Why is it that we must always pull teeth? Year end rant.

[Peter Nachtwey]

make that 504

I seem to add a few followers every beginning of a new semester. Most must be college students.


BTW, I do things much differently that what is taught in colleges. Colleges and books don't even cover what what I do.
At least they don't in the same way.

 

[I_Automation]

I seem to add a few followers every beginning of a new semester. Most must be college students.


BTW, I do things much differently that what is taught in colleges. Colleges and books don't even cover what what I do.
At least they don't in the same way.

I was told a long time ago (Pre-PLC's) that anything you can learn in college is old technology and ancient by the time you graduate.


The technology has to be commonly used to be worth writing a textbook about it,

writing a textbook takes over a year,
developing a course to accompany the textbook another year,
selling that course to schools,
training (maybe, at a good school) the faculty on the course,
the school adding it to their next semester,
you learning it for the semester and graduating maybe a few years later hoping to use that knowledge.


How long did it take to get courses on Windows 10, and now Windows 11 is out.

 

[Peter Nachtwey]

Another rant.

I was told a long time ago (Pre-PLC's) that anything you can learn in college is old technology and ancient by the time you graduate.

You mean like Z-N?
Most half of what it taught is ****. It is taught by professors that have NEVER had to use it.

The technology has to be commonly used to be worth writing a textbook about it,

And that this is another rant. Teachers teach what they have been taught.

writing a textbook takes over a year,

I have 30 years of Mathcad files. About 1000 of them of which about 800 are any good. Only Norm and Pandiani have seen what I have done. I could fill a few books.

developing a course to accompany the textbook another year,
selling that course to schools,

Naw.

training (maybe, at a good school) the faculty on the course,
the school adding it to their next semester,
you learning it for the semester and graduating maybe a few years later hoping to use that knowledge.

This is yet another problem and why I will never have that many followers. Only a few people need to know this information. Maybe one in a hundred if that. So I can see why there aren't that many interested.

 

[Tom Jenkins]

I understand the dystopian view, but I personally find it greatly exaggerated.

I was told a long time ago (Pre-PLC's) that anything you can learn in college is old technology and ancient by the time you graduate.

Newton's Laws are still pretty useful a few hundred years after their first teaching. I was taught that half of what you learn is obsolete in five years. I consider even that an exaggeration.

If your school taught specifics of hardware and software and not underlying principles then they did you a disservice.

writing a textbook takes over a year,

That I agree with, from personal experience.

developing a course to accompany the textbook another year,

Not at all - I've done it in a few weeks.

How long did it take to get courses on Windows 10, and now Windows 11 is out.

This is exactly the kind of specific instruction that is necessary but should never be the emphasis of a good education. The underlying concepts of logical thinking, the interaction between man and machine, and goal-oriented effort are universal and much, much more important.

 

[Tinine]

Hey guys,

Been lurking for a good number of years, mainly because of Peter's posts. Only registered a couple of months ago.
Strange that I often follow Delta-Motion because they are intimately familiar with servo-hydraulics.

I have only just joined the dots and I had never heard of Peter Ponders PID. I'll be there for sure.

My $0.02....Is it just me that finds it impossible to understand math notation?

Not exactly new to motion control, I was Galil's first machine-control customer back in the mid-80's and I had Jacob Tal on speed-dial

Shipped thousands of controls world-wide and shipped hundreds of my own CNC tube-forming machines, mainly to big-auto.
Today, I design/build my own motion controllers (latest is based on the Parallax Propeller-P2). I have only ever relied on borrowed code snippets and experimentation because this head finds it impossible to follow this old math notation. Why do we even have this garbage any more? What's wrong with pseudo-code that is easily understood?
I believe that many engineers remain ignorant of how this stuff works because they have the same problem and don't want to admit it.

Craig

 

[Peter Nachtwey]

My $0.02....Is it just me that finds it impossible to understand math notation?
No! I have trouble too. It is because the books/instructors are not consistent in the symbols they use for certain variables all the time. Math has a language all its own. Often there are no worked examples. I like examples.

Not exactly new to motion control, I was Galil's first machine-control customer back in the mid-80's and I had Jacob Tal on speed-dial

I went to a Jacob Tal seminar back in the late 80s or early 90s. I still have to book from the seminar. I learned few significant things. I remember twisting the motor that he had for a demo to get a "real feel" of what the P, I and D do. I learned the strengths and weakness of the Galil controllers.

I believe that many engineers remain ignorant of how this stuff works because they have the same problem and don't want to admit it.

Craig

They must have had the math, calculus, differential equations and linear algebra in college. The problem is that they forget what they learned or almost learned and consider themselves lucky to survived the course. The problem is there is a "Use it or lose it" factor. So many forgot what they were taught. Some we NEVER taught very well. Many professors have NEVER had to solve these problems in real life. Some have trouble applying what they know.

I remembered Runge-Kutta, calculus, differential equations, and linear algebra from college. The rest I had to learn on my own. PIDs, the closed loop part is easy compared to the target generator that controls the motion.

Back on topic. The problem is that the OPs don't even say what their application is and what they are trying to do. It is almost like the don't think anyone has done something like it before.

 

[plvlce]

Newton's Laws are still pretty useful a few hundred years after their first teaching. I was taught that half of what you learn is obsolete in five years. I consider even that an exaggeration.

If your school taught specifics of hardware and software and not underlying principles then they did you a disservice.

+1

For example... whether the school teaches mathematics using graphing calculators, Mathcad or slide rules, what you should be learning is math, not technology.

If all you learn is 'technology' then you will be perpetually behind as technology changes. Understand the principles that the technology is there to harness and you can transfer your knowledge to the new technology as it comes out with a much shallower learning curve.

 

[Tinine]

@Peter

"use it or lose it"...Exactly my point

Your s-curve solution came with a code example. Wonderful because it's immediately understandable.

The 2008 feed forward thread had me convinced that I was dropped on my head at an early age. Went through it a zillion times and I'm like "show it in actual code, dammit"

Craig

 

[Peter Nachtwey]

+1

For example... whether the school teaches mathematics using graphing calculators, Mathcad or slide rules, what you should be learning is math, not technology.

If all you learn is 'technology' then you will be perpetually behind as technology changes. Understand the principles that the technology is there to harness and you can transfer your knowledge to the new technology as it comes out with a much shallower learning curve.

The basics are what I call "forever knowledge". It never changes. PLCs, HMIs, motion controllers etc are all just tools.



This is exactly why I think teaching using Matlab is a crime. Matlab is good at getting answers but not so good at showing how they are derived.

 

[JeremyM]

The basics are what I call "forever knowledge". It never changes. PLCs, HMIs, motion controllers etc are all just tools.



This is exactly why I think teaching using Matlab is a crime. Matlab is good at getting answers but not so good at showing how they are derived.

You mean pencil and paper aren't things anymore?

 

[joseph_e2]

I last took Calculus/Diff Equations in 2008-2010. We were not allowed to use formula sheets or anything electronic at all during tests. Most of the students protested vigorously, but it was a math-department-wide policy for all classes so they got nowhere. I think it was a good rule. You need to understand the concepts before letting devices do it for you.



Of course, the perishable knowledge has largely perished due to lack of use in my day-to-day. I *hope*, though, that it would come back if I needed it...