I'm working on creating a product for controlling industrial equipment and I'm now confident in the cloud interface, particle device, and coding.
My focus is now turning to creating the circuit and PCB design. My product utilizes relays to control 480VAC motor starter contactors and takes some input data from sensors. I've built the proof of concept on a breadboard and tested functionality in the industrial environment. Now I need to make it work for 5 years.
Problem: I look at similar products and see boards jammed with opticouplers, varisters, ICs, and all sorts of other components. I don't know what I need and what i just don't know enough about to even know I don't know. I know Eagle and other products are good for building circuit layouts but there's the middle step.
Does anyone have a good reference, book, or resource for a crash course in circuit design? I'm needing help figuring out how to ensure my product can survive real world conditions. I realize there are companies that do this and might end up going that way. Before I reach that point and until I prove up a business case, I'd like to put some sweat equity in this myself to learn.
This is the only post I found on circuit design. Most post focus on PCB layouts, but not on determining what components and such are necessary by best practices.
@jonpasc I was in a similar position to the one you describe about 3 years ago. I would suggest that if you havenāt got a background in electronic design and you are serious about your product you should find a contract electronic design organisation to work with. They will help to explain why you need a lot of the components to āprotectā the rather delicate ICs from EMC and surges and help you get a reliable product to market quickly.
In addtion to @armorās comments, this is a good place to remind folks about your Local Safety Codes too.
Itās easy to forget about all the electrical safety requirements beyond your IoT device.
For instance, if your product will be used in the USA and itās responsible for switching motor loads, you should have someone perform a full NEC Code Review.
A H-O-A (hand-off-auto) switch should be included, since you are selling an automated device for starting/stopping a motor (NEC 430.102). A maintenance or service technician needs to be able to quickly and clearly identify how to disable your controller.
Another topic thatās sometimes missed is the additional hardware required to ensure the relays land in the OFF position no matter what failure mode happens internal or external to your controller (typically performed with Safety Interlocks).
Youāll be looking at a significant investment for a full Design and Code Review for an Industrial Controller Project.
You may want to start with an IoT product that monitors existing Controls first ?
In my personal experience, the āCloudā portion is the added value anyway, so you might be able to skip the dangerous stuff.
Seriously, you need to hire a professional Electrical Engineer with plenty of experience, either for you staff, or as a consultant, to perform this work. Especially when working with 480VAC. The future of your company, and the lives of your employees and users may depend on it.
These motor controllers are circuits Iāve been working on since I was 15. I did 2 full summers working solely on 480 VAC systems, so Iām versed enough in the electrical safety side. Iāve seen what it can do when not handled properly.
The motor controllers all use a Hand Auto Off controller, @Rftop. Turn to Hand and Press the starter to turn on, or just switch to auto to turn on. There is no sensor/trigger or relay in the Auto, itās basically on. There are 100,000+ units designed this way around the country. Current systems like Iām looking to create simply break the starter contactor coil current in the Auto-setting, so an operator, turns the relay to the Auto position and that essentially hands control over to the micro controller circuit. Power to the relay and the PCB come from a 480/120VAC transformer. Again, there are probably 10,000+ systems in the states wired up this exact same way right now, and Iāve studied these circuits extensively. Nothing is new, I even did it in a POC device setup using a transformer and 2 relays with optical isolation.
My question is more around what @armor mentioned. I understand the 480 VAC side and motor starter circuit. Iām more interested in protecting the IC from inrush current, surges, and interference.
At this phase in the market, for product #1, thereās not enough value in monitoring to justify creating a product @Rftop. Customers wouldnāt justify the cost. I have another 2nd product that does monitor. Iām working under the assumption that if i can get the more complex one designed, the second monitoring product for use case #2 will work as well, although it faces many of the same problems. Even just monitoring, how do I ensure longevity of a device is a question iām answering.
Thereās a 90% chance I end up working with a consulting firm on this project. All I was asking is if someone had resources that I could get a jump start with. I like to be able to understand the work Iām having done, be able to know why things are done, and to know that I can match why things are done with driving factors. Iād like to learn for myself even where to start on this. So lets move on from lecturing about safety. I understand my use case and environment. What i donāt understand is protecting ICs, building PCB circuits, and bringing a product to market, and budgeting the cost for all of this to decide if itās even worth pursuing.
While it is unfortunate that you feel that way, the comments you received are justified, even with 2 summers of experience.
Please remember Other people will also read this Topic in the future concerning their own projects.
Nobody apologizes for mentioning Human Safety, I hope you understand that.
That's free advice coming from a Professional Engineer and not a stab at you or anyone else.
In general, take a look at Solid State Relays since they employ optical isolation to protect your IC. The SSR drives an AC coil in the contactor (an Inductive Load), so Induction Suppression Cap is also needed. You may also want to investigate Phase Monitors, OV protection, motor savers, and Backspin timers (per application) if you're looking at eventually hiring a Consultant to finalize a turn-key panel.
But if your intention is to basically replace PLC's running typical Ladder logic with a Particle IoT device, that's a much smaller Scope obviously.
If your requirement is to collect some background understanding about what sort of components you need then I would suggest googling circuits for switching relays or 480 VAC. I have learnt the most from asking the engineer I employed why they added this or that component.
Basic learning points I can distill here for you without a block diagram of what you are trying to achieve are:
If switching AC load with a mechanical relay then you need to decide what DC voltage to use (5, 12 or 24) the voltage determines the current (the actuator coil has a resistance).
You need a flywheel diode across the relay coil.
You will need a MOSFET to switch this with a TTL signal from either the Particle MCU or better a GPIO expander - that way you are isolating doubly. Low side switching is better.
You will need a AC DC converter to create the voltage used for powering the relays - and that will need to supply enough current to switch however many relays you will have.
You will likely need a DC DC converter to create the voltage bus for the ICs (either 3.3V or 5V).
The GPIO expander will need pull-down/up resistors on the outputs to ensure that when powered / reset it defaults to OFF.
I would suggest I2C bus control of the GPIO expander. There needs to be shielding of the cable to the controller if off board. The signal needs pull-ups to 3.3v at the MCU end.
Any digital signals into the GPIO expander (from off board) will need a zener diode to limit the voltage (surge) and a small resistor for the current. Opto isolators can also be used but mat not be necessary.
Another learning point is that getting the circuit right generally takes a couple of iterations and requires some prototypes to get the board layout right.
If you are planning to sell a product then you need to know what certifications you need to have and pay a professional to do this for you.
Great post @armor, very detailed.
It reminded me of another possible solution for a simple controller using Particle Devices.
NCD makes a product line for Particle Relay Boards.
Iāve used a $50 board as an add-on to existing Sewage Pumping Station Control Panels to replace the Auto Logic (read the float switches), plus read pump pressures, water level, Cloud Reporting, etc. All the pre-existing protections remained in place, I just replaced the controls for the motor-run-signal when the Station is in Auto Mode. @jonpasc, those relay boards may or may not help for your project, but they have circuit diagrams on the resource pages to get some ideas about what armor has listed.
NCD can also design and produce a product for you. Theyāve done so several times for me.
Thanks for the input, @armor, great things to think about and check. Iāve written the specs that Iāll pass along to NCD and possibly others to quote out some board design.
If I was able to use a standard NCD relay board with a electron in an enclosure, for those of you who have actually created products, would that be considered a product youād release? The initial design is for 30-40 devices to control equipment on my fathers farm, so nothing Iāve got line of sight on selling in the next 6 months. I just want something that has a chance to operate trouble free for years (a could be professional product). If it works very well and I see opportunity, thatās where a business model might come in.
Iāve read on the forums, 2014, about a product checklist, but havenāt actually found it.