can use any tool to touch off
to the work or setup to replace a tool.
Have fun.......,
RICH
I do have a probe for the milling tools but with the different styles of lathe tool i couldn't see how to make it work for me. With the low count of tools i'll need i don't think it will matter.
So using Hood's method, the tool table will be relatively safe unless the home switches are tampered with in some way?
My next question is - how do you replace a tool or add another at a later date and keep its relationship with the rest correct?
I can see how it works if you populate the library in one go but if i remove and replace the spindle etc surely i have lost my reference??
-
So using Hood's method, the tool table will be relatively safe unless the home switches are tampered with in some way?
Nope, you can then just use the work offset to offset the X value and that should offset for every tool. You already do that for the Z Axis every new job when you zero to the end of the stock.
My next question is - how do you replace a tool or add another at a later date and keep its relationship with the rest correct?
I can see how it works if you populate the library in one go but if i remove and replace the spindle etc surely i have lost my reference??
The relationship of each tool is to the home position and not to each other, that is why it is better than using a master tool because in that case all tools are in relation to your master tool, replace that and you need to redo all.
Not sure what you mean about remove and replace the spindle. Is your tool holder not a defined position in relation to the spindle taper?
-
So using Hood's method, the tool table will be relatively safe unless the home switches are tampered with in some way?
Nope, you can then just use the work offset to offset the X value and that should offset for every tool. You already do that for the Z Axis every new job when you zero to the end of the stock.
My next question is - how do you replace a tool or add another at a later date and keep its relationship with the rest correct?
I can see how it works if you populate the library in one go but if i remove and replace the spindle etc surely i have lost my reference??
The relationship of each tool is to the home position and not to each other, that is why it is better than using a master tool because in that case all tools are in relation to your master tool, replace that and you need to redo all.
Not sure what you mean about remove and replace the spindle. Is your tool holder not a defined position in relation to the spindle taper?
Ok, i'll test the relationship ;)
Sadly nothing is 'fixed' on my setup - the spindle is a removable fixture and the tool-post block is clamped to the quill but without 100% location radially so there may be some small misalignment each time its fitted, in the axial direction it should be good as it uses the BT30 socket as a tool would.
-
I would try and get some means of locking radially so that it is repeatable.
-
I'll work on that, won't be easy but there'll be a way ;)
-
Still trying to get this to gel in my head :)
I tested it out this morning with some stock in the spindle and two tools, called t0101, used a feeler and touched off the Z, stored that, same on X and stored that.
Same on t0202.
Then tried calling and touching off on t0101 and set my work zeros, called t0202 and the zeros were spot on = happy days.
But, what happens after i remove and refit my spindle or use a different bit of stock to set a new tool - the spindle has shifted and the stock has changed.
Surely the tool ref distance from home to touch position is now out of sync as regards the other tools?
I can only envisage it working if there was a point fixed in distance from m/c home - the headstock on a lathe, the face of a chuck etc - I have neither of these as all things are movable.
??? ???
-
Different bit of stock makes no difference. Call a tool, jog to the end and touch off and set the Z work offset. Note I am saying the work offset and not the tool offset. Tool offset has already been set previously so it should not require setting again other than possibly using the wear offset.
To set up a new tool at a later stage is simple, just the same routine as before except first touch off the stock with a tool already set up and zero the Z (or to feeler gauge thickness if using one) then change to your new tool, touch off on the end of the stock and set the tools Z offset as previously done with the first tools. Do the same for the X by taking a cut and measuring. New tool is now set up.
If you think you will have repeatability issues on the X due to replacing the tool post then you can load a known tool, take a cut and then measure and if different from the X value of the work offset then just change the work offset, all other tools will also use the new work offset so will also be correct.
-
To set up a new tool at a later stage is simple, just the same routine as before except first touch off the stock with a tool already set up and zero the Z (or to feeler gauge thickness if using one) then change to your new tool, touch off on the end of the stock and set the tools Z offset as previously done with the first tools. Do the same for the X by taking a cut and measuring. New tool is now set up.
Hmm, ok
When you first store your tool z offsets, you stored the machine co-ordinate yes?
Para above - when you say "touch-off an already set up tool and zero the Z", you mean Z work offset?
I still can't see the correlation unless i am wrong in my first assumption?
-
Don't think too hard about it, it works ;)
-
BTW I assume you have added the DROs (175 and 176) and Touch off buttons (324 and 326) to your screenset, also the tooltable shortcut button (121)
If you have then they do all the calculating for you and enter the correct numbers into the tool table when setting up the tools initially.
Setting the zero of the stock is done with the normal Z Offset DRO with tools that have previously been set up in the tool table.
-
BTW I assume you have added the DROs (175 and 176) and Touch off buttons (324 and 326) to your screenset, also the tooltable shortcut button (121)
If you have then they do all the calculating for you and enter the correct numbers into the tool table when setting up the tools initially.
Setting the zero of the stock is done with the normal Z Offset DRO with tools that have previously been set up in the tool table.
I do now ;) - That was the missing link. My first understanding of your method would work but only if things did not move around, having these DRO's makes the difference.
Will hopefully have a play later, also looking at ways to locate the tool mount repeatably too.
Thanks for the patience BTW ;)
-
On my own screenset I put VB buttons rather than the X and Z touch off buttons and just put the following code in (change to 326 for Z)
DoOemButton(324)
Sleep 100
DoOemButton(121)
Doing that means the tooltable opens automatically saving you the extra button press or the possibility of forgetting to open the tool table and accept/save.
-
Ok, could add 316 instead of 121 and just save the table without viewing ;)
-
You could but if you are setting up tools likely you will be wanting to give them a descriptive name and thus you would have to go back into the tool table again to do so ;)
-
Still having fun with the setup ;)
If i take a light cut, mike it up and set the dia in the X DRO, why is it that if i then move the X to say 1mm less i.e. a 0.5mm cut, the size is wrong?
Seems to come over size each time?
-
First thing to say is don't alter the value in the X Dro unless it is all tools that are off by the same amount because you will be changing the work offset and that affects all tools.
Two ways to adjust a tool is by either resetting its values in the tool table by using the method previously mentioned (X Touch off DRO/Button) or use the wear offset DROs to add or subtract the entered amount from that tool. It is best to use the latter if it is indeed just wear but otherwise I preferred to use the former.
One thing to also mention, the last release of Mach was screwed up in quite a few ways for Turn so avoid it, think however you are using an older version.
-
Yeah i got the idea of tool wear comp, handy..
But this was just one tool, take a pass, ok, i did enter it in the X dry so thats one possible error, then immediately reduce X and take another pass - the second one comes out wrong size.
The tool was not removed, not called, just one pass, reset DRO, another pass, wrong result.
Can't see why yet??
Yes, i could 'tune' the result to be correct but from one pass to the next??
I need to test again without altering the DRO, going to be hard to get out of doing that as i do it all the time on the mill ;)
-
The tool table can perfectly reflect the tool offsets.
Just understand that everything associated with and when using / moving the tools controlled point can influence the final cut dimension.
As Hood stated wear and offset adjustments can be used, but, for the most part, I leave the original populated values for the tools alone unless the tool wears or breaks. My tool table has a few tools with names call TEMPx and the x represents some modified version of the original tool #. Logical or meaningfull names that reflect the tool
is helpfull. ie; Threading tool , then copy rename it to LRTO ( Left Right Turning Outside) and with a adjusted offset.Do what ever makes sense to you.
I added an X and Z adjustment DRO's and a button so I can adjust the current / any tool in the tool table. It realy comes in handy for drills, reamers, and other tools used on my lathe and when touching them off.
Using the tool to probe the work or a setter is a thumbs up over the cut and measure method. Just my thoughts and to each their on how they want to work.
RICH
-
Can't see why yet??
TOOL OFFSET ADJUSTMENT
1. Tool accurately machined the stock to dimension.
- no adjustment all ok
2. Tool when used caused the stock to be over dimension.
- The tool offset needs to changed such that the tool moves in more / removes more
material. So the tool offset needs to be increased.
3. Tool when used caused the stock to be "under" dimension.
- The tool offset needs to changed such that the tool moves in less / removes less
material. So the tool offset needs to be decreased.
Think the above is right! ::)
FWIW,
Machine coordinates are displayed in radii. The tool table offsets are in radii for the X axis.
KISS!
Be carefull when making manual adjustments to tools and consider radius or diameter mode.
I confuse easily.............. ;)
RICH
-
Thanks Rich, got that.
Two things,
1 - This was one cut after the other - cut, measure, move, cut , measure and the second cut was off. I will re-test this today hopefully.
2 - Does anyone know precisely what the tool DRO's/buttons actually do - 175/176 & 324/326 respectively???
-
Todays tests…
Tool was set by taking a random cut then measuring and entering dia into “touch X” DRO.
Next i requested a cut depth and noted the results, left is requested, right is result, error in brackets…
21.00mm = 20.69mm (0.31)
20.00mm = 19.62mm (0.38)
19.75mm = 19.47mm (0.28)
19.50mm = 19.23mm (0.27)
So the error is fairly consistent - 0.30mm would cure it, and it points towards the tool being set wrong by 0.30mm??? But how can it be wrong if i did a cut-measure setup??
Baffling me and clearly not right yet.
I tried another tool and got this…
21.00mm = 20.92mm (0.08)
20.50mm = 20.44mm (0.06)
These are more accurate but was setup in exactly the same way.
its a different insert shape tool.
Any ideas????
-
just an idea, is your tool cutting realy in the Center of the spindle?
-
Put a clock in the tool holder and command moves and see if the moves are accurate, going by your measurements I would say they will be but still worth checking. If in dia mode the clock will move half of the commanded movement.
As TPS has said centre height and orientation of the tool is critical for accuracy but it will only see small variances over small changes in dia so not likely your issue here.
Was your initial cut the same depth as your others?
Rigidity of your setup may affect things and thus setting up may require two passes to get right and doing the pass at normal DOC for a finishing pass is usually best on a set up with less than optimal rigidity
One thing to also consider with carbide inserts, it is often not a good idea to take a depth of cut less than the nose radius of the tool although it will depend on the inserts geometry as to how much it will affect things in doing so.
-
just an idea, is your tool cutting realy in the Center of the spindle?
Yes, it faced-off with just the tiniest pip in the centre.
-
Put a clock in the tool holder and command moves and see if the moves are accurate, going by your measurements I would say they will be but still worth checking. If in dia mode the clock will move half of the commanded movement.
As TPS has said centre height and orientation of the tool is critical for accuracy but it will only see small variances over small changes in dia so not likely your issue here.
Was your initial cut the same depth as your others?
Rigidity of your setup may affect things and thus setting up may require two passes to get right and doing the pass at normal DOC for a finishing pass is usually best on a set up with less than optimal rigidity
One thing to also consider with carbide inserts, it is often not a good idea to take a depth of cut less than the nose radius of the tool although it will depend on the inserts geometry as to how much it will affect things in doing so.
The depth of cut was variable, just pulling numbers out the air.
I have some more tooling on way, they use my favoured CCMT tips, the ones i'm messing with are more sacrificial and use triangle tips that i never really liked. I'll do some more tests when they arrive.
I'll also throw a DTI on there as soon as i get a gap this week, but i'm fairly confident it will be ok.
I did put a DTI on the quill when taking a cut to check for push-back etc but there was none, i also put some force on the quill by hand and again there was minimal movement so the quill looks pretty stable.
-
2 - Does anyone know precisely what the tool DRO's/buttons actually do - 175/176 & 324/326 respectively???
Precisely in general terms..............................
DRO 175 requires an input from the user. Button 324 uses the inputed value of DRO 175 and when pressed,
executes the following code:
x=GetOEMDro ( 175 )
call Set Tool (x)
For a description of Set Tool X see page 81 in the Mach3 Macro Programmers Reference Manual.
Now have a look at GetOEMDRO on page 24 and also look at SetOEMDRO on page 73.
Similar for 176 DRO / 326 Button which relates to the Z.
You actualy have three positional DRO's in Lathe, namely Machine Coordinates, Program Coordinates,
Part Coordinates MC / PC/ PartC respectfully. Additionaly there is Radius and Diameter Mode.
That said, depends, can't say "precisely".
RICH
-
That's odd because OEMDRO 175 and button 324 also affect/update the tool table as well, that is not documented though it seems?
-
i assume you have checked accurate x-axis moves via gauge and mdi moves.
so what is left? all your cut's are to small so IMHO only two things are left:
-spindle rack get's pulled up by cutting force
-tool get's pulled down by cutting force
just trying to think loud. ;)
-
Hi
yes the axes moves have all been checked and were all ok, all done with MDI moves, any backlash accounted for, but is minimal
I will do more tests when my cutters arrive, i do not like these cheap triangle tool holders, the bits are not located well and the fact i get different errors between two tools tells me not to trust the tool holders i think.
I have some SCLCR/CCMT and other shapes coming so will redo my tests when they are fitted.
-
Tried a lump of HSS and diameters came out as requested - best not to use cheap triangle insert tooling i guess :)
-
best not to use cheap triangle insert tooling
I sharpen mine first before using them on a universal grinder and they work fine.
Let me state that they don't seem to hold up under some conditions.
Carbide inserts grades and quality vary even between the best manufactures.
One problem with cheap carbide tooling is the grade of carbide and how made .....whatever it could possibly be..... even if stated!
Have a magnified comparison of the HSS and carbide tooling that you used and probably the nose radius is
smaller and cutting edges are sharper. Not surprised that they machined to spec.
Like I said before, the tool table can be right on as far as offsets but a lot of things affect machining to spec.
I sum it all up with the catch words " ones machines system" which is all inclusive!
RICH
-
Finally got some results i like. :) :)
Pic attached of a little test part, diameters requested were 11,14,17,20mm results i got were 10.97, 14, 17, 20 so it seems accurate enough i think.
Short video...
https://youtu.be/AlcihBYrKIA (https://youtu.be/AlcihBYrKIA)
After the facing cuts, the first few passes are blank because the stock was 1mm under size (running out of junk metal) ;)
Birds nest at the end as well, lots to learn.
Next up is the multi-grooving/turning tool, see if it can part-off without snapping my tool.....
-
happy to see you got it figured out.
so some cover's, some fixtures to get it off and on reproducable and a lot of learning about CNC turning (i think).
good Job :) :)
-
Yeah, pretty much.
Covers next but need to sort out why the bearings get so hot first.
Maybe too much preload, not sure.
Using deep groove balls which i know is not ideal but they do have a rating for axial loads and the radial load rating is off the chart for a spindle.
Fixtures not so important as it only takes a minute or two to tram-in.
-
A short video of some tool paths to create the part shown in my original post.
Does it look like it might work or i need to change something ?
The tool is only 1.5mm wide.
https://youtu.be/J3ZCPnOwcrs (https://youtu.be/J3ZCPnOwcrs)
-
Certainly not the way I would do things and may be prone to deflection depending on the actual tool used and the material being cut.
-
Hmm, ok.
Material is 6060 aluminium
You would groove the body away maybe, then just a final finish cut?
Was trying to go with just one tool :)
-
Have you got a dxf of the part and also dimensions or a link to the tool are you using?
I would use two tools normally, maybe even three but as I have a turret it is not an issue for me using multiple tools. If wanting to use a single tool and it is a multi directional tool then I think I would just be using a conventional style turning toolpath.
-
No dxf, its a simple shape, 23mm dia, 24mm long, 3mm rads on ends, 6060 aluminium.
The tool is a Korloy MGEHR with a MGMN150-G PC5300 insert, 1.5mm wide neutral angle, with clearance, max depth 14.5mm.
;)
-
Ok, I personally prefer the radius inserts with the MDT's and to save me the bother of setting a new tool up in BobCAD I just used the Seco one I already had in the tool table.
Here is how I would do it, not done the parting but you could use the same tool.
https://www.youtube.com/watch?v=_4YIGK5yl3E&feature=youtu.be
-
Interesting, not a million miles away from my attempt, i'll try CAM'ing it that way....
-
Just messing with the latest version of BobCAD and it makes setting up tools easier than the old version I am using so I decided to try your part with your tool.
https://www.youtube.com/watch?v=SGvw1F7qERQ&feature=youtu.be
-
Thanks,
Looks almost identical to my flavour I think, can't see any practical benefit for tool or work?
I checked BobCad and never knew it was that expensive ;) I'll stick with Fusion I think until they withdraw the free usage.
-
BobCAD is not as expensive as the listed prices suggest, or certainly wasn't.
Fusion isn't really free unless you are a student or a hobbyist, at least that is what I made of the licence agreement but maybe I am wrong.
-
Yeah, I count as a hobbyist because my turnover is way below the mark, luckily. When they move that mark, things will change and because its all cloud based I could lose all my work :( Being held to ransom really, hopefully it will continue for a good time yet though.
The rumour is that its free as we are effectively Beta-Testing it for them, long may it remain in Beta ;)
-
Ok, thanks to the tool path and dry run, i can see a problem, i think?
Arrowed in the pic attached is a path that bothers me.
This tool plunges ok but then makes a 45 degree retract - this will be cutting metal way deeper than the insert is allowed i think on its flanks.
I can see no settings in fusion that will stop this action, lead-ins and lead-outs are off.
Any ideas?
This is the code...
N20 T0101
N22 G54
N24 M8
N26 G95
N28 G97 S707 M3
N30 G0 X45. Z0.
N32 G48 S3000
N34 G96 S100 M3
N36 G0 Z-26.4
N38 X29.
N40 G1 X25. F0.1
N42 X21.986
N44 X23.986 Z-27.4
N46 X27.425
N48 G0 X29.
N50 Z-24.933
N52 G1 X25. F0.1
Line N44 seems to be the one.
-
Ignore the last post, a guy on Fusion forum just replied - its called "Back-Off" distance, setting it to zero cures the path.
:) :)
-
Dave,
Trying to use only one tool limits what and how you may be able to do things.The progression of machining steps
will vary but in the end the work gets done. Simplisticaly the thought process is just autonation and refinement of doing it manualy.
I use a quick change on my CNC lathe and serves me well. Just need to setup for tool change position and allow time to change tools.I KISS it by having home, referenced position, tool change position= Machine Coordinates 0,0,0 and define the G54 offset since my default G54 has no offset. Post cutoff off of a part one can just select another offset to make the same part again if you wish ....same gcode.
BTW, If I recall correctly :P G55 is used for rear mounted tool offsets, never mix front and rear tools in the tool table.
Hood can correct me if that is not correct.
There is no comparison of say LazyTurn to the higher end lathe programs, but, it can be a quick and dirty tool to learn and try different approaches to do work.
What is learned will have value in using the higher end programs.
Just something to think about...........,
RICH
-
Thanks
the idea or quest for one tool was not because of tool positioning etc - i have a QC tool post mounted and have 4-5 holders spare.
The idea was for simplicity - no tool changes is better than 1-2-3 changes (especially when you have no ATC) and with the tools i have seen available today, this part should be a walk in the park :)
I have indeed learnt a lot - no cnc lathe use before, just manual, and yes as you say its just automation - the skill is getting the software to imitate what would be done manually, tricky as i just found out when there are a myriad settings to tweak.
I only have one tool post, as far as CAM goes, its just a normal front mounted post lathe, my post is rotated 90 degrees so the tools face backwards as the mount is to the Quill but apart from that its bog standard.
Yes i did read warnings about not mixing front/rear posts in the TT but thats not an issue here. ;)
-
Ok so we have made a part (sort of) :)
Spindle axis ePid faults wrecked it by stopping dead mid part-off, twice.
So i still have settings issues with my servo, anyone have any thoughts here??
This is generally triggered by the following-error counter reaching its max setting, currently 10,000 steps or 1 turn.
When it occurred, the tool was maybe halfway in the slot.
The rest of the part went perfectly, finish was 100%, facing off was perfect, just the last bit :(
-
This problem has been there all along i think, pretty certain its the speed not a torque overload...
I can run ok up to 2980rpm, anything above that and the following-error runs away and it will fault out, same fwd and rev.
I have tried altering the voltage gain on the drive, tried altering the max speed on the drive
I can fudge it by setting Mach pulleys to 2980 but it seems some CSS code can get round that and will fault out when cutting as i found on my first run earlier.
I have tried other DAC channels on the controller, same.
Spent another day fiddling and got nowhere really. ;)
Any thoughts???
-
...
BTW, If I recall correctly :P G55 is used for rear mounted tool offsets, never mix front and rear tools in the tool table.
Hood can correct me if that is not correct....
It is best avoided, from memory you need to restart Mach for it to behave or at least rehome.
The solution I used was to set ALL tools as rear tools(or front if that is your default), that means even front mounted tools were set as rear. Then when you want to use a tool from the non standard toolpost you would call a G51X-2 (X-1 if using Radius mode) and that would scale things to allow your movements to be correct for the X axis, you had to remember to call a G50 when you wanted to move back to the defualt tool post or things could get nasty. In the end I removed the front turret as it was just too easy to make a mistake and smack a tool into the stock. I made an 8 position VDI turret up for the rear which meant swapping tools out was both quick and repeatable and I could have many tools preset ready for use.
-
Thanks for the reply Hood.
Did a quick look at front and rear tool post use since my recallection was not stated correctly.
WORK OFFSETS are manipulated when both front and rear tools and offsets exist in the tool table.
G55 is manipulated as related to the G54. The Part and Program Coordinates change to reflect a distance
relationship of front to rear tool. That is the extent to which I explored the use of front and rear tool posts.
FWIW,
RICH
-
But any ideas why my spindle faults above 2985 rpm?
-
Afraid not, the servo I have on the Chiron is 3800 rpm, it runs at 3800 rpm and will not fault even if I called 10,00rpm as the max it will do is 3800 rpm. Kind of sounds like there is a drive issue with your setup but what that could be I have no idea.
-
Ok thanks Hood.
No idea what to do here really, i should just have bought a small CNC lathe i think, this add-on is killing me.
Today i replaced the cable that joins the drive to the controller, used proper twisted-pair servo cable (damn expensive) thinking it might be interference but no change at all.
I tried it at 2980rpm and it ran ok manually, i edited my code so the max CSS limit was 2980rpm and it passed three dry-cuts perfectly.
Feeling optimistic i put tool to metal and it actually made a part, tweak the tool offset to get size on, made another - OK.
Another tweak because i went the wrong way ::) and it made another part OK.
Run another and I get the ePid fault, it wasn't even running at max rpm i think, that is the point where i gave up and turned it all off, another 6 hours of messing and no result.
The parts are in the picture, apart from issues with stringy swarf, it worked really well which makes it even more annoying.
I was going to pull a drive from another axis but i have to pull the cabinet apart for that and i can't risk upsetting the main mill setup as i use it a lot.
At present it's half an inch or maybe 12mm away from the dustbin >:( >:( >:(
-
You mentioned earlier that the bearings were heating up, could it possibly be the increased friction of that is causing the motor to have a harder time keeping the spindle going and thus faulting out?
-
They get warm, i'm paranoid about things that run warm ;)
At the end of four runs of 4 minutes each, they were fairly hot but ok to hold by hand so i guess below 40c ?
I can easily flip the belt off and do air-cuts to prove it out.
A friend has just suggested it sounds like the controller (CSLabs) is having trouble counting pulses at full speed.
The other axes have the same drives/motors BUT one big difference is that they never run flat out because i have the mill tuned pretty low as far as rapids go, maybe only running 2000rpm max on X & Y drives.
I can pull the belt off the knee axis and connect that to this drive - see if it faults out, that will prove its the motor or not, if it still faults it points towards the CSLABS box i think ??
-
Can't really see it being the CSMIO not being able to count the pulses unless you have some astronomical pulses per rev.
On the Chiron my spindle encoder is set to 32,768 pulses per rev and I don't have an issue at 3800 rpm.
-
Ok, thanks Hood, thats the ENC module, i guess you have the same in the spindle drive encoder setup too?
I'm not hooked up to the ENC module yet, just the controller.
-
Yes, as mentioned previously I use the one encoder for both IP-A and Enc module, just jumpered across.
-
Ok, thanks.
I'm going to throw the 'scope on it again and watch what happens as speed increases.
Then I'll try swapping some motors.
I'm reading back through my massive build thread because this sounds very familiar from when i was first tuning my axis drives....
-
Right, before i go any further, i did some poking with the 'scope, my problem is i don't know if what i am seeing is an issue or not so thought it best to get this on the table first off.
four pics attached, 3 taken at the controller end of the cable, 100rpm, 1500rpm and 2800rpm
one pic taken at the drive with cable removed and the spindle just flicked by hand.
So, my questions:-
Do these show a problem?
Why is there so much 'crap' in the signal at the controller end of the cable when i am using proper encoder cable with the pairs twisted and used for each channel as required?
This interference appears at all encoder connections on the controller, I can't easily get data though as i'm on my own and holding the probes plus jogging etc is a bit dangerous.
Its not a ground loop as disconnecting the cable screen made no difference at all to the signal.
If its normal, i'll carry on with motor swapping etc but don't want to start pulling it apart for no reason
-
Don't look great to me but then I only have an old CRT scope and not a digital one, so maybe my signals would show up in a similar manner to yours if it were digital. Having said that when I used the PoScope I don't recall seeing bad signals like you have.
-
When you go to Plugin Control button and then choose the CSMIO do you see any red numbers for the Internal position counters. A friend did on one axis on his lathe, can't remember exactly what the cause was but I seem to recall it was bad encoder signals and the CSMIO was picking it up and showing the values in red which indicated errors.
-
Hmm, whats bothering me is the fact that merely connecting open cable caused the degraded signal - you can pull the connection from the controller and probe the open wire and the result is the same yet the drive seems to be sending good waveforms out form its connector.
I just really don't know if this a false lead or not yet but i don't like it simply because its there.
-
When you go to Plugin Control button and then choose the CSMIO do you see any red numbers for the Internal position counters. A friend did on one axis on his lathe, can't remember exactly what the cause was but I seem to recall it was bad encoder signals and the CSMIO was picking it up and showing the values in red which indicated errors.
Yes, always have done, sometimes get the "check encoder connections" warning at start up too.
It seems connected to the fact that my controller is powered up before the drives are - nothing i can do about that as the drives only have the one power option, not split power & control supplies.
The spindle encoder does not show up in that window though.
-
Its not the motor/encoder/spindle bearings
I swapped the motor for my X axis one and got same fault, the X Y Z motors never ever reach full speed so never see the issue.
Its not the spindle bearings/heat causing as it faults with belt removed.
-
Shame you are not closer as you could have tried out one of my drives/motors on your CSMIO to see if there was an issue.
-
Yeah, thanks for the offer.
Been looking at drives, I know these are cheap for a reason but surely they should work?
I still don't know if the signal quality is an issue.
It runs fine open-loop so its definitely encoder related.
I have no other drives to try out, if i was to buy another i would have to be 100% certain it would work first and it would need to be affordable i.e. probably 2nd hand.
I know this issue popped up on my first build, checked the thread but cant find it, i think it was when bench testing a drive and CSMIO - it would fault out but back then i put it down to lack of set up etc, since fitting on machine, none of the drives have ever reached full speed so the problem never returned.
Doesn't help but i'm sure it happened.
-
On the Chiron I initially had problems with the spindle drive faulting. Can't recall exactly in which way things faulted, whether it was the drive or the CSMIO. Ended up being the cable I had for the encoder to drive, made a new cable up and all was fine.
I don't really think it will be the drives or encoders as such but maybe just the signal output from the drives is not strong enough to push out to the control. Are they just straight through outputs or are they buffered?
-
Manual shows them as buffered outputs, AM26ls31 line driver.
I have tried two cables - my go-to tester which was a centronics printer cable, now i have made a new cable with proper encoder cable, twisted-pair, shielded etc, damn expensive stuff too.
The cable is less than 1m long.
I cant find any tech spec on the CSMIO inputs though other than max freq is 6Mhz
-
Don't suppose you have a spare encoder you could chuck up and use that for feedback to see if it is any better?
-
Not really, i do have a single-ended one somewhere, would have to order a line-driver to get differential output and its only 1024 pulse IIRC
-
Does your drive have a 5v and 0v connection for powering encoder? If it does you could maybe try connecting the 0v from the drive to the 0v on the encoder input of the CSMIO and see if that helps. Not sure if I have mine connected that way or not but I do remember that I had to use it on the lathe when I was using a CNC Building Blocks breakout board.
-
Great minds and all that...
I did that this morning after RTFM again - its a bit vague and one diagram shows no connection, another shows the 0v connection.
The are connected now but it made no difference. :(
-
Oh well, it was worth a try I suppose.
-
Its not encoder related as far as i can see :(
I had a spare 600ppr encoder so wired that in and connected it by clamping in the spindle.
It still suffers the same fault, just slower as there are only 2,400 pulses instead of 10,000 per rev - the error count just goes up slower.
It still worked ok until 2980rpm.
-
Using my bench PSU at 10v it was giving 2965rpm one way and 2971rpm the other.
It was also creeping with the inputs shorted out.
I ran the auto zero speed set function which killed most of the creep.
Then adjusted the gain to 305rpm/V
It now gives 3000rpm forwards and backwards.
Did it fix the issue?
No :(
A short video of the Pid screen at 2990rpm and 3000rpm
https://youtu.be/i7GEe7jAoY4
-
Using my bench PSU at 10v it was giving 2965rpm one way and 2971rpm the other.
It was also creeping with the inputs shorted out.
I ran the auto zero speed set function which killed most of the creep.
Then adjusted the gain to 305rpm/V
It now gives 3000rpm forwards and backwards.
Did it fix the issue?
No :(
A short video of the Pid screen at 2990rpm and 3000rpm
https://youtu.be/i7GEe7jAoY4
as mentoined in Reply#120
-
as mentoined in Reply#120
Of course, i'll butt out on this topic, apologies again.
-
I know I have stopped this thread due to it being against rules but was hoping to add one more entry for closure - might help someone in the future.
I think I have fixed it :)
Two parameters -
PN020 - Motor rated speed - this is set by default motor settings
PN051 - Motor Max Speed - this is a user adjustable one
PN051 Max was set at 3000 - i had tweaked this before but not the other one as it's a default setup one.
I have now set both to 3050rpm and away she goes :)
My working theory is that the controller was calling for 3000 but the drive was maxing out somewhere around 2990 - where it faulted so it was stuck in an endless fault loop - controller saying "I want more” and the drive saying “you can't have it”
Short video…
https://youtu.be/ksGlgcCWO4s
I think it makes sense??
No idea about tuning yet but that is not for the mach Forum, thanks for all your patience on this one guys :)
-
I think TPS actually meant post 119 rather than 120, he mentioned about 2950rpm being the max at 10v. nothing to do with post 120 so I wouldn't worry about that :D
Good to hear you have it running, it is always the things like that that are a headache, had similar nightmare with the VFD on the lathe, somehow it changed itself to be looking for a serial command connection rather than a voltage command and it took ages for me to work out why it wasn't working :D
-
Ah that's good to know, I know talking about other products is frowned upon so will keep it minimal ;)
Yes, oddly though it still faulted when i set the max rpm to 2950 but at least its happy now.