Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: BR549 on July 17, 2015, 11:44:33 PM
-
I was part of a group discussion the other day about WHAT is needed IF one was to build a new DIY controller for "TODAY'S" hobby users.
It was VERY surprising as to the feature needed not needed list. This was from a group that uses CNC every day as a living and a hobby. A mixed list of users.
As a conversation topic WHAT would your needs be and WHAT could you do WITHOUT as far as useable features. NOW be honest with yourself. I am not talking about wannabe features just real life I use it everyday features.
Popcorn is ready, (;-) TP
-
DIY cnc controller
LEAN and mean and super fast 3d
no turn
NO mouse needed good tab control
SIMPLE GUI no wood grain panelling (;-) One size fits all
4 axis only A synced feedrate on diameter
large toolpath area
Dry run
Simulate/estimate
3d probing XYZ creepy crawler built in
2d probing contour XY creepy crawler built in
Tool height probing built in
Table Probing for surface comp
A to Y function
G68 YES
analog out spindle, OR relays control KISS
Run HOT feeder
Run Laser
Run plasma ,sense arcOK for motion , NO THC control
RefHome switch
Limit switch
BASIC Standard Gcodes /modes
limited Std canned cycles Drill /tap only
GCode MODAL cycle (roll your own G66 style )
NO G10 stuff
NO tool offsets (tool Comp)
NO fixture offsets
No tool tables
No fixture tables
NO G52
NO G92
Multiline MDI with build in SAVE option (conversational Command line)
Standard Mcodes
Mcode scripting(roll your own)
Probing 4 ways
G31 inverse motion/ Trip then back off xyza
G31/g2 probe on arc
G31/g3 prob on arc
Safe mode motion for non G31 moves
NO #var
No SUbs
No conditional
No G65 macro
No soft limits
No wizards
(;-) TP
-
Honestly,
What one needs now / today will not be what they will need tomorrow / in the future.
So I would like everything available and then pick and choose for my use.
So maybe Mach installation could give installation option list where one could tick off what is wanted or would like to change
based on need over time.
So maybe there are a few levels of "Todays" Hobby user and based on some general questions ( below) one would get
an appropriate level of installation. The questions automatically imply what will be used / not used.
1. What machines will be used?
2. Have software to generate code?
3. Will you do any programing?
4. Is tool probing be required?
5. Is part probing required?
6. etc, etc
BTW, I am working on the Ultimate Hobby Screen which is truly PLUG and PLay.
Just has 3 buttons....Start, Run, Finish, requires no configuration or knowledge, no mouse or key strokes required,
as all is done visually. Just have some technology issues to solve and it will be done. ::)
RICH
-
Rich just stop and think about how it was done yesterday and how it is perceived today then it is fairly easy to predict what tommorrow briings. People are creatures of habit.
Take tool comp for example MOST people today have NO IDEA how to use it and don't as Cam does all the work. IF you don't have TC you do not need a tool table.
Fixture offsets are the same MOST have no idea why they need it as they never use it. Tkae that out and you do not need th fixture tables.
Probing absolutely need that as they WANT to replicate things and recreate them in CNC. PUT some muscle into that section of code.
#vars NOPE
Subs NOPE
Conditional NOPE
parametric NOPE
Macro scipting NOPE
Soft limits NOPE
120 ways to run a spindle NOPE
Now you are down to a simple set of features that you can spend time with to make a very FAST planner that can do 3d really well. and IF it can do 3d well it does 2.5d extra extra well.
YOU have to take YOU out off the picture as you are abnormal compared to teh average User TODAY. SAME as with myself .
Remember today cad/cam is a dime a dozen and some are free so you do NOT need any drawing or cam support in the package. Just partner up with a cad cam package as a mutual interest project.
I think it its time to step back and take a REAL look as to HOW todays users are actually wanting it to be like and HOW they actually use it.
I bet that it is NOTHING like you and I perceive it to be. The discussion was an eye opener.
Just a thought, (;-) TP
-
I have to admit this mirrors much of my thinking this past year. Were I to do a new controller it would have
no G42/G43 type offsetting, no tool tables unless the program did CAM . In the old days GCode was hard to
make, took a long time and the machine op had nothing to do with it. Want to use a different tool? great, use G43,
but these days, if I want to use a wear of .3mm or swap tools for a job, Id just repost the code with a wear on it,
or a smalller tool diam.
Hobby people make code and cut it ( Ive heard from a lot of them over time.). I never use tool offsets, why would I?
I start up vectric and repost my code, takes me a second, and I run the program with no concerns and I can read it easily.
GCode is darwinism in motion, its evolved to be glorious in its power, but buggy in its implementation and getting worse
in complexity over time. I say leave it to the Boeing type companies for which it makes sense. A contoller for a hobby guy
shoudl be much less encumbered with industrial type complexifications. Im firmly in the keep GCode simple camp.
Whats mentioned for mach4 , a screen with 3 buttons is probably a great idea for hobbiests. Im not against a controller
having full capability, so long as it doesnt impact performance, but I do think a hobbiest has no desire for the added
complexities and they actually hinder forward progress. One good thing about Mach4 is it is very configurable. Having
screen design makes it possible to be all things to all men. Hopefully over time that power will show.
Art
-
I would LOVE to test a stripped down version of M3 or M4. I bet 99% of all the weird would dissappear instantly and leave you with a very nice running package.
The screenset WOULD be very simple and easy to follow.
Basic Gcode is a VERY simple language. It is like talking to a 3 year old. UNTILL it is MADE complicated. AND it does not have to be that way.
To add to the list you do NOT need encoder support
NO modbus support
DO NEED pendant support from the hardware side.
-
Hi Terry, great topic, thanks for starting it, will be interesting to follow.
I like the idea.
"Hobby" today (and tomorrow) spans a pretty wide spectrum.
From brand newbies without a clue to model turbine builders, retirees from the industry .. real "Pro" hobbyists. Oh, and gear, watch and clock makers too. ;)
Keep it as simple as practical.
Everything in it works flawlessly.
Keep the price down accordingly (if possible).
Make it appealing to the "entry level" user.
I do not have a specific list of features, but I can say that I would've liked to have seen something like what you are proposing back when I jumped in.
Thanks TP,
Russ
-
HIYA Russ, one of the questions that came up in the discussion was . IF you only had a controller that did basic Gcode COULD you still make your parts as easy as you do now with teh support software that is currently available (Cad/Cam).
The answer was 99% YES. No problem, of course, sure. That was a jaw dropper to some.
SO what could you really do without ????? and still make your parts.
(;-) TP
-
I do use G52 on occasion, and do like the Soft Limits.
Other than that, your list above looks good.
Russ
-
The main reason that a lot of hobbyists don't use a lot of features is because nobody has ever taught them how to use them.
No offense to hobbyists, but a great number of them are doing things very inefficiently, and would be utterly amazed if they spent some time learning from someone that really knows what they're doing.
5 or 6 years ago, if you asked a question on a CNC forum, chances were very good that's you'd get a good, informative answer, and you'd learn something.
Today, the chances are very good that you'll get a lot of answers from fellow hobbyists that are either flat out wrong, or not the best way or most efficient method.
There's a lot of the "blind leading the blind" going on in hobby cnc today.
A group of some very skilled and well known users are getting together to try to rectify this situation, by possibly offering training and a new forum where questions will be answered by knowledgeable people.
What Art is describing is a control targeted strictly at hobbyists. There's probably a lot of money to be made in that market right now. Look at the success of small turnkey machines like the Nomad from Carbide3D. And the ShapeOko and X Carve.
But what happens when these people start learning and want to do more? You're catering exclusively to entry level users.
To some extent, Vectric does the same thing. While they are constantly adding new features, plenty of really good features have not been added, strictly because they don't want to add too much complexity for their hobby users. Ease of use has often trumped more power with them.
Sorry for the rant. Next post will be about the control. :)
-
But what happens when these people start learning and want to do more? You're catering exclusively to entry level users.
Then I would move up to a more full featured product. Might eventually require commercial equipment.
I think that in addition to entry level users, it would cater to hobby folks that do not require the additional features, the (probably) large percentage of folks that jump in, get their feet went then leave and those that desire to make simple projects ... and nothing more.
One might start a hauling company with a used pick-up truck and see how it goes. Might eventually by a semi or 2 but surely would not start out with a fleet of locomotives and 747's. :)
-
The simple answer? A bug free Mach3 with the Tempest planner.
Seriously. What features are lacking from Mach3? (If everything worked correctly)
The vast majority of users that are moving on from Mach3 are doing so because of bugs that they can't workaround.
OK, let's see.
I think the ability to customize screens is very important. Unfortunately, the people writing control software have show that they are not very good at GUI design.
And no matter how well thought out the GUI is, there;s a good chance that someone else can improve on it.
G41/G42 - To me, this is mandatory. Anyone that's ever run a "big iron" router knows that using G41/G42 is very common. In our shop, I have at least $500-$1000 in 1/2" router bits. After sharpening, none of them are 1/2" anymore. We often run hundreds of unique programs every day, and being able to program for a 1/2" bit for all of them saves an incredible amount of time.
What if I program for a .489" bit one day, and it get's dull before I can run my part? The new bit is .474". without G41/G42, I have to go back and redo all my programs. With G41/G42, just change the diameter in the tool table and start cutting.
Having worked this way fro 20 years, why would I want to go backwards
G43 - Imo, not needed unless you have an ATC, and even then you can work around it. Thousands of people are letting Mach3 automatically zero tools during toolchanges with my screenset and other macros with similar functionality. G43 might be a little faster with a lot of tool changes, but requires a method to have fixed length tooling, which most routers do not have.
Macro or scripting support - I could live without this if the features I'd use scripting for were native functions.
What I need is support for two independent Z axis, via a SwapAxis() type function
Auto Zero during tool changes.
Customizable homing.
User definable park positions.
If all these were native functions, then I wouldn't need scripting.
I don't use offsets very often but they are needed in some situations.
Actually, a bug free Mach3 with Tempest would be all that I'd really need.
-
On the screensets I feel different all l that is needed is a basic screenset that has all the basic functions of CNC. IF you need more than that you are using the WRONG software. IF it did all the basic functions of CNC what MORE would you REALLY need.
Most users to day just want to WATCH the machine make a part not spend days trying to figure out HOW to do it.
Keep HOBBY in the back of your mind IF I were having to make a living with it the SPECS would change dramatically . Been there done that and have the hat, teeshirt and coffecup.
Even M3 has a great deal of overkill for TODAYS average user.
(;-) TP
-
Probing is a neccesity You need it for tool height setting and replicating things. It needs
G31 straight probe with inverse motion. It runs to trip then backs off. That alone can save a ton of programing.
G31 /G2 probes in G2 arc as defined These are the cats meow for 3d probing routines of the creepy crawler methods both 3d and 2d contouring
G31/G3 probes in G3 as defined
Saved points ONLY on trip not end of motion. End of motion would be an error could write error 999.999 code in saved points
Protected mode motion other than G31
-
You're equating hobbyist with beginner or basic user. Unless that's exactly who you're targeting??
I would think that many hobbyists tend do be very serious or hardcore users.
Pick any hobby, and you can find multi million dollar industries that cater to the whims of these "hobbyists".
They don't want the basic entry level setup. Because they chose it as their hobby, they really like doing it, and will want the latest and greatest, with as many features as possible.
-
HIYA Gerry , I have found that WANTING features and actually USING features tend to be 2 very different things. It is like buying a new car Why buy a new Porche when a standard sedan will do the same job at a fraction of the cost and complexity of owning.
That would be apples vs oranges.
I have been setting down and actually taking things and creating and running code to test the theory. I don't NEED a great deal of the features in Mach3 just to create accurate parts. The is the main theme of the discussion WHAT do you actually need to make parts.
G00 yes
g01 yes
g02 yes
g03 yes
go4 yes
G10 NO
G12/13 NO
G40 -49 NO
G50 yes
G54-59 NO
G68 yes
G80 /81/83/71/84 yes
the rest of the canned cycles NO
M98/99 subs NO
G65/66 macros NO
#vars NO
Wizards NO (use your conversational cam )
Screens sets other than a basic function screen NO
MOST Other controllers are doing FINE with a preset functions SCreensets Most do NOT have a screenbuilder or even need one. All the basic functions are all ready on screen.
Sit down with a stock 1024 screenset and then strip out everything you do NOT need. You will have a LOT of room left over (;-) I know I did. AND could still cut parts out just fine. ADN when I mean strip out I mean strip out. NO background stuff JUST the basic DROs buttons and inputs.
IS it a different mind set YEP. But so is the mind set of the newest generation of DIY CNCers. Dorthy we ain't in Kansas no more .
(;-) TP
-
OOPS don't really need teh canned cycles at ALL CAM will do it all quite nicely.
That is my WANT sneaking back in there.
(;-) TP
-
Come on Guys I KNOW there are more than the 5 of us here. WHAT do YOU say ??
SPEAK UP
(;-) TP
-
I was part of a group discussion the other day about WHAT is needed IF one was to build a new DIY controller for "TODAY'S" hobby users.
I know I am a bit late here but I think the first thing to do is define Hobby users Vs. CNC hobbyist. IMO they are worlds apart.
Who would this controller be aimed at pleasing? Those using a CNC in their Hobby or those who's hobby is CNC?
-
You're equating hobbyist with beginner or basic user. Unless that's exactly who you're targeting??
This illustrates the 'spectrum" I mentioned earlier and I think you should target the lower end for this type of endeavor.
I would think that many hobbyists tend do be very serious or hardcore users.
I'd say many do as well, but probably most don't.
Pick any hobby, and you can find multi million dollar industries that cater to the whims of these "hobbyists".
They don't want the basic entry level setup. Because they chose it as their hobby, they really like doing it, and will want the latest and greatest, with as many features as possible.
Some certainly do, many also use full featured hobby software to retrofit "Big Iron, Thousand files a day" commercial operations.
It's all in how you choose to define Hobby .... and where to divide the spectrum.
One size WON'T fit all.
There are products for the upper end, this would be great for the lower end.
IMO, mid to lower end, :)
Russ
-
I was part of a group discussion the other day about WHAT is needed IF one was to build a new DIY controller for "TODAY'S" hobby users.
I know I am a bit late here but I think the first thing to do is define Hobby users Vs. CNC hobbyist. IMO they are worlds apart.
Who would this controller be aimed at pleasing? Those using a CNC in their Hobby or those who's hobby is CNC?
Great points Brett.
Yet another division in the spectrum(s).
Russ
-
First you need to figure out WHAT you can make with your upper end controller that I cannot make with my lower end controller (;-) I have been working at that for over a week now. The answer has been very surprising so far EVEN for a mid range commercial operation. The group fed in some surprises as well Several reveiled that it would surprise everyone(and it did) as to HOW FEW of the advanced CNC features they actually use Day to Day. And they had been at it for over 30 years and still going AND have no plans to change the strategy as it makes GOOD money.
Now would that work for EVERY situation ? NO but for the vast majority it would.
How important do you think G17/18/19 is ??
(;-) TP
-
Hobby CNC Definition - SMALL LEVEL
1. No knowledge of Gcode at all and could care less about it.
The only thing needed to learn is to be able to move the axes in a straight line
at some rate in permissable directions.The gcode is totaly transparent to the user.
2. No knowledge of basic machining which includes setup of material, all the associated
feedrate stuff,etc and does not want to have to tell the machine anything.
Will only purchase a generic mill tool set ( small, medium , and large set) and no
machinist tools.
3. No desire to create a screen, do programing, etc.
4. No knowledge of controller configuration and could care less.
So simply said, the user just wants to put some stock in the machine, press START
( to start the process) RUN ( to let the machine do the program) and Finish ( to shut
everything down when done).Start just prompts you to insert a tool, define what
machine you are using, and then you just move near the material, close the machine
door, and select the program to be done. Run just starts the machining.
So the screen set is just one page and has almost none of the typical options.
Just DRO's to show axis movement "and they are only numbers", a graphics area, and
there are no additional tabs, like disagnostics etc. All you have are some jog buttons
using only the keyboard to move the axis. New language required so get rid of all
those goofy terms and definitions.
SO...........
For the user the only thing they can do via the screen is move the axis
up/down/left and right or click a button to find the material.
So the burden is shifted from the controller interface to the 3D CAD/CAM.
Yes you will need advanced "finding" to automate things.
So the "Package Deal" may be a dedicated computer to a dedicated contoller system
and supplied with 3D CAD/CAM system.Heck throw in a dedicated machine.The difference
between "levels" of hobby are just how much control and what options are provided to
the user for the "system" based on what will be done.
The new generation lives in a visual world.
WYSIWYG....WHAT YOU SEE IS WHAT YOU GET! ......remember that saying for CAD.
To summarize it all.................
There is no setup, configuration, or gcode required to speak of.
User wants more control to do something or control the machine then buy the next level of
package!. The levels comes in small, medium and large.
Maybe also need the supersize for the industrial guys.
Removed myself from the picture, ;) :D
RICH
-
RIch that is funny as you just decribed the growth path of PRINTERS/Plotters (;-) in the OLD days to plot you had to load the paper, program the plotter as to WHERE the paper was located THEN define the size of the paper and WHERE the plotter was to begin the actual plot. Then DEFINE which PENS did what. ALSO the OP needed to learn all the ERROR codes involved so he could fix the problem to get it to actually plot.
SOUND FAMILAR ??
Plotting was more a Black art than a science .
Today you simply fill up the tray or Roll . The printer KNOWS what size paper is in the machine, can find the corners all by itself KNOWS where to pint it to and KNOWS the all the pen color choices. The op just creates a drawing and presses a button grabs a cup of JOE and the MACHINE does all the rest. THe Print Driver does ALL the hard work and teh CAD produces ALL the need code to draw the shape. It will EVEN email/text you when it is done.
Does THAT sound familar ???
CNC is at that point where it needs to be more like a modern plotter/printer. In the DIY world ALL the pieces to make it work that way have already been developed they just need to be assembled into one Controller. SOme are clser than others to where they are in that quest.
The OPS just want to USE it not earn a Doctorate Degree in Machine design/programing.
BUT to SOME the doctorate degree IS the HOBBY not the use of the machine.
NOW back to the question, WHAT do you actually USE feature wise to do your work ??
(;-) TP
-
Hey Terry,
For the Mill it's easy, as I just put the tool into the collet, make straight moves to where I want it begin ( I know that spot since I defined it in CAD and CAM).
then open the file and let Mach run the program.
I don't have any switches, homing, motor control, tool changer, no need to even open the tabs, etc, etc. All that stuff is just there should I need it for some reason.
All that other stuff is on the screen to give me something to look at while and keep me from failing got to sleep.
All the wizards are not really needed as the CAD / Cam does all the work generating pre compensated code. The configuration is probably never changed. The machining may be piece meal work, but, usually the work is one of a kind.
That's as bare bones as one can get and has served me well for years.
The only thing that comes to mind of particular use are the MPG's and they are used to go fast or go slow.
Same could be said for the engraving machine.
Now make no mistake about the following. I design my work, not just draw it, and put good machining and engineering practices into the
design. I also like programs that give me a lot of control over what I define for the machining, yet there are those that say they would never use
that program because it needs input! Yet they will use a whole bunch of wizards to do something.
When the machine shop that is making the parts is just a telephone call or 5 minute walk from you are you quickly get feedback
on your design and sometimes it's not what can't be repeated in public.
--------------------------------------------------------
Now let me inject some ridicule into the CAD / CAM side and it is true in the industry.
There are those that can generate the most complex things in CAD , BUT, they couldn't design something if their life depended on it, and it is getting even worst
today. Ever wonder why some companies won't even employ someone as a designer unless they have a 4 year degree.
RICH
-
BTW,
A little history.
It's around 1970 and 8 hours of CAD is mandated in the course program. The program was to give vision of how things would be in the future.
All that could be done with CAD was make the curser go across the screen!
We have come a long way baby!
Fast forward to about 1980. AutoCAD is the equivalent of "etch a sketch" , the Bentely brothers are employed by a company which decides to make a better
in house CAD program. We just ask them do create a command and 4 hours later it's done and being used, one of my requests was cut, copy, move and paste selected items from the drawing. ...it took 8 hours since the initial version had a few bugs that had to be worked out. That CAD station cost a 1/4 million!
Talk about working with the best in industry and cheaply priced software today!
OK back to topic .............,
Sorry one more....
The niece noticed my slide rule and asked what it was. I said it was my old computer. She said that they sure did make computers small back then and that it didn't have a
any batteries. Yep, I replied , they were the good old days! :D
RICH
-
Rich the sad part of that story IS that the degree only means you showed up to class, passed a test AND paid your tuition on time. It does NOT mean that you could even design a toothpick. The world has seriously dumbed down a lot. BUT the machines HAVE gotten a LOT smarter (;-) We are Borg,You will be assimilated, resistance is futile.
I think IF one were to design a new controller it would NOT need all the advanced bells and whistles that ADD huge complexities to the Source code and slow the running of the control down for things like 3d work. I have proven to myself that there is really not much that I cannot create with basic Gcode calls. AND never use ANY advanced Gcode calls. AND do it just as fast OR faster .
That is BASICALLY what the 3d printer dudes did and it sure seems to work ok for them as they tend to process a Megaton of micro segmented lines /points into smooth flowing shapes. And can do it all with a BeagleBone(;-)
THe early CAD I used was just a command line input ONLY and it all came on 1 floppy disc. But then again I remember paper tape nc machines AND it would surprise you as to how MANY extreme torerance things are still created from a TAPE driven CNC machine only now the paper tape is Kevlar tape. But it still uses the same old hole pattern technology as the OLD NC machines did.
(;-) TP
-
for me I like to use the fancy bits and bobs other than tool table stuff my cad does that, all types of probing always use it, laser zero for x and y, soft limits don't use it on big machine, small I use soft and hard limits and MPG.
for my dad tool height probing, laser zero for x and y,cycle start, feed hold, stop, E stop( x, y and z work pieces max and min dro) and that`s all he just wont`s put ya code in chuck the material on the table zero it and start cutting all done with his MPG when M4 is working better his screen set is one page x,y,z dro with ref home and zero, G code and tool path with a couple of buttons why I do the fancy stuff so he can just cut
-
OK here is a thought,
We create a new verison of MACH4 called Mach4 EASY. It is a simple version of Mach4 with basic Gcode features. It uses a simple USB control box that DOES NOT require any setup from teh user. It is like a new USB printer just plug it in. The Control box has terminals for the machine setup AND a diagram as to where everything goes.
4axis
home/limits
probe
spindle control
It is ALL predefined .
Mach4 EASY would be a basic 1 screen control AND have an integrated version of MILL wizard AND you can add moduals for things that are developed for the USERS that want something special. It would have a Modual importer to load the addons. Plug N Play.
For the DIY group You BUY Mach4 EASY and it comes WITH the USB card Ready to plug n play OR you buy a new machine with it already installed ready to cut.
Later on IF you think you need to be a CNC Master you could UPGRADE Mach4 to the Commercial (Mach4 C )or Industrial ( Mach4 I ) version.
MOST would never need to upgrade(;-) But you do have that option.
Simple and EASY , (;-) TP
-
It was supposed to have been released over a year ago, and called Mach4 Lite.
If they don't do something like this, then new users will be flocking to the alternatives. Unless they come up with a way to setup Mach4 that's considerably easier than it currently is.
-
That rings a bell Gerry.
At one time it was Lite, Pro and Expert versions planned.
http://www.machsupport.com/forum/index.php/topic,21703.msg150899.html#msg150899
Starting to notice some of the flocks gathering too.
Thanks,
Russ
-
I think it was 2 years ago, that Automation Technologies (Keling) was selling boards with Mach4 Lite as a pre-release, and it was supposed to be released around October 2013. After that, it seems to have vanished.
I can't see how Mach4 Lite could have been ready 2 years ago, with the current state of Mach4 hobby 2 years later.
-
This is a simple screen I made a few years back. As you can see the operator didn't have much he could screw up.
No menus
No tabs
No other screens
Ray
-
Just to put in my 2 cents. We recently bought a Servo 5000 mill vintage 1987. The owner never used it because manual G-code programming was too hard for them. So the machine sat until 2008 when they were offered a Windows 'upgrade' which they bought, thinking it would be easier. They couldn't even get the Windows computer to talk to the drive boards so the machine sat again. So we picked it up a couple of months ago expecting that while it had maybe 100 hrs of spindle time it might need a new CNC. However when we plugged everything in it ran perfectly. It is 20x40x21.5 travels includes a 10" fourth axis with an 8" 6 jaw chuck and 21 Cat40 tool holders with 7 Jacobs Super Chucks, for $6500!
Of course programming it manually hasn't gotten any easier. However its a Windows machine and for $150 we installed CamBam right on it. So all the BS you needed to do to program manually is mostly gone, and we make parts after a couple on minutes of programming. So the reality is that the CNC itself can be really really dumb, while the CAM program with however much horsepower you need can now be very smart. So the need for canned cycles, and lots of other stuff moves off the CNC and onto the CAM computer. To me that means that the motion controller boards like smoothstepper become the CNC, and I am not really sure that Mach 3 or Mach 4 needs to be much more than the HMI for the iron, which is probably a better task for my cell phone or an Ipad.
-
>> I am not really sure that Mach 3 or Mach 4 needs to be much more than the HMI for the iron
I suspect you underestimate just how much M3 is doing and overestimate how much the smoothstepper is doing. The smoothstepper is a 6 channel organ, m3
is the orchestra, the meistro and the music interpreter. There is far more to it behind the scenes than most realize, even if you stripped out the macros
and many other things. Not that it might not get there someday, but controllers still perform a very important function in the process.... :)
Art
-
Hmm..........
The newbie CNC "Shark's" ask the simple question in consideration of investing.....
So what's the product that you will be selling?
We are confused as we think the base product, still in development, lacks third party application for which you have no control over.
It must work with all the outside influences. I don't want to buy a car and wonder if the rims and tires I have will fit.
What's the investment cost today? When will I be able to upgrade?
Commercial break...............
So what do you think the newbie's said after the break?
Just a different thought process to consider,
RICH
-
Maybe I should have posted reply #35 here after reading the last comments in this thread,
http://www.machsupport.com/forum/index.php/topic,30530.msg212005.html#msg212005
But the reply still challenges / has implications to the basis of this thread,
RICH
-
My ideal controller would be able to do all that Linuxcnc does with a built in screen designer and perhaps a more helpful way to setup the machine requirements
-
>> I am not really sure that Mach 3 or Mach 4 needs to be much more than the HMI for the iron
I suspect you underestimate just how much M3 is doing and overestimate how much the smoothstepper is doing. The smoothstepper is a 6 channel organ, m3
is the orchestra, the meistro and the music interpreter. There is far more to it behind the scenes than most realize, even if you stripped out the macros
and many other things. Not that it might not get there someday, but controllers still perform a very important function in the process.... :)
Art,
You are probably correct, I really don't know how much is done by a smoothstepper or even what you feed it. It seems like when unload the realtime motion control to the smoothstepper that most of the work left would be user interface and I/O but as you say that may be a gross misinterpretation.
Art
-
Gary:
I guess I forget that most of you dont really know whats under the bonnet. Smoothsteppers, printer ports, most controllers are fed motion in waypoints. The printer port was fed 200uS points, the rest are almost all 1ms waypoints. So the controller, m3, m4, linux..whatever, take in the Gcode, they hold the state engine which is the guts of cnc. They interpret the Gcode , control the various states, then plan the trajectory to 1ms waypoints.
Motion engines are told whats up for the next 2 seconds or so, the controller makes sure this 2 seconds is extended as it depletes. The controller
takes care of acceleration, mixing speeds between axis, maximum and mixed velocities , offsetting, and toolpath displays. Scripting and configuration as well.
Engines take the waypoint data, and are programmed to take a step count for each millisecond of motion, how many steps each axis takes in that ms
determines end velocity and such. So jerk constraints,acceleration curves and constant velocity and so on are computed by the millisend and continuously sent to the device. Its all a complex dance of internal motion before the motion engine gets sent its data. How smooth the motion is, how complex the logic and scripting can be, thats all controller, the actual steps are made by the engine.
Im not saying the controller is more important than the engine, there is a certain chicken and egg flavour there.. I just wanted to clear up
that the engine itself doesnt even come close as yet to replacing the controller. Its been tried many times to do so, to have usb sticks with the programs on them plugged into a board and away you go.. .. its always failed, for various reasons. That day IS coming, but it isnt here quite yet..
Over the years Ive been told many times by people that hardware will replace the controller, the ncPod, the "Brain", and a few others had tried while I was still making Mach3. They all had truly planned to do so.. but its only when you try to actually write a controller you find just how vastly complicated
it truly is. Mach4 was supposed to be ready years ago, and I suspect theres a few humbled ego's from the experience of bringing it to where it is today.
Mach3 is far from perfect, but replacing it has proven a hurculean task.. which when you consider its age..is a real testement to its abilities, and the large group of contributers to its growth.
Art
Art
-
Art,
Yes I was unaware that all the operations you cite was still in Mach 3. I used to build a low class robot for the commercial greenhouse industry. It used a 6809 microprocessor at 2 Mhz, 32K eprom, and 64K ram. All the programming was done in Assembler and burned to the eprom. It also had a speech synthesizer and PWM for the motor controller so there was a lot going on. When my programmer got hired away after 4 years we turned the programming job over to the company making the boards. They were shocked when they printed out the Assembler code and it was over 40,000 lines! Fortunately he had done a great job thoroughly documenting nearly every line of code and they had no problem working with it. Production on that control system went on for about 15 years and ended when some of the critical chips were no longer available even on the surplus market. At the assembler level you can do an amazing amount of work with very few resources. Extremely time consuming though.
-
For a plasma cutting machine, you need forward and reverse on contour, you need jog off contour and either relocate or return to contour, and you need a kerf compensation routine that works, mach3 doesn't.
-
Fascinating thread, even if old, but I will vehemently DISagree with some of the suggestions.
NO #var
No SUbs
No parametric programming
Any of these would make it NO DEAL for me.
No wizards
Why??? They can be useful. Imagine having to replace the Text Writer!
Conditionals would be nice, but not if they brought in a lot of bugs.
G65 ... a bit of a can of worms I think, although the basic idea is interesting. But the idea that G65 would stomp on some #var variables - VERY debatable. OK, you would just have to skip using the first 30 #vars in any program.
The whole idea that any small changes in dimensions or cutter sizes should be dealt with by going back through the entire CAD/CAM/Postprocess/check cycle is ridiculous - especially as you can deal with any of these changes by changing a single #var parameter (or even by cutter compensation). Current CAD/CAM is just too mindless to be useful for many of us. Until you get to the $50k - $100k level.
The idea that hobbyists don't need anything more than Start/Run/Stop is amusing but also commercially stupid and very insulting. If someone becomes serious about their hobby, which is normal, they are going to want to do a LOT more than that.
Rhubarb rhubarb rhubarb.
Cheers
Roger