- Thread starter northwest
- Start date
So LinuxCNC uses raw C++ for its scripting purposes? Nice if so
Dumb clanky question then, why ? what advantages would that give? Bearing in mind, I know absolutely zero about programming, is it possible to adapt to suit?
- Messages
- 6,158
- Location
- SW Scotland
That’s my “daily driver” language so to speak. Advantage to me personally is that I use the language a lot. If it was something like, I dunno Python, then I’d have to look up syntax and APIs more often which is boring and slows you down.
Rough analogy would be if you work on Cast Fe MMA all day long and you have to do a bit of Al MIG. You can do it but it’s not what you’re tuned in to right now IYSWIM?
mtt.tr
Member
- Messages
- 8,444
- Location
- Essex/Suffolk border
It uses its own language, its no benefit unless you enjoy and more importantly as @Hood stated have time to prat around or want/need to achieve something super unique very unlikely
Fortunately i do have the time and i love the puzzle that programming produces.
Unless you are after that challenge i would just go professionally made setup
mtt.tr
Member
- Messages
- 8,444
- Location
- Essex/Suffolk border
I would say if you have that mindset you can learn how to alter the HAL files but you will be looking for syntax
see what you mean,; so if you were really up to speed you might be able to adapt to suit a specific issue, or improve the interface.? could be handy, but I'd guess at that level you'd need to be really tuned into it both the cnc part and the C+++.... well above my head then, that said, there's a lot of things in mach3 that are a bit clunky, and if I had the means / method to adapt it would be useful. Mind you if I need to go to a programming level, probably better for me to use off the shelf, I already have enough problems with that, don't need to go further into why or how it does it to just use it.
cheers
Hood
I am obsessed.
- Messages
- 17,624
- Location
- Carnoustie, Scotland
Does Linux have a means of testing your macro code in a similar way that you can in Mach via stepping through the editor? I would imagine it does, the Chinese controller however doesn't and was one of the things I found to be a pain. I am not a programmer and only want to learn enough to get it doing what I want and thus need to keep trying different things/syntax etc to get it working. Being able to do in the editor in Mach is great as you can step through and hover over registers or I/O etc and see what they are as you step through.
The other advantage Mach has for me is it uses a form of basic and although it can be limiting it is also fairly easy for a non programmer to get into and understand. The Chinese control is likewise, it is a home brew language but fairly logical and easy enough to do, problem is you need to keep uploading to the control to test and if not correct then alter again on PC then upload again.
The other advantage Mach has for me is it uses a form of basic and although it can be limiting it is also fairly easy for a non programmer to get into and understand. The Chinese control is likewise, it is a home brew language but fairly logical and easy enough to do, problem is you need to keep uploading to the control to test and if not correct then alter again on PC then upload again.
Search to see what other people have used. Early interacts used big stepper motors, with later ones moving onto 1KW DC servos.
However bridgeport size, you'd be looking at Nema34 steppers (ideally with high voltage drives to maximise performance), or probably 750-1000W servos. Personally, unless you're really going for as cheap as possible and aren't bothered about performance, I'd go servo.
To give an idea of prices, looking at Lichuan - Lichuan CNC Servo Factory Store - Amazing prodcuts with exclusive discounts on AliExpress
2phase 10Nm Nema 34 stepper with 80VDC drive is $91 - 10nm cnc stepper motor driver manufacturer MC856 4A nema34 stepper motor kit 2phase LC86HS2801|stepper motor kit|nema34 steppernema34 stepper motor - AliExpress
3phase 6Nm Nema 34 with mains voltage drive is $142 - 3 phase stepper motion controller LC3722A nema34 stepper driver 6nm 3.0A cnc stepping motor LC3913 H|cnc stepping motor|step motornema34 stepper - AliExpress
Most common steppers are 2 phase, but 3 phase steppers are smoother in operation. You could also use a 2 phase Nema34 with high voltage drives (leadshine do them). The issue with stepper motors, is the rated torque is only available at stall speed. As speed increases, the torque drops off, and high voltage drives reduce the level of torque drop off.
Then there are the closed loop stepper options.
2 phase 10Nm with 70VAC/100VDC drive is $116 - 2 phase 10Nm hybrid stepper servo kit 1416Oz in Nema 34 closed loop step motor with 5m encoder cable for lathe polisher grinder|step motor|nema 34hybrid stepper - AliExpress
3 phase 9Nm wiuth mains voltage drive is $201 - 9Nm 3 phase Nema 34 closed loop stepper servo motor driver kit AC220V digital easy servo driver LCDA2260H and motor LC86H3129|nema 34|nema 34 closed loopclosed loop - AliExpress
But all those options will suffer from torque reduction, which limits maximum speed.
Moving onto servos-
750W 90 frame size servo $212 - Lichuan 90st M02430 for 750W servo motor flange 90 size with servo driver kit 3000rpm 2.4Nm AC servo drive ac servo motor kit|AC Motor| - AliExpress
1.2Kw 110 frame size servo $280 - AC motor 220v for 1.2KW servo motor with driver kit 110ST M04030B flange110 4Nm servomotor 3000rpm use in CNC plasma cut machine|AC Motor| - AliExpress
I've listed two sizes of servos, as the bigger one may be a better match for a heavier machine.
In terms of pros and cons of all the options.
The stepper options are easier to fit, and connect. They all use screw terminals for all the connections.
The servo drives use D-sub connectors for signal/control wiring, so you either have to solder them (the drives come with empty shells), or spend extra for suitable breakout boards.
Mains drives you just need to connect mains power. The lower voltage options will either require either a transformer (if they take AC input), or a suitable DC PSU.
The steppers you'll likely get better performance by going for the larger torque options, and direct coupling them. Milling machines don't really need maximum speed, but if this was something like a high speed router, I'd say go for smaller steppers and gear them, as you'll get better acceleration/performance that way - bigger isn't always better for stepper motor performance!
Servos, I'd run via a 2:1 belt reduction. The servos will spin far faster, and maintain full torque over their rated speed range, so to get most performance out them, you need to gear them.
The reason I picked Lichuan for the above examples, is I've dealt with them, and they were very helpful, although I actually ordered via Alibaba, as I needed a couple custom options.
I also know somebody who now mostly uses their products for building machines, as he can't fault their prices, products, or service. If you want to see manuals, their website is Shenzhen Xinli Chuan Electric Co., Ltd. (xlichuan.com)
Hood
I am obsessed.
- Messages
- 17,624
- Location
- Carnoustie, Scotland
That may be possible, depends. What are the "clunky " things?
usually me......lol
Think a lot of my bugbears stem from going back and fore to XP from 7 for the fusion software then back to XP for the mach... Mostly layout of the screen, spend a lot of time moving between them with diagnostics etc at set up time, but once it's set up, I used them less and less, so live with it.. I think it's more for me to adapt to mach than the opposite, just old habits from other softwares, for example multi screen layouts would be ideal, so can check ongoing functions at the same time especially at the beginning when I was trying to watch everything at the same time to get more familiar with the functions.
Think a lot of my bugbears stem from going back and fore to XP from 7 for the fusion software then back to XP for the mach... Mostly layout of the screen, spend a lot of time moving between them with diagnostics etc at set up time, but once it's set up, I used them less and less, so live with it.. I think it's more for me to adapt to mach than the opposite, just old habits from other softwares, for example multi screen layouts would be ideal, so can check ongoing functions at the same time especially at the beginning when I was trying to watch everything at the same time to get more familiar with the functions.
In terms of controllers, I've never been a fan of the LinuxCNC approach.
They were hell bent that unless it's realtime, it's not worth it, which means until very recently, they've relied on dedicated internal PC hardware.
As somebody who runs my machines to make money, I know the PC is likely the weakest link. If a PC fails, I don't want to have to faff around swapping hardware in the hope whatever has failed hasn't wiped out the controller cards. I just want to connect another PC, load the software/machine profile, and get running again.
I'd also echo @Hood 's comment about dedicated controllers.
I quite like the dedicated controller my new lathe came with, but I'm not sure I'd like the same on a milling machine, unless it was a production machine, where it would have a limited number of setups. PC based controls still give you a lot of flexibility, and often nice graphical displays for the lesser used functions. I know a dedicated controller can do edge/centre finding with a touch probe, but if you only do it every few weeks will you be able to remember the required code to do it?
My personal preference for controllers is the Dynomotion KFlop with the various optional boards, as it lets me do quite a lot things you'd struggle to do with other controllers.
Things like my little CMM runs standard steppers on all three axes, but two of which run closed loop via linear scales and returns the encoder position for probing points.
My last lathe ran a KFlop with analogue servos, a customised spindle output to handle the spindle VFD, and talked to a PLC via MODbus for the turret control. The PC is simply there as a user interface, and to send code when required.
My current mill runs three axis step/dir, with the spindle running an analogue servo and capable of rigid tapping.
The biggest problem with a KFlop is you need to understand C in order to configure them, but understand that, and it'll do things you'd normally need a far more expensive controller to do.
Every controller has pros and cons.
Anybody who tells you differently is either lying, has a vested interest, or thinks how they done something is the only way to do said something.
Mach3 was revolutionary for it's time. It still has a huge user base, big community support, and lots of compatible controllers. For new users, it is still a good option.
Mach4 is still good software, but it's now pretty expensive, and you still need hardware to go with it, which makes it even more expensive compared to other options.
Centroid Acorn is a good option, if the hardware meets your requirements, but by the time you buy the extra software options, it can get pricey.
UCCNC controllers are also a good option, again if the hardware meets your requirements.
As I said in a previous post, most controllers will manage to run a basic 3-axis machine. Where the cheaper options will often become unsuitable, is when you start to do more complicated things that require critical timing, or lots of inputs/outputs.
They were hell bent that unless it's realtime, it's not worth it, which means until very recently, they've relied on dedicated internal PC hardware.
As somebody who runs my machines to make money, I know the PC is likely the weakest link. If a PC fails, I don't want to have to faff around swapping hardware in the hope whatever has failed hasn't wiped out the controller cards. I just want to connect another PC, load the software/machine profile, and get running again.
I'd also echo @Hood 's comment about dedicated controllers.
I quite like the dedicated controller my new lathe came with, but I'm not sure I'd like the same on a milling machine, unless it was a production machine, where it would have a limited number of setups. PC based controls still give you a lot of flexibility, and often nice graphical displays for the lesser used functions. I know a dedicated controller can do edge/centre finding with a touch probe, but if you only do it every few weeks will you be able to remember the required code to do it?
My personal preference for controllers is the Dynomotion KFlop with the various optional boards, as it lets me do quite a lot things you'd struggle to do with other controllers.
Things like my little CMM runs standard steppers on all three axes, but two of which run closed loop via linear scales and returns the encoder position for probing points.
My last lathe ran a KFlop with analogue servos, a customised spindle output to handle the spindle VFD, and talked to a PLC via MODbus for the turret control. The PC is simply there as a user interface, and to send code when required.
My current mill runs three axis step/dir, with the spindle running an analogue servo and capable of rigid tapping.
The biggest problem with a KFlop is you need to understand C in order to configure them, but understand that, and it'll do things you'd normally need a far more expensive controller to do.
Every controller has pros and cons.
Anybody who tells you differently is either lying, has a vested interest, or thinks how they done something is the only way to do said something.
Mach3 was revolutionary for it's time. It still has a huge user base, big community support, and lots of compatible controllers. For new users, it is still a good option.
Mach4 is still good software, but it's now pretty expensive, and you still need hardware to go with it, which makes it even more expensive compared to other options.
Centroid Acorn is a good option, if the hardware meets your requirements, but by the time you buy the extra software options, it can get pricey.
UCCNC controllers are also a good option, again if the hardware meets your requirements.
As I said in a previous post, most controllers will manage to run a basic 3-axis machine. Where the cheaper options will often become unsuitable, is when you start to do more complicated things that require critical timing, or lots of inputs/outputs.
Hood
I am obsessed.
- Messages
- 17,624
- Location
- Carnoustie, Scotland
Why XP for Mach?
Regarding screens, you can make your own up easily to suit exactly how you work, here is mine, lots don't like it visually but I do and it has all I use day to day on the one page and the other pages I have more uncommon things.
Chiron main screen. Ignore the couple of out of place LEDs and DROs, they are just things I added for testing out things for other people and are only on the screenset on my home computer and not the actual machine one.
Tool setting page on Chiron.
Tool table page that displays the tools I have loaded and set up ready to go and which tool changer position they should be in, this gets updated from the tool table in Mach when you change or rename tools etc.
The Plasma is similar except I really only ever use the main page which is this.
mtt.tr
Member
- Messages
- 8,444
- Location
- Essex/Suffolk border
Those are some seriously good prices for servos even with the postal cost and the duty they will no doubt incur. I don't have any need for servos or drives at the moment but I will be keeping a link to them
Im envious i can never find anything on AliExpress
Thanks for that really in depth answer - before I buy anything, and I'm a long way from there yet, I will consider carefully. The performance thing - I guess, as long as the motor is strong enough, it's speed we're talking about. Well this project will be hobby only, maybe supply a few request items, so speed is not of the essence for me really.
One more question, 3 axes I see plenty of talk about, could a rotary table or dividing head be added, if it were needed?
Hood
I am obsessed.
- Messages
- 17,624
- Location
- Carnoustie, Scotland
Lots of people say that but often they soon regret going that route as it soon gets monotonous with slow positioning.
Depends on the control you go with, hardware especially can limit things. For example I have a CSMIO/IP-A controller on the Chiron and an IP-S on the plasma, both are capable of 6 axis. The CSMIO/IP-M that the OP is putting on his mill is only capable of 4 axis.
