What's new
What's new

Is it worth it to learn how to manually program or CAD/CAM?

Higgins909

Aluminum
Joined
Nov 19, 2018
I'm trying to figure out what I should learn first. With CAD/CAM it's so complex that it would be very difficult to edit manually. Programming manually would be useful for simple parts. I'm talking about programming on a Haas with G/M Code or learning Fusion 360 as I've had a hobby license for a while, jut never actually got into learning it. I bought some books for Fusion 360 almost a month ago hoping to teach me the basics but haven't even opened them. (I have a shelf of books that I do this with) Shop currently doesn't have a CAD/CAM but I think they're working on it. Even if it's a different program, I will have some knowledge. Although for all I know if could be in a week or they're just looking into it.

Thanks,
Higgins909
 
I'd take the time to learn CAM. With your tool library in there, you can set up templates to cover a lot of the repetitive work and program simple parts in no time at all, with the ability to simulate the toolpaths and be confident you're not going to crash the machine (especially if you import your workholding too, you can collision check with it). You'll also be capable of programming parts you'd never be able to hand code. Minor dimensional tweaks are easy at the machine if you turn on cutter comp in your CAM, you can tweak wear offsets as usual if you'd like.

That said, if you're running a machine at all, even from CAM, it's CRITICAL to understand what the gcode is doing for troubleshooting purposes (cus stuff happens), for writing a quick command in MDI (Zip to the current WCS XY zero to sanity check something, manually activate something during setup, etc), or for making/troubleshooting postprocessors if you have an oddball machine or want to do fancier stuff than is included with the stock post.
 
It depends on what you are trying to accopmplish and what your future looks like.

I learned G-code manually inputting it at the control. I now feel confident walking up to most any Machine and making chips, on relatively simple parts (and some not so much). I think it's integral to understand what the code is telling the Machine. To be able to stop it and look through the code to see what is coming next, to problem solve starts with Code, second is Mechanics. BUT in some shops it isn't needed. You have to know what environment you are working in. IE if you are the only guy there that needs to make parts and there isn't a second set of eyes, you probably want to be as independently capable as possible. Meaning not requiring outside assistance including CAD/CAM to make chips.

OTOH, I learned that way and now I use Gibbs, Mastercam and Esprit. I can edit my own post processors to output the code that I want, not the code that someone else thought was best for my application. The only reason/way I know what code I want, is because I am well versed in G-code programming.

CAD/CAM is not the be all end all of everything. It is a very powerful tool. Very few CAD/CAM programs can post to ALL the Machinery there is. There are blank spots. But that's all my opinion and you get what you paid for.

R
 
I would have never said this 20 years ago and I kinda can't believe I'm saying it today but cam is what's getting all the nice 3D surfacing work done today. The better the cam package the more easy and useful it be, we have Gibbs, Catia and Unigraphics. That stuff is big money programs and the chances are good you'll have to use whatever is provided so learn the basics.

We have some folks doing some awfully nice work that can't read G code worth jack shit. As a matter of fact if the network is down and Gibbs and models are not accessible might as well hand them a broom.

Yes reading G code is a plus but fact of the matter is I'm amazed at some of the work these guys do but know almost nothing about programming manually.

From an old school G code guy that was difficult for me to admit.

Brent
 
Yeah I get that. I didn't mean to give the impression that knowing how to program manually isn't a benefit, it is a great benifit. That's kinda a no brainer. Like the other guy mentioned being able to do a WCS xyz sanity check is huge confidence booster when you're unsure about something.

I'd recommend learning both but when a new job opens up here and someone not so experienced takes over he can be doing pretty nice work in a fairly short amount of time just by learning how to run Gibbs.

I'd still bet most folks drive with some sort of Cam package even if well versed in coding manually. I don't mean 3D work either.

Brent
 
Last edited:
If you want to be along for the ride when a new machine comes in then just learn CAM. If you want to be the guy that can figure why the new machine is not working with the posted code then you're gonna have to learn the finer points of G and M programming. Someone has to be able to make or modify post processors or at least tell someone that doing the post what needs to go where and when in the posted code.
 
If you want to be along for the ride when a new machine comes in then just learn CAM. If you want to be the guy that can figure why the new machine is not working with the posted code then you're gonna have to learn the finer points of G and M programming.
That and six bucks will buy you a crappy cup of Starbucks coffee :)

Seriously, Vanc, maybe our day is done. What's in it for the extra effort ? Maybe just poosh de button and go home is more sensible. Put your mental energies into something more worthwhile, like making a model of the space needle out of toothpicks.
 
most cnc operators need to be able to program
1) do a tool change
2) set gcode defaults
3) turn on work offset and bring to XY position
4) turn on tool length comp and bring to initial Z safe part clearance height
5) bring to cutting height
6) turn on tool dia comp and bring to contour start at feed
7) program a contour shape
8) bring away from contour turn off tool dia comp
9) bring up to initial Z clearance
.
same with being able to program drilling a hole using G98 and G99 or bring drill to initial Z or to retract point between holes
.
not going to do anything complex manually programming but being able to do basic things most need to be able to do
 
most cnc operators need to be able to program
1) do a tool change
2) set gcode defaults
3) turn on work offset and bring to XY position
4) turn on tool length comp and bring to initial Z safe part clearance height
5) bring to cutting height
6) turn on tool dia comp and bring to contour start at feed
7) program a contour shape
8) bring away from contour turn off tool dia comp
9) bring up to initial Z clearance
.
same with being able to program drilling a hole using G98 and G99 or bring drill to initial Z or to retract point between holes
.
not going to do anything complex manually programming but being able to do basic things most need to be able to do
There's an answer in there some place. I just can't seem to find it??? Which part of learning finger code, or Cam where you addressing?
I'm trying to figure out what I should learn first.
 
I'd say that you at least need to know/understand the basics of finger banging..

When I say "Basics", I mean The Basics... Which direction is X+ type of thing.
Basic Cartesian coordinates.. If you don't understand that, then you are going to be
lost forever.. Have you ever seen a CNC knee mill that had an arrow on the table
pointing LEFT that said "X+"??? You don't want to be that guy...

The basics. G0 is go fast. G1 is go slower. G2 is take a left (or is it right) G3 is go
right. Understanding how I's J's and K's work... Understanding what a reference plane is
and WHY that is important....

Specific syntax in a canned cycle can be looked up... Basic safety block at the top, I don't
even know what half that stuff is without looking it up...

I would say its important for both finger banging and CAM to understand how cutter comp works..

Understanding the difference between absolute and incremental is important...


If you got excited in school when plotting points and making graphs, then you would love hand coding,
I hate it, takes too damn long..


Personally, I do a little bit of hand coding, but I never start from scratch... Even if I would
start from scratch, I'd modify an existing program, so that all the crap in the beginning is
already there and all the crap at the end is already there...
 
I had the joy of training my replacement at my last job. I had been trying to quit for almost 6 months and finally told the boss I was walking if he didnt give me someone to train. He gave me someone with little to no machining background, they didn't know what a Cartesian plane was, and couldn't make it into work by 8, when the rest of the ship started at 5. Did I mention it was his son?

In a period of about 3 months I had him to the point he could model, program, and machine some parts with minor complexity. He could program and machine the dumbest, I mean, most to complex parts engineering came up with.

He had a very basic understanding of G code, but could produce good work.

So, point of the story, learn some G code, at least the basics like bobw said. Enough that you have an understanding of what the machine is doing and why.
 
CAD/CAM all the way, without it I'd have been super screwed and probably never even gotten into CNC.
Now I do edit a bit of stuff by hand at the control but that's usually just changing the depth of something or feed rate/RPM, minor stuff, not full paths or strategies, for that I go back to cad/cam do a few tweaks, run the simulation and repost which only takes a few minutes. Just finished programming another part while having breakfast.
 
It depends on what you are trying to accopmplish and what your future looks like.

I learned G-code manually inputting it at the control. I now feel confident walking up to most any Machine and making chips, on relatively simple parts (and some not so much). I think it's integral to understand what the code is telling the Machine. To be able to stop it and look through the code to see what is coming next, to problem solve starts with Code, second is Mechanics. BUT in some shops it isn't needed. You have to know what environment you are working in. IE if you are the only guy there that needs to make parts and there isn't a second set of eyes, you probably want to be as independently capable as possible. Meaning not requiring outside assistance including CAD/CAM to make chips.

OTOH, I learned that way and now I use Gibbs, Mastercam and Esprit. I can edit my own post processors to output the code that I want, not the code that someone else thought was best for my application. The only reason/way I know what code I want, is because I am well versed in G-code programming.

CAD/CAM is not the be all end all of everything. It is a very powerful tool. Very few CAD/CAM programs can post to ALL the Machinery there is. There are blank spots. But that's all my opinion and you get what you paid for.

R

I couldnt agree more.
I am o.k. at reading G and M code. (I know many who are better) Most of what I do is in Gibbscam. Some Heidanhain, Mazatrol and Powermill. But when the code your CAM system post' doesnt work, you need to know what does work so that you can either fix your own post or at least be able to tell the post developers what you want / need. Most post developers are not machinist'. if you request them to make the post do something that a machinist or programmer would consider stupid ....often times they will do it....exactly as requested.
 
It would be pretty much impossible for me to make any money without CAM. Not saying it couldn't be done, just wouldn't make any money since I'd spend all day writing the code. Adaptive clearing aka HSM aka Dynamic Mill? Forget about coding that by hand. Yes, you can knock together some g-code that does HSM-style cutting, but automagically getting a full roughing cycle for a 4" x 6" pocket with a dozen standoffs and protrusions? I don't think so. Engraving lettering? Probably some conversational controls can do those nicely, but you are then just doing CAM at the control instead of at the workstation.

Regards.

Mike
 

Attachments

  • IMG_0975.jpg
    IMG_0975.jpg
    96.9 KB · Views: 277
  • Grip 1.jpg
    Grip 1.jpg
    96.5 KB · Views: 292
  • Univalve Back Cap.jpg
    Univalve Back Cap.jpg
    94.8 KB · Views: 301
I'm trying to figure out what I should learn first. With CAD/CAM it's so complex that it would be very difficult to edit manually. Programming manually would be useful for simple parts. I'm talking about programming on a Haas with G/M Code or learning Fusion 360 as I've had a hobby license for a while, jut never actually got into learning it. I bought some books for Fusion 360 almost a month ago hoping to teach me the basics but haven't even opened them. (I have a shelf of books that I do this with) Shop currently doesn't have a CAD/CAM but I think they're working on it. Even if it's a different program, I will have some knowledge. Although for all I know if could be in a week or they're just looking into it.

Thanks,
Higgins909

On a HAAS what? I am going to assume mill?
I believe the line should get drawn between mills and lathes.
2-azis lathe? Sure, you can finger bang your way through most anything there. If your really good at it, maybe even faster than CAM.
As soon as you walk over to the mill? Unless you are doing the absolute simplest of 2-D shapes? Or just drilling holes?
Quit wasting time, and learn the CAM.
But, you STILL need to be able to read/understand the code the CAM spits out. If for no other reason than crash avoidance.
Because, if you cant read code, and your sketchy with CAM, your gonna crash.
 
I'm trying to figure out what I should learn first.

Both.

You're kinda asking should I learn how to speak english first or use Microsoft Word. I know, bit of a stretch but it makes a point. You are trying to create gcode with your CAD/CAM software. If you have no idea what gcode is, you are putting 100% trust into your software. Then, what if the code is not correct? How will you fix it if you don't know gcode?

My point being you can learn both at the same time. Besides, a basic understanding of gcode is not a huge undertaking. You can know enough about gcode in a few months...CAD/CAM, you'll be forever learning more.
 
You need to learn the code first so you can understand what the CAM is spitting out. How are you going to know that the code isn't going to cause a crash if you can't read it?
 
The OP's question is simple. Is it worth it? Yes.

But if you're asking if you should learn one or the other, I don't think there is just a plain answer to that. I guess in 2018, if I knew what I know and had to choose between the two (which would be asinine and ridiculous BTW) I'd invest in learning a solid CAD/CAM suite.

R
 








 
Back
Top