Arduino in industry...

[JordanCClark]

Nothing Arduino here that I use. I'm more of a Parallax Prop guy.. Agreed, it's nice to have a standard form factor, but to be honest, even the shield format doesn't fit all Arduino boards anymore. So now you have to pay attention to shield format vs Arduino board you use, or roll your own board with whatever micro and components you wish to use. That's what I do.

Altech makes the track and mounts I use for mounting my boards. The tack is 73mm wide, so that's how I design my boards. Usually readily available from Allied Electronics.

Got a meeting to run to, but afterwards I'll post my RFID stuff for y'all to look at. The antenna thing is unfortunately proprietary, else I'd share that as well.

 

[Peter Nachtwey]

No one is listening. Use the Arduino if you want but heed JesperMP's word about getting the product UL/CSA and CE approved. Generally this is too expensive unless you are making a product like I said at the beginning of this thread.

BTW, we take our motion modules to a lab and tested to ensure they meet the CE requirements. It is like JesperMP said above. You self certify but you better have documentation that prove the tests were done and that you met the requirements.

You guys have no idea how much it costs to get certified and then maintain certification. That is one of the reasons products costs so much. We charge a hefty premium for our controller certified for the oil and gas industry because it isn't a big market and the tests are expensive.

A few years ago Mike Granby of Red Lion and I posted ( ranted ) on this forum about how much UL and CSA costs.
http://www.plctalk.net/qanda/showthread.php?t=53137
In most cases it is just a paper work shuffle because there is no testing being done after a product is certified. UL/CSA is a rip-off.

The only tests that we do that are actual worth a damn are the CE tests and the Class 1 Div 2 and other oil field related type of tests.

Oh, yes, there there is Profibus, ProfiNet, Ethernet/IP, this and that protocol certification so you guys can plug and play. It all costs money.

 

[rpoet]

Peter,
I'm on your side. Since there is no testing with a roll-your-own controller (UL, CE, etc...) I don't want to use that solution.

A commercially available and listed/recognized controller (PLC, motion controller, etc...) is required by most any AHJ if the machine needs to be inspected before it's commissioned, and it's a damned good idea anyway.

Interestingly, I was talking with a friend who runs a UL508A panel shop, and according to UL, you can basically put anything you want behind a class 2 power supply. Being power limited, I guess they figure there's not enough juice available to do any damage. That's no excuse for a homemade solution causing unintended movement of a machine or interference to another device. It would be hard to justify to a court and your insurance company if a non-listed device failed and caused injury or damage.

UL is largely a farce and a way for people to make $$, but their regs do lead to safer panels I think. It prevents people from using any old thing and building dangerous and messy machines (what happens after the panel is out in the field is another thing entirely).


-rpoet

 

[ndzied1]

...and messy machines

-rpoet

UL508 says nothing about whether your panel is messy or not.

There are exactly two driving forces behind UL508A.

1. Prevent Shock Injuries to people
2. Prevent Fires in electrical enclosures

This is what they told me at UL when I took the class there. UL508A doesn't care if your panel works and UL508A doesn't care if your panel looks good. Just that is doesn't create a shock or fire.

It does force you to think about the installer more than has been traditional which is a good thing but there is a lot of chaff that has developed in there over time. However, I will say that you can generally determine why something is in the spec even if it doesn't make sense in practice.

The other thing that is bad in the current UL system is now that UL is a for profit company, there is a vested interest in inspectors nitpicking to find obscure variances from spec and issuing reports that cost the builders money. There also seems to be a lot of variance in inspectors from location to location which leads to confusion and inconsistency.

 

[Peter Nachtwey]

[*]Prevent Shock Injuries to people
[*]Prevent Fires in electrical enclosures

Yes, and because our controller runs off of 24 volts we are safe. If there is a short then smoke is released but there is no fire. So in our case it is a ripoff.

 

[rpoet]

UL508 says nothing about whether your panel is messy or not.

I thought I remember reading something about "workman like manner" but maybe I'm conflating UL508A with the NEC. I suppose even that's subject to interpretation.


-rpoet

 

[JordanCClark]

No one is listening.

Actually we were listening. And even agreeing. The thrust of my post was that as an end user, I have more flexibility than an OEM in what I can do. It's why I said my take on it is a bit different.

Now I'll ask it again: As long as any applicable rules and regulations are, in fact, being followed, then there really should not be an issue, right?

Forgive me if I'm coming across as an ornery son-of-a-gun. I'm just adding some perspective here. I see us as one big somewhat dysfunctional family. But we also put the 'fun' in dysfunctional!

 

[shooter]

Following regulations is not enough, you must put it in a conformity document.

 

[rootboy]

I thought I remember reading something about "workman like manner" but maybe I'm conflating UL508A with the NEC. I suppose even that's subject to interpretation.


-rpoet

Yup, somewhere in section 100 IIRC.

Ah, found it! "National Electrical Code, Section 110-12"


As for using an Arduino, I was just looking over a gondola full of one-off Z80 and Opto 22 boards, programmed in Forth, that used to control our HVAC air houses (that are destined for the dumpster).

We have a brake fluid boiling point tester that uses a 8052 as its processor, and AGVs that who knows what they use.

The DVT booths (Dynamic Vehicle Test) used to use a Motorola chipset on a S100 buss. Our Novatek's use a 8051 IIRC, etc... I could go on and on.

So what do these all have in common? None of them are PLCs, and all of these devices are/were used in industry....


So I don't see any problem whatsoever using an Arduino to control a piece of equipment. It would have to be approved, but that goes without saying.

 

[rootboy]

Nothing Arduino here that I use. I'm more of a Parallax Prop guy.. Agreed, it's nice to have a standard form factor, but to be honest, even the shield format doesn't fit all Arduino boards anymore. So now you have to pay attention to shield format vs Arduino board you use, or roll your own board with whatever micro and components you wish to use. That's what I do.

I just dug into the P'Prop the other day, 8 cores? Wow!

As for the 'duino, the shield thing is overrated, the strength is in the user base IMHO. And it doesn't hurt that they are dirt cheap...

Look at the FriendlyARM boards out there (anything but friendly...) All the horsepower you could ever want and completely unusable.

Got a meeting to run to, but afterwards I'll post my RFID stuff for y'all to look at. The antenna thing is unfortunately proprietary, else I'd share that as well.

I'd love to take a look at it.

 

[Peter Nachtwey]

I just dug into the P'Prop the other day, 8 cores? Wow!

But you have to bit bang all serial ports. Essentially all peripheral devices consume a core. Also, there is no multiply or divide and limited memory.

I would stick to the ARM cores or something more powerful. To me the difference between a $5 part and a $50 part is small. What is more important are the tools and the ability to get the job done easily. Programming time is expensive.

 

[kolyur]

But you have to bit bang all serial ports.

No problem, Parallax provides assembly code for this. You could have multiple serial ports running simultaneously and completely independent.

Essentially all peripheral devices consume a core.

That's the nature of the Propeller. Obviously when you are beginning a project you'd need to look at your requirements and determine if it can be handled with 8 cores. It's certainly not the right device for every application. Personally I'm a big fan of the multi-core setup, especially when compared to the PITA of programming interrupts on single processor devices.

Also, there is no multiply or divide and limited memory.

Granted, 32k of memory is a limitation, but they're working on the Propeller 2 which I think will have 128k.

I developed a commercial product a few years ago which uses the Propeller as its main MCU. This single controller continually reads serial data from two ATD converters, checks the state of the user controls, handles some numerical computation, controls a few relays, and writes serial data to a vacuum fluorescent display. Sure it could have been done with virtually any old microcontroller, but to me the multi core arrangement just makes more sense. And in this case, the difference between a $5 part and a $50 part was a very big deal.

 

[wwlederer]

I came across to Arduino in 2009. The task was to control a pump motor counting short pulses from liquid dispensing sensors. Before, for the same task a Schneider Twido PLC and Mitsubishi FX were tried. No doubt, Arduino offered the best solution and the cheapest as well.
Before we had 2 methods to control something: a special electronic board design or using PLC. Arduino is in the middle. The board design is much easier now. Download the *.sch and *.brd files of Arduino boards and use them as a basis. The time and cost of developing dropped. Even, if You consider to manufacture only few units, it still worth now to think about designing a special board.
And every year Arduino doubled..

 

[wwlederer]

I came across to Arduino in 2009. The task was to control a pump motor counting short pulses from liquid dispensing sensors. Before, for the same task a Schneider Twido PLC and Mitsubishi FX were tried. No doubt, Arduino offered the best solution and the cheapest as well.
Before we had 2 methods to control something: a special electronic board design or using PLC. Arduino is in the middle. The board design is much easier now. Download the *.sch and *.brd files of Arduino boards and use them as a basis. The time and cost of developing dropped. Even, if You consider to manufacture only few units, it still worth now to think about designing a special board.
And every year Arduino doubled..

 

[Peter Nachtwey]

wwlederer, did you get the Arduino board tested? Did it meet the CE specifications?
That is where the real costs are.

Downloading schematics and board files is good but even the board layout requires some 'magic' and experience.

Perhaps the Arduino runs a low enough frequency where radiating noise isn't a problem.