Go Down

Topic: Advice for transitioning from Arduino prototype to consumer market design (Read 543 times) previous topic - next topic

bengine45

I would like to preface this thread by stating that I am a complete novice in the domain of production level design and manufacturing. I have not yet before ventured beyond the phase of prototype design in any of my hobbyist endeavors. I am an aerospace research engineer, and my experience with hardware development and circuitry design is limited. I plan to soon attempt to building a production prototype from what I currently have and am looking for advice and insight to help me in getting started.

For the past 15 months I've iteratively worked on a product concept and have finally finished up a nice physical working prototype. I feel that I will soon be ready to evolve my Arduino-based prototype into a production-level prototype that is safety certified and "green lighted" for consumer market entry and retail. My product idea identifies as an in-home appliance (heavy and bulky) and it will be used daily by consumers in either their living rooms, kitchens, bedrooms, etc.

The functionality is rather simplistic: motor actuation via Bluetooth control with LCD displays for various visual feedback indicators. My design includes:
  • 360W power supply
  • Arduino Mega
  • 2.8A NEMA 23 Stepper
  • TB6560 Motor Driver
  • Bluetooth (BT receiver fob with a Nano)
  • 3 LCD displays

Though I feel that this mix of components is not very complex, I am constantly racking my brain with figuring out how I am going to remake this system to meet production grade certifications and requirements. Currently I'm weighing 2 potential approaches: to either 1) create a custom PCB design that integrates all of the above components (yikes!) or 2) invest in an industrial/professional grade controller as well as compatible versions of the above components. It's probably obvious by this point in reading my post that this is where I'm at a total loss with my limited amount of experience in this knowledge domain.

Given the numerous certifications and safety requirements I will likely need, how should I go about creating a production prototype? How would someone in my position, with limited experience, begin with this daunting phase of design? As far as I know, there isn't an online handbook for newbies that discusses in detail of the process for unique product designs. I welcome any guidance and ideas for how to go about the process. Any bit of information will be very helpful; books, videos, blogs or tutorials, targeted communities/forums, etc.

Paul_KD7HB

Before you get too far with your brain wracking, spend some time minimizing the count of different components. If you can use the same component in various places, including electronic components,you will save a bundle in manufacturing costs. If you have a bill of material prepared, try to reduce the number of individual component entries.

Paul

bengine45

Before you get too far with your brain wracking, spend some time minimizing the count of different components. 
Thanks, Paul. At best, I can get by with eliminating an LCD screen and try alternative approaches to that portion. I've spent the past few months minimizing hardware and optimizing software and machining simpler parts needed. I'm satisfied with this final prototype version enough to replicate it at near-production quality.

lastchancename

  • Minimal point-to-point wiring - ease of assembly & reliability
  • Designed in fault-tolerance and fail-safe strategies
  • Testing, more testing
  • Complete the testing
  • Good documentation - so the user can figure it out without you there.


Remember the smallest problem - if they have to return the unit or call you on site will cost you at least $20 per repair (time., parts & skills), maybe $200 or more including travel or shipping.

OK, good news, you've sold 1000 units at $100 nett profit each !
Great $100K total profit !
: : : :
Uh-oh, one of those unexpected problems has surfaced.
1000 x $100 average to make good. Oh sh|t, there goes your $100K

A lot of work to get nothing back - except a dented reputation.
Be careful.
Experienced responders have a nose for laziness, (they were beginners once)... Sure, there are trolls, chest-beaters, and pretenders - but the help you'll get here is about as good as it gets - if you try to help youself!.

Paul_KD7HB

Your problems will probably get more expensive in the long run.
Is your power supply UL certified and is it plugged into your device or is it part of your device. If part of your device, the the WHOLE device must be UL certified.
I had a customer that needed UL certification. We had to do a production run of 400 devices with serial numbers. Then the lab doing the certification selected a random number of serial numbers for testing.
Since your device also includes an oscillator with 400kHz or more frequency, you will also need FCC certification by a certified lab. And if you make future software or hardware changes the FCC certification must be repeated.
Welcome to the real world!
Paul

bengine45

Uh-oh, one of those unexpected problems has surfaced.
1000 x $100 average to make good. Oh sh|t, there goes your $100K

A lot of work to get nothing back - except a dented reputation.
Be careful.
Thanks for the input. While I do appreciate your scenario, you seem to focus heavily on the possibility of a devastating design flaw that would require a call back and result in a ruined product batch, if I'm not mistaken? This process will take time. I understand that and choose to proceed with caution over haste. I'd be more worried at this point over a potential future lawsuit than a royal screw up in design/manufacturing, though I understand that either can of course happen.

bengine45

Your problems will probably get more expensive in the long run.
Is your power supply UL certified and is it plugged into your device or is it part of your device. If part of your device, the the WHOLE device must be UL certified.
I had a customer that needed UL certification. We had to do a production run of 400 devices with serial numbers. Then the lab doing the certification selected a random number of serial numbers for testing.
Since your device also includes an oscillator with 400kHz or more frequency, you will also need FCC certification by a certified lab. And if you make future software or hardware changes the FCC certification must be repeated.
Welcome to the real world!
Paul
It's UL, however; it's not a wall wart or plug and play. There are wired leads for the AC and DC terminals. I have been planning this entire time to acquire full device certification under UL and FCC with CE and CSA after becoming established. I've been recommended to meet ISO and NFPA standards as well. As you stated, all of these large expenses will be need to be accounted for in the cost. I would classify my idea as a luxury novelty, so I do have some room to work with in total expenses vs. functional value.

I could really use more insight and help on the transitioning from hobbyist tier electronics to production grade components for safety and reliability. For example, would replacing an Arduino Mega with a low tier Omron PLC ($250+) be overkill for a consumer-available home application with high ampere, NEMA motor with high torque, bluetooth, etc? Similarly, would I need to replace my current components with ones that are UL recognized in order to achieve full UL certification? I'm lost as to what minimum reqs. and standards are needed for my application in order to be listed as consumer safe and given the thumbs-up for retail.

lastchancename

DESIGN FLAW - Not at all,

If your user inputs a value that locks up the code, or it suffers from heat stroke, or maybe once in two years, the power supply can't meet the demand of exceptional circumstances (two fans, and a servo while running in reverse on a Thursday !!).  Maybe a power spike loses all your retained data. 
If we could predict it, the issue would have been tested and corrected before shipping !

Depending on the product, any of these may require a return, or support visit at a minimum $20-30 per hour plus parts and travel/shipping.

No replacement unit, or design flaw, just an unexpected operational problem that could destroy your reliability.

For example, I have over 100 units located more than 50 km from anything like support or a person to press reset....  Without a watchdog, or remote capabilities, if anything had gone significantly wrong... someone would need to drive out there with parts, diagnose - or simply restart the device.  Maybe replace the unit. (One was kicked by cows, another had a relay failure, yet another was hit directly by lightning - not product failures thank goodness !, but all of them were more than 200km away from me.)
Experienced responders have a nose for laziness, (they were beginners once)... Sure, there are trolls, chest-beaters, and pretenders - but the help you'll get here is about as good as it gets - if you try to help youself!.

Paul_KD7HB

UL certification could care less if the components in your project are UL certified or not. It is YOUR product that must be lab certified.
Seems like it is time for you to pay personal visits to businesses that specialize in manufacturing products for their customers, like my former company did. Ask them all your questions and options you are considering to get professional help and it will also give you a feel for their interest in building your product.
Good luck, Paul

bengine45

DESIGN FLAW - Not at all,
I understand better now what you're getting at. Thanks for the additional context. Heat, component wear, surge, hysteresis, etc. I am trying to pin point at a component level in data sheets and component testing. Though combined in my design, for example, I know for a fact that it, as it is with the current components listed above, isn't suitable or reliable to handle power spikes in the long term. It needs a proper redesign for a production grade prototype. I am not sure of how to accomplish this based on my project requirements and the type of components (high amp/torque motor, bluetooth etc.) needed in my design. whether I need to create a PCB design with high quality production grade components (a daunting task considering the TB6560 intricacy and my lack of experience in hardware design) or purchase industrial/profession PLC with compatible professional grade components.


UL certification could care less if the components in your project are UL certified or not. It is YOUR product that must be lab certified.
Perfect. I was unaware UL certification is only concerned with the final product.

Seems like it is time for you to pay personal visits to businesses that specialize in manufacturing products for their customers, like my former company did.
I have a family member who is the VP of a small local manufacturing and engineering services firm. Perhaps my best immediate option?

Paul_KD7HB

I have a family member who is the VP of a small local manufacturing and engineering services firm. Perhaps my best immediate option?
Start there! Ask for names of other companies to visit. Just don't become a pest.
Paul

Go Up