Learning Electrical & Automation

Jashley73 · Jul 9, 2014

I want to pick the brains of some of you here regarding learning the building blocks needed to get into automation - particularly automating machine tools with robots and other peripheral equipment.

For some background, I've been in machining for about 9 years now, proficient with manual and CNC machines. Have some basic understanding of electrical stuff, but not enough to really count for anything. I can poke around on the ladder on a CNC control to try and 'visually' determine if something isn't quite right. My employer installed our first robot a little less than a year ago, so I'm familiar with the basics of I/O's and how that relates to the robot and the cell in general.

But getting into interfacing, PLC's, ethernet communications, and other equipment, I'm in the dark.

I've considered taking some technical-college courses, that offer individual courses up to an associates degree. Maybe this would be the way to go, maybe not...

To further muddy the waters, I'm not sure that I really want to get away from machining entirely, but would be interested in understanding how to build a CNC Machine/Robot automation cell if needed. (I know this includes RIA safety stuff as well...) I might be interested in learning some general control's too, in case I decided to build machinery later on.

So, I'm just curious, how would you advise someone interested in building their own CNC/Robot cells? Thanks.

Jashley73 · Jul 11, 2014

Hmm, I thought by now some of our resident automation folks would have chimed in. 'Was really hoping to hear som of your thoughts on this. Anyone?

ConstantWave · Jul 11, 2014

Not an automation expert by any means but a community college course if you can find a good one can be very helpful and possibly the fastest way to learn the basics. I took an evening course last summer that covered industrial maintenance along with some basic automation principles and found it very helpful.

greenbuggy · Jul 12, 2014

I'll also be watching this thread. I've done some automation work at my current job and really enjoyed it. It would be nice to get the papers to prove I know what I'm doing when I'm job-searching further down the road.

Tonytn36 · Jul 12, 2014

I'll try to answer sometime Sunday. Currently in the 3-Brother build and LATE! (Haven't been able to get parts in fast enough.) Drove to the heat treater's last night and picked up the last major component, so I can installed that assembly this morning (Hopefully).

Tonytn36 · Jul 25, 2014

You really need a good course in electrical controls. Not so much for PLC programming as to understand how the PLC works, how the different types of sensors work, signaling, etc. You also need a programming class - again, not so much specifically for one thing, but to understand programming syntax and basic logic flow, etc. Programming is programming...the differences in the types is the syntax and features available. C++ would be a good place to start - many of the robots are based on it. Then there comes the safety aspects, the RIA is a good place to start on the robot side, but ANSI / ISO / OSHA / NEC are going to cover the remainder.

There is a lot to be learned on the engineering side as far as process goes. A good course in industrial process would help here. Then there are the engineering aspects of payload, reach, motion, timing, end effector design, peripheral equipment design and layout to meet the required cycle time, throughput and part quality. How the part is presented to the robot can make all the difference in the world in how the cell runs and the quality it produces.

Jashley73 · Jul 26, 2014

So am I right in saying, based on what you said above, that focus on Controls should come first?

Like I said above, I'm mainly interested in how to build a CNC-Robot-Gaging-Peripheral cell for down the road, so I imagine that communication would be a big area of focus too.

The more I think about this, I am beginning to think this might be an ambitious goal, but I still would like to have a better understanding. It would be nice to be able to piece-together components to build a working system - say robot tending a machine, or integrating peripheral measuring into CNC machines, etc...

Thanks Tony...

omrc7771 · Jul 26, 2014

Most tech schools refer to this program(associates degree) as electro-mechanical technology. I finished the program in the mid 90s and it covered everything your looking for and then some. Well worthwhile.

Miguels244 · Jul 26, 2014

I never had a chance to take a course of study.
I learned everything under fire so to speak.
If you can then take some basics classes so when the SHTF you know where to start.
You can also get specific training from most vendors, call the local apps engineer.
Some are free some are not (Rockwell) but they DO have a serious interest in you knowing how their stuff works.
Know the difference between NC/NO sensors and Sinking/Sourcing.
Know when to isolate.
Understand when latency (slowness) is an issue.
Learn EthernetIP and a few other ethernet protocols...and the old serial ones as well.
There are probably a dozen reasonably common comm systems...largely the same but all different, figure each will need to be learned as you go.
If you plan on using any euro PLC except Siemans learn IEC 6-1131 programming, the IDE is by CodeSys and is free still I think.
Understand closed loop versus open loop motion.
Remember that timers are bad in motion, you really want to use position switches.
Head over to *******, it's not industrial grade but you might learn something.

But seriously, if there's a decent program near you take it.

matt_isserstedt · Aug 1, 2014

I learned what I know from digging into systems built by others, sort of a low-level understanding at best. I've worked a lot on adding and modifying parts of the big system but never built a complete one. But in a factory where there can literally be 100 or 1000 different machines you can see why intimate knowledge of a single one never gets built up. What I have learned is that things like networks and communications between devices, and now part-dataflow(replacing or improving on hard-wired I/O) are the most challenging to understand. Program a PLC...is just logic, and understanding the advanced functions. Program a robot...is steps, points, and logic, and understanding the advanced functions. (Similarly, program a CNC is just Gcode, steps, and understanding the advanced functions). But to get everything to talk and function and work together (and...for the maintenance guy...self-diagnosis about what is wrong and why the automated process has stopped!!!

) is very challenging.

Machinefixer · Feb 12, 2015

Look for a class on instumentation and process control. It will cover basic PLC / Ladder Logic, Digital and Analog I/O's such as sensors for Input. And VFD's, Servo Motors, Actuators, ect. for Output. Things that, that robot your employer bought, has to have to know where to go, when, and what to do when it gets there.

CalG · Feb 13, 2015

I would say that IF a person were wanting to get to know CNC and automation,
Linux CNC and a set of 3 servo motors with drives and limits would get you more experience than two years of "formal tech school" But without the valuable piece of paper.
That would end up as a PC based system with all it's limitations. Proprietary hardware just makes it easier. ;-)

Combine that with information gained from such online boards that are open and available, progress will come quickly!
If one can grasp the basics of motion control and logic along with the many twisted avenues of "electro-magic" A return on the time investment is assured.

Here is an example of a PLC information source
Your Personal PLC Tutor Site - Interactive PLC Questions and Answers

No affiliation or endorsement. (warning! there is a strong commercial aspect at that sight)

Think of a PLC as a "standardized industrial language". logic programming could take place in many languages. But why learn on a platform not universally acknowledged in the field?
Pick a big name brand and get fluent!

Just my 2Cents

Jashley73 · Feb 13, 2015

After putting this whole process-redesign to bed for several months, we're beginning to re-investigate our options. On option that looks the easiest, and cheapest so far, is to simply move the measuring inside the machine, by using the Renishaw touch-probes that we already have, but haven't been using to their full potential.

I did some testing several weeks ago, and with the appropriate settings, the probe will measure a diameter repeatably with less than 2 microns total deviation, on a known standard - well within our tolerance. The idea is to utilize this in-machine measuring at the appropriate intervals, and to insure that we are getting accurate readings, use a known surface inside the machine to "master/confirm" the probe is in good working order.

We can even output information from the probe/machine control as well, and we'll need to talk with our SPC guys, but my hope is we can output the information from the machine control straight to the SPC system...

So hopefully, we can save ourselves $100,000 by making use of something that we already have...

CalG · Feb 13, 2015

Sounds like you are "getting fluent".

It's not rocket science! ;-)

adama · Feb 14, 2015

I have a ONC in electrical and electronics, ok it was only day release for 2 years and some evenings, but it really covered all the basics. Yeah you need to know about plc's etc for automation, but having a good understanding of electronics and especially 3 phase motors and there quirks is a great bedrock to learn the automation world off of.

To learn to programme PLC's and such its just a case of playing with them. Eventually the language just becomes as common as G code or even english. Its not so much hard, just needs some practice and understanding. Same with setting up drives and such, first ones hard, second one even if its a different make you have some idea at least and by the time your on your 4-5th one its easy.

A fair few pro's knock the PC controlled cnc stuff like Linux cnc, yeah its no Siemens control, but it encompasses nearly all the same basic functions, it needs the same basic thing to work, its a great place to practice stuff and you do end up with something usable at the end of the day too. Whats more the gap between "real" CNC and Linux CNC is closing, rigid tapping is now a real reality. The motion controls continue to improve ETC.

The hard thing i find with automation is staying current. Whilst the print stuff i deal with does not use robots, the sensor tech is ever changing and knowing what options there are for sensing is a key part to keeping machines running reliably.

Jashley73 · Feb 14, 2015

Jashley73 said:
After putting this whole process-redesign to bed for several months, we're beginning to re-investigate our options. On option that looks the easiest, and cheapest so far, is to simply move the measuring inside the machine, by using the Renishaw touch-probes that we already have, but haven't been using to their full potential.

I did some testing several weeks ago, and with the appropriate settings, the probe will measure a diameter repeatably with less than 2 microns total deviation, on a known standard - well within our tolerance. The idea is to utilize this in-machine measuring at the appropriate intervals, and to insure that we are getting accurate readings, use a known surface inside the machine to "master/confirm" the probe is in good working order.

We can even output information from the probe/machine control as well, and we'll need to talk with our SPC guys, but my hope is we can output the information from the machine control straight to the SPC system...

So hopefully, we can save ourselves $100,000 by making use of something that we already have...

Oops! I though this was the thread where I was asking about building a post-process, automated measuring gage. Shoulda' read my own thread...

Muffler Bearing · Feb 19, 2015

Start with reading ALL of Tony Kuphaldt's free (Creative Commons) books. I am linking to Socratic Instrumentation but I think he has a half dozen other books.
Socratic Instrumentation
They are very dense in information, and really can teach quite a bit about the specific industrial concepts that really aren't touched in any other books.

WILLEO6709 · Feb 19, 2015

I started with a book on programmable logic controllers.... I did not understand a lot of it when I first read it....

then I got into a job where I had to use a low level koyo ( fanuc series 1 ) plc and it all started to click. Then I bought some old HMC's that hadn a hydraulic jig hard coded in the ladder.... I spent 2 days with a fanuc tech spent mainly getting me to the ladder and understanding what not to do....

then I created a pair of m codes to make the " jig happy " input so I could run the spindle.

PLC's are the basics of automation,,,, start there and many other things follow the same logic.

CarbideBob · Feb 19, 2015

You learn this best by jumping in way over your head.
You seem on a good track and in my shop I'd push you hard forward on it.
I won't touch your in machine measuring for final part SPC, you will get how this can screw you up.
(Fine to trust everyone, but one should always cut the deck)
Bob

Jashley73 · Feb 19, 2015

CarbideBob said:
You learn this best by jumping in way over your head.
You seem on a good track and in my shop I'd push you hard forward on it.
I won't touch your in machine measuring for final part SPC, you will get how this can screw you up.
(Fine to trust everyone, but one should always cut the deck)
Bob

Bob - No doubt there's room for caution when having the same machine cut and inspect the part. The good news however, is that there are a couple warm-bodies that handle and inspect these features before the part is "done." But certainly there's several aspects that we need to think through before this becomes a "trustable" process. In-machine mastering is one. (Is there something else I need to be aware of?)

Regarding electrical and automation education anyway, as of recent that has kind of taken a back-seat to other things. Here lately I've been wrapped up in fixture/tooling design & build, and now have an "apprentice" that I'm training. So between keeping the machines running, designing new stuff, teaching the apprentice how to make said new stuff, and everything else I stay pretty busy. And no new automation related stuff on the horizon for now either... Maybe soon though - time will tell.

Learning Electrical & Automation

Titanium

Titanium

Aluminum

Stainless

Diamond

Diamond

Titanium

Hot Rolled

Diamond

Diamond

Plastic

Diamond

Titanium

Diamond

Diamond

Titanium

Aluminum

Diamond

Diamond

Titanium

Similar threads