Answer the question
In order to leave comments, you need to log in
What is the algorithm for switching from an Arduino craft to a series?
Necessary caveat: I'm another "amateur" who wanted to imagine "smart" 2in1 turn signals
and through this discovered the world of addressable RGB tapes and Arduinos. After a couple of days of "Top 100 Arduino projects" videos - the feeling of the omnipotence of this platform and the feeling of " yes, I can make things and sell them ."
And so, before a huge order from the Chinese for all kinds of Arduino stuff (boards with GSM, gas sensors, etc. servos), I looked at Habr. I read similar topics: one , two and comments under them. Although the opinions of those who answered were divided, the ardor diminished, I postponed the order.
Is it right that if you make real "things" for sale, then after testing the prototype on Arduino, you need to make your own board, place only the necessary components on it and program them in a low-level language? With the goal of " less wires, lower power consumption, higher reliability "?
At the same checkout, the question is: is it possible to program microcontrollers in Java ? Previously, I learned this language and often came across "1 billion devices run on Java." Just a piece of iron - everywhere we are talking about C ++.
ps Sorry if the question is chaotic and slightly out of Habr tradition.
Answer the question
In order to leave comments, you need to log in
I will answer as a person who has been making various prototypes for a long time (both on Arduino and PP), but does not sell them.
In general, it depends on the buyer, someone can and will be satisfied with such an alignment.
But in general, Arduino is not about reliability, if everything is on these wires of DuPont 2.54 popping up, then these "turn signals" will fail at the very first bump. And even worse, if all this is connected to a battery, the wiring will close around the protection and there will be broads and a fire.
Designing your own printed circuit board can really be called a certain necessity, and besides, in order to save energy, you really need to remove everything unnecessary. And you need a lot of tests to not only make sure that the device works as it should, but also that it is really safe.
In Arduino it's "harder to shoot yourself in the foot", but if you really want to make things for sale, then you really need to understand all the nuances and errors in the MK (and sometimes there are errors from the manufacturer, see errata), low-level languages \u200b\u200bare not bad with this help. Plus, it may also be that all these things in Arduino that make life easier will “gobble up” all the resources and you will rest, although in fact you won’t even reveal all the capabilities of the MK by half.
You also need to understand circuitry and electronics. Safety and reliability above all!
On the first question - yes, it is highly desirable. Another plus is the ability to fit the board into the desired size and case configuration, the ability to throw out unnecessary parts (do you need a USB-Serial converter? What about a linear power supply? What about a USB connector?) and so on. But you need a pack of power filters (this is a car), protection of inputs and outputs.
On the second question - java drags jre along with it, and it is rather fat, it needs a fat controller that likes to eat electricity deliciously. In addition, the behavior and delays of a C ++ program are quite predictable, in java - additional tricks are needed so that the garbage collector does not suddenly visit you.
You can develop the topic further, noting that in conditions of severely limited resources (how much RAM is there? 8 kilobytes?), a language with manual resource management (C, C ++) is much more predictable than all such automatic and convenient toad.
Well, and one more thing - in all my radio-destroying practice (more than 15 years) I have never bought an arduino - I immediately designed a prototype and made a board (first with a LUT, then I ordered it in production, since it is better and not much longer + they will make it in production right away a dozen boards that are not afraid to screw up during debugging) - this
is orders of magnitude more reliable than picking in a ball of wires. Well, some devices with an analog path (spectrum analyzer, myoactivity analyzer) in the form of a ball simply do not work.
then after testing the prototype on Arduino, you need to make your own board,
is it possible to program microcontrollers in java
Only normal connectors or soldering, then it has the right to life.
In general, all this is already there:
https://www.drive2.ru/o/b/542750551503274174/
https://www.drive2.ru/b/453795387891778971/
https://www.drive2.ru/b/2106319 /
https://www.drive2.ru/b/450863540136314977/
https://www.drive2.ru/b/535129561533252166/
https://www.drive2.ru/c/2505784/
But on the organization of production:
https ://habr.com/ru/company/promwad/blog/211975/
https://habr.com/ru/company/promwad/blog/194878/
https://habr.com/ru/post/476384/
https ://habr.com/ru/post/477802/
https://habr.com/ru/post/472154/
https://habr.com/ru/post/437950/
https://habr.com/ru /post/513816/
https://habr.com/ru/company/antilatency/blog/320204/
The correct answer is written in the question.
Yes. If there is a series, then you need to make your own board and put the controller and all the necessary components there.
With languages, too, everything is NOT difficult. The program can be written in any language that is supported by the machine code compiler.
In the Arduino environment, the C ++ compiler is used. In the development environment from Microchip there is also Assembler (for true Aryans).
These are, as it were, compilers officially supported by the iron manufacturer. In my (subjective) opinion, C++ is great.
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question