Arduino PLC is a real thing now

Thread Starter

strantor

Joined Oct 3, 2010
5,206
I've seen or participated in many a thread over the years discussing the use of Arduino as a machine controller, and the soundest advice most often given, is that it's not a great idea. Arduino is not rated for the rigors of industrial control. It can't be relied upon like a PLC can. It hasn't been torture tested the way PLCs are. It doesn't have native support for the signals used in industry. It's a kludge at best. Well AutomationDirect went and changed that. They beefed up the Arduino and stuck it in a DIN-rail mountable package with a backplane that interfaces with their line of Productivity 1000 I/O modules. You can now connect hardened 120VAC, 24VDC, 4-20mA, et. al. industrial I/O modules "Directly" to arduino, as well as shields from wherever. Best thing is, they stuck to the open source theme that Arduino embodies. Even their IDE is open source and available on GitHub (you can still use Arduino IDE if preferred). Schematics and gerbers can be downloaded for their CPU module and their shields (but not for the I/O cards which were developed for another PLC line, and predate arduino)

https://www.automationdirect.com/open-source/home




I'm putting together a shopping list for a huge project at work, and I'm slipping a little bit of this into it for evaluation. Below is the configuration I'm ordering; just a little of everything. I might use it to automate my smoker so it will smoke delicious meats without constant oversight. Once I've had a chance to dabble, I'll come back and post a review. Until then, I just wanted to put the word out there; this is pretty exciting for me; it is one of the only times my professional and personal interests have converged without me forcing them to. Total nerdgasm. I've not been paid or requested to post this, and if it turns out to be crap, I will be sure to rag on it without mercy.

1581576365030.png



I hope I'm not late to the party; this info is a couple weeks old by now. Was anyone else in the know?
 
Last edited:

MaxHeadRoom

Joined Jul 18, 2013
19,747
I have put together simple PLC's using a small micro and Opto22 I/O boards that cover many I/O style Input/Output devices.
The down side is the ability for the ease of 3rd party trouble shooting of the system.
The original main premise of the PLC was that it replaced the hardware ladder style control that used physical relays etc.
It's main strength IMO, is the ability not only to display the Boolean logic on screen in ladder format, but indicate the status of all rung conditions. As was the original requirement by GM to be included in its development
If you can replicate this with Arduino, then I would agree it could be a replacement, otherwise it is not living up to the required original features, which having worked with on floor maintenance personnel is a definite must..IMO.
Max.
 

Thread Starter

strantor

Joined Oct 3, 2010
5,206
I have put together simple PLC's using a small micro and Opto22 I/O boards that cover many I/O style Input/Output devices.
The down side is the ability for the ease of 3rd party trouble shooting of the system.
The original main premise of the PLC was that it replaced the hardware ladder style control that used physical relays etc.
It's main strength IMO, is the ability not only to display the Boolean logic on screen in ladder format, but indicate the status of all rung conditions. As was the original requirement by GM to be included in its development
If you can replicate this with Arduino, then I would agree it could be a replacement, otherwise it is not living up to the required original features, which having worked with on floor maintenance personnel is a definite must..IMO.
Max.
Good point. And with the cost of actual PLCs with said features becoming more affordable, it's not the solid victory I made it out to be. I mean, for $20 more you can get a Click PLC which has I/O already on the processor. By the time you add a single I/O card to the arduino PLC it's already more expensive than the actual PLC. So, I got too excited but this is something I've waited for for a long time. I even built my own PLC years ago from an AT chip programmed in Arduino and a bunch of relays and voltage dividers. It's something that I had an illogically strong desire to happen, and it did, and I'm going to stay excited about it for however long it takes for the "new" to wear off.

On the other hand though, many OEMs go to great lengths to circumvent the very features and purpose of PLCs that you cite. They don't want their machines troubleshot so easily. They want their technicians flying across the pond for exorbitant fees every time a machine misbehaves. They will surely see some value in this.
 

MaxHeadRoom

Joined Jul 18, 2013
19,747
Incidentally all the high end CNC machines that I serviced and retro-fitted also had a co processor for the built in PLC in order to do the more mundane processes on the machine, coolant, spindle, tool changer, part stacker etc.
All these use the standard PLC ladder format display for ease of trouble shooting.
Max.
 

dwdohm

Joined Mar 8, 2019
12
What I don't see mention of in the discussion is operating temperature rating and EMI susceptibility testing.
Has Automation Direct released test information on these for their CPU module?
 

Thread Starter

strantor

Joined Oct 3, 2010
5,206
What I don't see mention of in the discussion is operating temperature rating and EMI susceptibility testing.
Has Automation Direct released test information on these for their CPU module?
Per the datasheet: "Agency Approvals: UL 61010-1 and UL 61010-2-201 File E139594, Canada & USA; CE"
If I'm not mistaken, approvals for those standards are contingent upon demonstration of testing in all the areas that you identified, and more.

They have not uploaded any videos of EMI testing, but vibration and temperature are demonstrated.


 

GetDeviceInfo

Joined Jun 7, 2009
1,666
And at a price point that is sure to attract. I’ll be ordering a similar configuration for evaluation as well. For me, regulatory compliance at the price beats all attempts at home brew.
 

Janis59

Joined Aug 21, 2017
978
Arduino and PLC??? Have an opinion the PLC is sth needed for near the MHz or even faster, but Atmel 328 have no more than 16 MHz with bipolar Vcc or 8 MHz at unipolar 3V3 regime. Thus the sistem is at least order of magnitude too slow for be implemented for any imaginable practical task. The more cooect tablets are 4046 or some cases even better the 567 (LM, XR, XR etc).
 

Janis59

Joined Aug 21, 2017
978
RE PLC in CNC router??? Oh my... Okay, but then why to not realize it on the central desktop machine which is at least 500-fold faster than any ever best Arduino.
 

Thread Starter

strantor

Joined Oct 3, 2010
5,206
The UL standard referred to appears, to me, as defining and testing to acceptable levels of EMI radiation (hazards to personnel in close proximity), but not susceptibility of the equipment to EMI induced from external sources or satisfactory operation at temperature extremes.
This is based upon the UL web site description of the standard at https://standardscatalog.ul.com/standards/en/standard_61010-1_3 .
I'm not sure how you saw enough information in that link to draw that conclusion. It seemed intentionally vague to me. So vague as to basically say nothing whatsoever. I think anyone with an equally strong confirmation bias of the opposition, would likely read it in a way that supports their own position too. I however, have no dog in the fight, and I'm not going to pay to uncover the truth. If you don't like the arduino, don't buy it.
 

AlThePal

Joined Sep 9, 2018
1
Like strantor says this appears to join together my personal (hobby electronics) and professional (automation) interests quite nicely. I'll be interested in having a look, although AutomationDirect don't sell direct in the UK and the sole importer's price for the new CPU is over $91 - an astonishing 86% markup!

Discussions like these often centre on "which is best, a PLC or an 'Arduino'". It's not a case of which is best but rather what are their strengths. Standard industrial stuff should normally be done using a PLC. As Max says it is often more important for it to be understood and supportable. Where the new module shines is in implementing the weird one-offs, for example in joining the industrial world to the web world. This could let you implement stuff that's either not possible with a PLC, or the province of very expensive, high-end PLCs - ironic given how cheap it is.

Interesting times!
 
The whole intellectual property thing kind of has me puzzled. While I'm sure there are tricks and secrets to some programmers that others may not know in the end in order for a PLC program to be of serious value wouldn't you have to replicate the whole machine?

Downtime = lost income... there is no other way to put it. Unless you are planning routine repairs you don't want to wait for a tech to fly across the country to work on a machine.

As far as running a machine from a desktop computer there isn't enough I/O to start with... it would probably take some sort of rack system that would end up being entirely way overkill for running a machine in terms of what it can do instead of what it is doing. If it were feasible it would already be standard practice.

Being Arduino and open source someone out there will soon be scratching out the beginnings of something better programming wise. I've always wanted to play with PLC's, but never really had an idea worth spending the money on one. Maybe this will open the doors for the next best thing.
 

MaxHeadRoom

Joined Jul 18, 2013
19,747
RE PLC in CNC router??? Oh my... Okay, but then why to not realize it on the central desktop machine which is at least 500-fold faster than any ever best Arduino.
Not really sure of the point being made?
The PLC does not really need to be really high speed in an Industrial applications, a typical method used is to scan all input devices/rungs and then update a state table with the OUT results.
The scan time of even high end PLC's is relatively slow by todays standards.
Most of the outside world devices a PLC deals with are relatively slow actions.
In a CNC controller, the whole intention is that the relatively slow machine functions are dealt with the PLC, the H.S. positioning motion is dealt with its own high speed processor.
Max.
 

Thread Starter

strantor

Joined Oct 3, 2010
5,206
The whole intellectual property thing kind of has me puzzled. While I'm sure there are tricks and secrets to some programmers that others may not know in the end in order for a PLC program to be of serious value wouldn't you have to replicate the whole machine?

Downtime = lost income... there is no other way to put it. Unless you are planning routine repairs you don't want to wait for a tech to fly across the country to work on a machine.
Your points are good. Logical. I can't explain to you or anyone, the logic behind source protection for PLCs. I can only tell you that it is most definitely "a thing" and it is rather pervasive in machinery coming out of Europe. There are of course other industries, American ones too, where it is found. For example, talk to any farmer about the problems with John Deere tractors.

But on topic, where I work, we have mostly machinery from a German OEM, and the levels they go to, to lock their customers into perpetual dependence upon their service department, are unbelievable. I've seen the same with Dutch OEMs & Norwegians. I've seen them go as far as to file off MFG part #s and stamp their own over it, so that the customer might never know exactly what parts to purchase locally, only that they must order from the OEM overseas, pay a 100%-1,000% markup, pay the part's weight in gold for emergency shipping, and in many cases pay an overseas technician to fly over and commission/install/program it. Before I started here, there was no professional automation person. My employer took these conditions lying down. They did not know there was an alternative. They relied upon the OEM in many situations to remotely diagnose failure, and propose solutions, which invariably were the most costly solutions conceivable. For example, in the case of a SEW Movimot (gearmotor with integral VFD) with an obviously failed braking resistor, they would prescribe an entire new movimot, to include motor, gearbox, VFD, mechanical brake, brake control module, digital I/O module, encoder, and finally, braking resistor.

This same OEM has removed MFG P/Ns, put block protection on every block in their PLCs, password protected connection to the PLCs, installed custom firmware on many automation components such that they can no longer be upload/downloaded, etc. I have taken massive exception to these practices, and beat my drum loudly all the way up to the president of the company. I have described to him the extent of how he's being taken advantage of, and the fact that there's no logical reason behind it, other than to keep him perpetually at the mercy of this foreign OEM; a permanent, chained source of blood for a vampire. He has taken my words to heart. The time has come to purchase several new production lines, worth several million $$ each, and he has laid down an ultimatum to the OEM. Told them that this is not how we will do business going forward; that if they wish to sell us these machines, their PLCs will be delivered unlocked, with programs delivered on CD, containing all comments, tags, etc., along with the programs for existing machines unlocked as well. They have conceded for the most part. They are still holding out on the programs to a specific machine component of the line, which actually does perform a rather unique process, and may legitimately contain novel IP.

Novel IP, is the only excuse for this in my book. If you've invented something unique, that does what no other machine in the world can, and the key to its function lies in the programming of the PLC, then fine. But that is NOT the case for the vast majority of PLC programs. The vast majority of PLC programs are nothing different than a wiring diagram, which is always considered a deliverable with the machine, anywhere in the world. Nobody would buy a brand new machine without wiring diagrams. And I don't know why we have allowed ourselves to be fleeced out of delivery of PLC programs. I guess it comes down to the enlightenment of the people placing the orders. Most of them probably don't understand.

The justification initially provided for these actions when I inquired with the OEM was that they must maintain 100% control over the PLC programming. They cannot allow unsanctioned modifications to the programming, as that would compromise their ability to provide support for the machinery. They must maintain the ability to predict how the machines will respond in given situations, as they will respond in the way they were programmed, which is the way that they programmed them. And that it is impossible to provide access to view the programming, without providing the ability to change it. They must have thought they were speaking to an idiot. Also, they claimed it as their IP, just like microsoft will not provide the source code for Windows. The difference, I told them, is that I cannot copy a production line onto a thumb drive and distribute it to all my friends. And, I have no need to modify windows to support future manufacturing processes.
 
Last edited:

Thread Starter

strantor

Joined Oct 3, 2010
5,206
Not really sure of the point being made?
The PLC does not really need to be really high speed in an Industrial applications, a typical method used is to scan all input devices/rungs and then update a state table with the OUT results.
The scan time of even high end PLC's is relatively slow by todays standards.
Most of the outside world devices a PLC deals with are relatively slow actions.
In a CNC controller, the whole intention is that the relatively slow machine functions are dealt with the PLC, the H.S. positioning motion is dealt with its own high speed processor.
Max.
+1
PLCs have been running machines since the 70's.
Instead of comparing PLCs to modern desktop computers, they should be compared to the control board of an ice maker. Then the Arduino's competence will be obvious.
 

dwdohm

Joined Mar 8, 2019
12
I'm not sure how you saw enough information in that link to draw that conclusion. It seemed intentionally vague to me. So vague as to basically say nothing whatsoever. I think anyone with an equally strong confirmation bias of the opposition, would likely read it in a way that supports their own position too. I however, have no dog in the fight, and I'm not going to pay to uncover the truth. If you don't like the arduino, don't buy it.
I worked in the UL Santa Clara Labs, and think I have a reasonable perspective of wording of their standards. This standard, like many of those for electrical utilization equipment, is primarily focused on product safety, not performance.
In general terms, we used to explain to clients "This isn't Better Homes and Gardens, we test your appliance to help assure that when used as directed it won't burn down the house or electrocute someone, even if it burns the toast or sounds crummy when playing music".
Over the years since I left, they have begun to test to some performance standards, such as for network cabling, but are generally very clear about incorporation of TIA or IEC standards into their own.
As for dislike of Arduino, or similar microcontrollers, I use them frequently in applications where they are not exposed to extreme temperature or EMI.
PLCs are often in cabinets at elevated temperatures and near devices that generate significant conducted and radiated Electromagnetic Fields. That is why I raised the question about certification against industry recognized performance specifications.
Since I don't know where Automation Direct sources their present PLC offering (they seem to have broken off with Fanuc as their PLC supplier years ago) it was an honest question.
Cheers,
David
 

GetDeviceInfo

Joined Jun 7, 2009
1,666
Arduino and PLC??? Have an opinion the PLC is sth needed for near the MHz or even faster, but Atmel 328 have no more than 16 MHz with bipolar Vcc or 8 MHz at unipolar 3V3 regime. Thus the sistem is at least order of magnitude too slow for be implemented for any imaginable practical task. The more cooect tablets are 4046 or some cases even better the 567 (LM, XR, XR etc).
Why would you even think that a 328 be deployed in such a system? Actually it could be as a I/O module supervisor.
 
While I don't know anyone personally I do know of the whole John Deere issues. The older stuff is stating to gain value I read somewhere. Grandpa had a 3020 when I was a wee one... first tractor with a cab he bought I believe.

I guess forcing dependence on your techs would make business sense to the machine manufacturers, and hiding it in the fine print probably isn't too hard either.

The tech support argument is a good one, but in the end makes for headaches I'm sure.

Since I was just a machine operator in those days I never was involved with real technical things. I just studied everything I could to understand what does what and how it does it so I could either make the right choice and fix the right problem, or at least get in touch with the right person to explain and work out what we could without them having to leave what they are doing.
 
Top