Industrial Control Teensy

[MForlong]

Hello everyone,

I know the Teensy was not designed for industrial applications but...

I have been using them in industrial applications for over 6 months now after replacing Arduino megas that were running for two years. The value for the price is absurd (teensy 4.1 was $26 at the time). The Teensies (Teensys?) currently do things like motion control with stepper and servos, PID, RS-232 communication with other devices and open and close pneumatic valves.

The machines I have built have increased productivity by 3-8x on some processes. The machines themselves are designed to fail to safe positions. For example valves close if power is lost ,secondary sensors detect abnormalities etc.

Currently the machines run 5 days a week 8 hours a day with minimum issues (read not teensy fault).

My background is chemical engineering not electrical but I haven been using MCU's since I was a kid in the 90's using basic stamps.

Problem
I am having a hard time convincing myself to move away from the system I am currently using to a more professional grade. The business side of my brain says I need to outsource this problem soon but I am having a lot of fun designing, implementing and improving these systems. Where should I go from here?

I don't have a broad or deep enough knowledge to make an informed decision. I was excited to see the Arduino PRO stuff come out to the market but I am apprehensive.

Should I move on or would making the teensy more industrial through an applications board (24v/optically isolated signals) be worth it?

Thanks
M

 

[BriComp]

Should I move on or would making the teensy more industrial through an applications board (24v/optically isolated signals) be worth it?

There definitively would be a market for something like that. The only problem would be the Marketing required to make it happen.

In the past I designed the electronics and software for the Renishaw Rapid Manufacturing system (RAMTIC).See here.

The idea, originally conceived by David McMurtry (then Chairman and CEO), was that carousels stacked with machining blanks and cutting tools would be wheeled up to a CNC machining centre, plugged in and the GO button pressed. Come back some hours latter and the carousel would now hold the completed machine parts.

I was asked if I could do the control system for it. I must have been having a good day, because after some deliberation, I decided that I could do as good a job as anyone else.

The first system was controlled by a PC, using a solid state drive, controlling the relative hardware using I2C and custom electronics. I had experienced an electronic/software system in the past that had a mind of it's own when production arc welders were operating on the same floor. I managed to overcome the problem with software control. So I determined that any control system MUST be robust with confirmatory feed back of completion of action demanded.

Sorry I have gone on a bit but any application board must be flexible enough so as to be able suit a myriad of application requirements.
Possibly a motherboard to accept the Teensy with hardened IO (hardened input I/O - clamps to supply rails, etc) with user selectable links to different functions, perhaps also plug in devices - 24v I/O, RS232 IO, RS485 I/O etc.

Having had a quick cursory look at the Arduino Pro, I don't see much more than a fancy name, trying to separate the Hobbyist Arduino from the PRO Arduino. But I guess most of it is in the NAME.

I hope this rant helps.

 

[LenSamuelson]

Depends on legal / insurance answers

If your target application has safety implications (a question you and your co-workers would need to answer), then the you can change the Big Decision into a few smaller ones:

• Are you willing to risk having Arduino software involved?
• If using Arduino software, are you inspecting it for correctness?
• Are you planning to inspect PJRC software that your work will use?
• Can you determine whether the PJRC hardware is expected to behave correctly in your industrial environment?
• Are there legal / insurance implications whether you use PJRC Teensy vs other industrial product?
• Could you establish a test plan that manages risk properly?
• Or alternatively as you suggest in your original post, the environment is intrinsically "safe" and the question comes down to validating the resulting control box in a well-defined environment?
• Can you be sure that another system would meet your needs (you know your requirements way better than an external supplier)?

That list is not comprehensive but offers some ways to think about the problem space.

My coworkers and I have been using Teensy hardware for testing and development for many years, both with Arduino and with custom-built-from-ground-up firmware, and have never had any issues. Note however that all hardware that we sell is built using classical methods, fully custom. Which means that we have not faced the decision whether to use Teensy in customer production. Bottom line: Internally, it's a no-brainer the Teensy is a very good choice. Getting lawyers involved though could produce a different answer.