Way to deflect, artful dodger.

Ok. I'll try to answer your questions even though you have already answered them for me. Yes, MCUs can do analog, digital, Ethernet, all sorts of networks, touchscreen and LCD as well. Plus millions of other things which PLCs cannot do because they're limited by rather restrictive software.

I'm not going to discuss the part where you said that you were a duped idiot. I don't know you, so with your permission, I will skip that.

 

but the discussion is about PLCs and substitutes. Do we need a bios engineer to run this thing or can an electrician use it as a troubleshooting tool?

Every DIYer or hobbyist or engineers can be use the tester that I attached as a good tool, it is quite easy to use, I made that example just shown that he can do and modify the multi-meter, and some multi-meter used the ATmega Chip to design, and that is the same with arduino.

I have no idea the situation in Canada, but in Taiwan, not every electrician or engineer know about PLCs, only when they are interested in that area or their job have to used it, otherwise most of electricians or engineers no any related with PLCs, the situation similar as not every Chinese know Chinese Kung Fu or Tai-Chi.

 

There's a difference between form and function.

"sub-$100 digital bricks". This can be done with small 50-cent MCU plus few discrete components. Will cost you $5 manufactured and shipped in small quantities. Takes few days to manufacture, including pre-programming of the chips. It will take less space and will be more reliable because there's a much simple design. What is the advantages of the "sub-$100 digital bricks" over this?

["What's the advantage?"] depends on ["what's the application?"] and ["Who's doing it?"]

Let's say the application is a small automatic toroidal transformer winder, and ["Who's doing it?"] is the OEM, a chinese manufacturer of bargain quality machines. In this case, it absolutely makes sense to use a small custom PCB with an MCU on it. All that's required and expected of of a chinese OEM is that before the machine goes into a crate, it winds toroids. A year from now, not their problem. They don't offer tech support and they don't send out field service guys and they don't care if it quits working and you have no way to fix their "proprietary" control board. Maybe they sell replacement control boards, but if they do, there's something "different" about themthat makes the replacement less than straight forward.

Now let's say the application is a 2 y/o chinese bargain toroidal transformer winder and ["Who's doing it?"] is me (in my previous role), a field service technician. I get called in after the owner has already wasted 2 weeks with his bread&butter machine down, trying to get some answers out of the chinese to no avail. Now it's an emergency. Now he just wants the damned thing working ASAP, and willing to pay for it. Am I going to waste his time taking the crippled board back to the shop and trying to reverse engineer it and duplicate it and have it made for me (in china probably) and then get started trying to reprogram it? Hell no. I'm going to pull out the brick PLC that's already in my service vehicle and get started writing a control program for it right immediately, on the spot. I'll have the full story from the operators on how it's supposed to work within 30min, and I'll have it wired up and programmed, winding toroids again before the end of the shift. The owner will call me a hero and the $100 PLC I sold to him for $200 will not even enter his thoughts. It will be totally eclipsed by the loss in revenue he's already sustained. He'll kick himself for not calling me sooner, and in the future, he will. And I'll gladly sell him more marked up PLCs and he'll gladly pay for them.

Now, let's say ["Who's doing it?"] is an Austrian OEM and the application is a $2.5M cabling line with 40 conductor bays, with 5 analog sensors and a VFD in each bay. 3 HMI control stations along the length of it, and Profibus communications between all components, as spec'd by the American buyer. Do I really need to go into this? Is anybody (even you) going to argue in favor of this thing being controlled by a hackjob circuit board?

 

, touchscreen and LCD as well. Plus millions of other things which PLCs cannot do because they're limited by rather restrictive software.

I have added an HMI to PLC's, The likes of Allen-Bradley have there own, Maple Systems offer anything from a 3 line display to a kiosk graphic touch screen for most makes of PLC's.
So no restriction there for me.
Max.

 

There's a difference between form and function.

"sub-$100 digital bricks". This can be done with small 50-cent MCU plus few discrete components. Will cost you $5 manufactured and shipped in small quantities. Takes few days to manufacture, including pre-programming of the chips. It will take less space and will be more reliable because there's a much simple design. What is the advantages of the "sub-$100 digital bricks" over this?

Somebody's time is real cheap, and it ain't mine.

 

Ok. I'll try to answer your questions even though you have already answered them for me. Yes, MCUs can do analog, digital, Ethernet, all sorts of networks, touchscreen and LCD as well. Plus millions of other things which PLCs cannot do because they're limited by rather restrictive software..

I'm starting to wonder if we are talking the same thing. An MCU in an LCD couldnt hope to perform as a PLC because of it's restrictive software. A PLC wouldn't drive an LCD because it's not intended nor configured for such tasks. An MCU is the heart of a PLC, configured to serve a broad range of tasks. What makes the configuration so usefull is it's ability to be reconfigured. I'm not sure where you practice, but my exposure to industry (30 years) haven't discovered this gross misuse of PLCs that you describe. Yet, as I've posted, MCUs in many configurations have been utilized in industry before the advance of PLCs. I think you need to get out more.

 

Every DIYer or hobbyist or engineers can be use the tester that I attached as a good tool, it is quite easy to use, I made that example just shown that he can do and modify the multi-meter, and some multi-meter used the ATmega Chip to design, and that is the same with arduino.

I have no idea the situation in Canada, but in Taiwan, not every electrician or engineer know about PLCs, only when they are interested in that area or their job have to used it, otherwise most of electricians or engineers no any related with PLCs, the situation similar as not every Chinese know Chinese Kung Fu or Tai-Chi.

The last two installs I performed had every control wire in the plant terminate in a PLC rack. As an Electrician whos job it was to troubleshoot and repair, you have to know how to search your logic code (analogous to reading a schematic) and identify your IO, otherwise you have little to no chance of successful repair.

 

The last two installs I performed had every control wire in the plant terminate in a PLC rack. As an Electrician whos job it was to troubleshoot and repair, you have to know how to search your logic code (analogous to reading a schematic) and identify your IO, otherwise you have little to no chance of successful repair.

About the troubleshoot and repair, there are three parts, maybe you never meet that problem and don't meet that is better.
1. The first is in the operation manual, I met once the problem was in the communication code of the PLC, I was used VB to control PLC with a PC-Based HMI(Human-Machine Interface), the PLC is a big brand of Japan, I was followed the instruction, but whatever I tried that I still can't communicate with the PLC, and I called to the seller to asked that problem, the service employee said the code is correct, because I tried too many times, so I gave up to used that code, and try my method, it was spent no much time, the communication is working, so sometimes we will saying that don't trust the books otherwise would rather than no books, when we buy a ee book, the descriptions could be wrong, the circuit(schematic) could be wrong, even we find the circuits from Internet also could be wrong, so the book just for reference, sometimes we need to via the experiment to prove the circuits are correct.

2. The second is in the PLC module itself, if a PLC module was damaged then we almost will send back to the seller, they have repairing engineers to fix it, but if the warranty date was expired then it is depends on the ability of the engineer of buyer.

3. The third is the whole system, many PLCs will be working with other devices, so when the problem is occurs, the ee engineer or electrician have to identify where the problem is, this will be needs the experience to judge.

 

I'm starting to wonder if we are talking the same thing. An MCU in an LCD couldnt hope to perform as a PLC because of it's restrictive software.

Don't understand what you're saying. I merely said that you can program anything into MCU. On the other hand, what you can program into PLC is limited by its (PLC's) software.

I only mentioned LCD because Strantor was wondering if MCU could drive LCD, and I said that yes it could.

 

Now let's say the application is a 2 y/o chinese bargain toroidal transformer winder and ["Who's doing it?"] is me (in my previous role), a field service technician. I get called in after the owner has already wasted 2 weeks with his bread&butter machine down, trying to get some answers out of the chinese to no avail. Now it's an emergency. Now he just wants the damned thing working ASAP, and willing to pay for it. Am I going to waste his time taking the crippled board back to the shop and trying to reverse engineer it and duplicate it and have it made for me (in china probably) and then get started trying to reprogram it? Hell no. I'm going to pull out the brick PLC that's already in my service vehicle and get started writing a control program for it right immediately, on the spot. I'll have the full story from the operators on how it's supposed to work within 30min, and I'll have it wired up and programmed, winding toroids again before the end of the shift. The owner will call me a hero and the $100 PLC I sold to him for $200 will not even enter his thoughts. It will be totally eclipsed by the loss in revenue he's already sustained. He'll kick himself for not calling me sooner, and in the future, he will. And I'll gladly sell him more marked up PLCs and he'll gladly pay for them.

An interesting story. Now I understand. You don't mind buying PLCs (which are not much more than $10 MCU in a nice box) for $100 because you can unload them to the next guy for $200. This makes perfect sense. Sure as hell, you would never be able to sell him Arduino or anything alike for $200.

 

An interesting story. Now I understand. You don't mind buying PLCs (which are not much more than $10 MCU in a nice box) for $100 because you can unload them to the next guy for $200. This makes perfect sense.

Yep, supply and demand. It's pretty awesome when you understand how it works and play along. I give people what they want, charge them well below what they are willing to pay, and everybody walks away happy.

Sure as hell, you would never be able to sell him Arduino or anything alike for $200.

You're right, I wouldn't. By the time I would have returned to my customer's location after a stint behind the bench assembling all the custom interface circuitry to turn an Arduino into a drop-in control system for an industrial machine, my customer would have called my competitor, someone who knows what "JUST MAKE IT WORK, NOW" means, and they would be happily winding toroids with a $200($100) brick PLC that I didn't install. And if they did wait around for me, I'd have a pissed off hot head breathing down my neck while debugging the on-the-fly programming with no I/O status feedback, just "guessing" WTF is going on by passing a bunch of klugey print commands to the serial port. I would never get called back, and I would expect that after such a mishandled job.


Your point would have a lot more validity in any other market sector. In this industrial setting (the very setting that PLCs are designed for), time is #1 priority. Manufacturing, Oil Field, etc., they all demand speed and efficiency, no matter the cost. They don't care how much you charge them (within reason, of course). That's speed of delivery of new equipment (applicable to OEMs) and efficiency of repair/troubleshooting (by marginally trained, average intelligence people). UP TIME, DOWN TIME, bottom line. If you're working on consumer electronics and you can just tell people "I should be done designing a new control board for your microwave by the end of next month," they'll probably be OK with that, as long as they're saving that $130 that an OEM replacement would have costed. But this guy with the toroid winder, he's losing $130 every hour. A $200($100) PLC is NOTHING to him. It simply does not make sense to pinch pennies and go back to the drawing board every time you need a control solution, when there's and infinitely versatile one sitting on a shelf in a box already that just needs to be installed and programmed.

 

About the troubleshoot and repair, there are three parts, maybe you never meet that problem and don't meet that is better.
1. The first is in the operation manual, I met once the problem was in the communication code of the PLC, I was used VB to control PLC with a PC-Based HMI(Human-Machine Interface), the PLC is a big brand of Japan, I was followed the instruction, but whatever I tried that I still can't communicate with the PLC, and I called to the seller to asked that problem, the service employee said the code is correct, because I tried too many times, so I gave up to used that code, and try my method, it was spent no much time, the communication is working, so sometimes we will saying that don't trust the books otherwise would rather than no books, when we buy a ee book, the descriptions could be wrong, the circuit(schematic) could be wrong, even we find the circuits from Internet also could be wrong, so the book just for reference, sometimes we need to via the experiment to prove the circuits are correct.

2. The second is in the PLC module itself, if a PLC module was damaged then we almost will send back to the seller, they have repairing engineers to fix it, but if the warranty date was expired then it is depends on the ability of the engineer of buyer.

3. The third is the whole system, many PLCs will be working with other devices, so when the problem is occurs, the ee engineer or electrician have to identify where the problem is, this will be needs the experience to judge.

And how would your proposed replacement resolve such issues

 

Don't understand what you're saying. I merely said that you can program anything into MCU. On the other hand, what you can program into PLC is limited by its (PLC's) software.

I only mentioned LCD because Strantor was wondering if MCU could drive LCD, and I said that yes it could.

are you suggesting that the millions of embedded applications used in industry are useless because most of them have proprietary software where the only programming available is through menu screens, keys, or other simple interfaces.

 

@NorthGuy ,
I think we can resolve this disagreement and benefit each other at the same time.
You design a replacement for my brick PLCs and I'll buy it.

I want (10) 24VDC/120VAC tolerant transistor I/O (configurable as input or output) with >50kHz Pulse capability
I want (10) relay outputs 120V/10A inductive contacts
I want (2) analog inputs and (2) analog outputs, configurable for 0-10V or 4-20mA
I want it to speak CANBUS, MODBUS, RS485/RS422, RS232, Ethernet I/P, and USB
I want all the inputs and power input polarity protected, spike/surge protected, and I want everything hardware failsafe.

I want a software interface where I can see my I/O status, in the programming, in real time, and an LCD on the unit which displays I/O status.
I can handle text programming, but I don't want to use the uC IDE. I want a software package applicable directly to the "PLC-like" device. Try not re-invent the wheel; use an existing language, I prefer Python, but whatever you decide I'll go with.
I want it capable of floating point math and with several kB of non volatile memory.
I want an accurate RTC.
I want it to understand commands from other PLCs and OPC servers (produce/consume/msg/etc) and play nice on a network.
I want it smaller than 6"X6"X3" and in a NEMA/IP-rated dust-protected DIN rail mountable enclosure.
I want 24/7/365 tech support and a warranty.

To sweeten the pot, I'll not require you to obtain any type of certification or perform any exhaustive compliance testing. AND I'll pay YOU the $200, so I can charge my customers $300 for it.
I'll need these available with less than 2 weeks lead time, in any quantity from 1 to 1000.

How does that sound? Do we have a deal or what?

 

are you suggesting that the millions of embedded applications used in industry are useless because most of them have proprietary software where the only programming available is through menu screens, keys, or other simple interfaces.

I'm not suggesting or implying anything. I said that PLCs have restrictions imposed by the programming software - the one which provides "nice" graphical interface. I also said that MCUs do not have such restrictions.

 

How does that sound? Do we have a deal or what?

So you propose me to build a super-PLC for you. This all doesn't sound too hard, except for 24/7/365 support, of course. The problem is that it is rather big and will require somewhere 8 months to a year to develop during which time I need to earn some living.

Just out of curiosity, how much would the set of PLC modules which do all these things would cost?

 

Just out of curiosity, how much would the set of PLC modules which do all these things would cost?

A lot more than $300, but that isn't really an apples to apples comparison since you wouldn't be building a PLC (for your sake, i know you wouldn't want your name associated with that acronym) but instead PLC-like device (call it digital cerebrum or whatever you fancy). And your motives are anti-profiteering so I'm not really sure why you would care.

 

A lot more than $300, but that isn't really an apples to apples comparison since you wouldn't be building a PLC (for your sake, i know you wouldn't want your name associated with that acronym) but instead PLC-like device (call it digital cerebrum or whatever you fancy). And your motives are anti-profiteering so I'm not really sure why you would care.

I don't.

So, if you agree to give up the support, want to pay $200K now and $200K after about a year, you will get 2,000 of these brand-new PLC-like devices, plus whatever you want to order afterwards with 2 week lead time and minimum quantity of 100. I can start in March.

 

And how would your proposed replacement resolve such issues

If you mean that to change the brand of PLC?
I can't do that, because I was a Technical Adviser, and the factory had their thought to decide which brand of PLC should be use, the things always can't follow as our will, because my job in the beginning was only to apply the new tech engineers, and training them, but because the executives of company didn't like the applicants, you can say that the company has it's own request, so it can't be easy to find the good engineers get low pay, and that's why I have to involved and should be learn how to write the VB program within two weeks and start to write the program later, but I never touch the VB before, so you know what is the pressure ...

 

I don't.

So, if you agree to give up the support, want to pay $200K now and $200K after about a year, you will get 2,000 of these brand-new PLC-like devices, plus whatever you want to order afterwards with 2 week lead time and minimum quantity of 100. I can start in March.

You're asking an awful lot from a guy who's repeatedly praised the solution he's already got. You're the one who sounds like he thinks the industry has their collective head up their collective ass, and I was issuing you a challenge, to put your money where your mouth is. Don't try and turn it around on me.

Point is, if you really believe in all this crap you're spouting, then from your point of view, this should appear as the opportunity of a lifetime, with a red carpet rolled out just for you. You're the only one who hears the market silently crying out for this new solution that it doesn't yet know that it needs. Go out there and answer the call. Take the opportunity and boldly go where no man has gone before. I don't know why you aren't doing it yet. Why are you on this forum wasting time, filling the internet with your brilliant ideas? Why are you giving away these golden nuggets?

Why not? Because not even you believes what you are saying. I theorize that it's part of the canned social/political ideal package you subscribe to, to hate PLC manufacturers because they make the big bucks. That whole 1% and 99% thing. They're evil because they're greedy, capitalist pigs. Well, whatever. To each his own. I don't have time for this discussion anymore. I have opportunities to follow up on. Good talking with you.

P.S. If you do invest in your ideas and you do make some kind of device, let me know, I might be interested. If you make that exact device I described, I really will buy it for $200/pop.