I’m looking for general guidance on prototyping devices and upgrading them to a “production” release.
I’ve been prototyping a battery powered inventory level scanner using a heltec cubecell htcc ab01 V2 and multiple vl53l0x and just using a breadboard with the basic no solder wires. I have some basic CAD design and 3d printing ability and am developing a prototype case to go with the program I am writing, which I basically want to wake the device every 4 hours or so and get a few distance readings from the vl5310xs and send them to my gateway. From there my plan is to use those distances in some exterior program to determine the inventory levels for each part and generate an email to send to our supplier.
When considering potentially making this a professional-ish product, it made me realize I really have no idea where to even start. I’m currently designing a case to be used with the dev board, but should I not be using a dev board like I am now and instead have a custom PCB made?
If I were to stick with just using the dev boards, is it okay to solder wires directly to the pin holes of the abo1 and vl5310xs directly? This seems like the easiest route to me, but yeah probably not what would be considered a “release” model. Honestly just looking for the path of least resistance that will still generate a useable product.
Definitely a bit in over my head, but the alternative is using expensive vendors that honestly don’t have a very good product to begin with (based on my testing)…
Before going any farther with the project, consider how you will maintain the devices. Something will fail, batteries need replacing, software needs upgrading. If you have more than one device, how will you know which one is failing?
Going from a solder-less breadboard, to a soldered prototyping board, to a PCB is a usual way of proceeding.
Today, PCBs are quite inexpensive from China. Many people go through several versions before the design is cast in stone.
As mentioned, there are a lot of things to consider and incorporate than just the circuit on a PCB. Suggest you finalize things as much as you can before attempting your first commercially made board.
Good points for sure, the battery replacement is most of the reason I chose the htcc ab01 as it can get to a current of roughly 3 uA, so with only waking them every 4 hours and with a couple large enough batteries the hope is they will last a couple years or so at a time. The htcc ab01 also comes with the ability to read battery levels, although I need to double check if that’s compatible with getting down to 3 uA. So I would make some sort of view to view that sort of info and set up email alerts for certain battery thresholds. So really instead of just sending the distance the sensor measures it’ll also send battery life, and whatever else I think might be important. It would of course mean very periodically a decent amount of time spent replacing batteries if there are possibly over 100 of these, but could be worth it.
As for failing devices, that’s definitely something I will have to think more about. Thanks for the thoughts
In the real world, a master device would poll each of the remotes and expect an answer. The poll and the answer message both have an identifier. If the remote does not respond to a poll message, then the master must report a failing remote.
No you should start where you are now. Building a prototype using a dev-board is fine. If solder everything you can test it out in the field and see what you run into. While you are doing that you could start on designing a first version of a PCB, and use parts that will allow to solder almost all of it yourself. If the processor is beyond your skill level, you might continue with a devboard or get the processor assembled. Same with other parts. Make about 10 pieces and start testing those and in the meantime you can optimize your code. This may cost you about a 100 bucks but will provide you with all the technical information you need. You can then move onto fully assembled PCBs
What kind of processor are you intending to use ? Are you planning to get the pre-programmed ?
Make sure you make everything Lead-free. That will be your main concern for being able to sell your product.
All of the marketing and such is of course a completely different matter. Warrantee etc. These are not technical issues.
This will change depending on where you live and in what countries you want to sell to.
Can you please tell us those two things?
I have had extensive experience in these matters from when I worked in industry. It is not a cheap nor easy experience, and it can take up to a year or more to complete. You do not strictly speaking need UL certification to sell in the U.S. , but a lot of your customer’s insurance policies might need it.
Otherwise, in the U.S. you do need FCC approval to sell anything, this can cost a lot as this requires testing at an FCC approved test house, in what ever country you are making it in.
I should clarify, when I said “production” release, this is purely for internal use within the business that I work at, not something I would be looking to sell. There are a few companies that offer something similar, but so far through my research I’m not very impressed with the options available, so playing with the idea of making something in-house.
I haven’t gotten that far yet as I only tested the program while plugged in to my computer via USB, I will definitely make sure I test it and get an accurate idea how much is used per wake.
I still need to actually finish a complete and working prototype, I am basically at a prototype of a prototype right now, I just wanted to make this post to point me in the right direction once I am at the point to finalize the design
I’m sure there have been others, it’s not an original idea, but that being said the vendor options for this sort of thing are limited- hence the inspiration to try it myself
It is possible to just keep using devboards, it will make updating the code simple without having to go through the trouble of making a USB to TTL bridge and include that in your design.