[asch]

I have once implemented such a feature that the program could reconnect without triggerning reset. It is possible to disable the use of DTR if serial is programmed low level on the PC side.

[asch]

When you plug the USB from a PC then you can connect to it using a serial terminal. (Minicom for example in the Linux world but there are plenty alternatives in all OSes.) Your program has to initialize and read the serial interface which is connected to through the USB-serial to the PC.

Arduino IDE also has a simple serial console that can also be used. There are plenty Arduino examples for this. So you don’t even need Minicom or any similar.

When the serial client is connected to the Controllino then it triggers a reboot (similarly as on Arduino – see documentation there if you are interested in the details) which first starts the bootloader but after its timout your program is restarted and you can communicate to it through the serial connection. All settings of the serial connection must be the same on both sides. (Baud rate, stopbit and parity.)

When you have serial connection from PC then you can define any protocol you like. For manual update of password a simple line based communication can be enough.

When the password is inside the program that can be saved into the EEprom part of the chip. This library can be used: https://www.arduino.cc/en/Reference/EEPROM

Other methods for serial connection:

* Serial connection can also be simply automated from scripts using the socat program. Example that connects serial to stdin/stdout: socat /dev/ttyACM0,b9600,raw,echo=0,crnl –

* It is possible to connect from program directly from languages where there is a proper serial library (C, Java, etc). If Data Ready is correctly programmed before connect it is also possible to connect without triggering a restart. Because reset is triggered by an edge on the DTR line which is a control line of serial connection but it is possible to prevent from changing. I have done it once in the past with Arduino but I don’t remember the details exactly.

[asch]

To: Mr. Controllino

I can understand the concerns of Ran because I am in the same boat. Let me clarify this:

* I have programmed 8 bit AVR based products so I am confident in the chip itself that it can work fine if it is put into a good circuit.

* I can handle the programming in C. No issue there.

* Our concern is that we have to decide whether the product is really industrial quality or not.

If I offer an integration solution to a factory with my company based on your hardware I have to be sure that the hardware itself will prove to be a good choice. For example if there is some weakness in the electrical properties that makes the hardware to fail after 2 years of operation and I have installed 100 pieces of the hardware to a production line then I will be in trouble. There will be constant downtime in the factory. And at the end it will be caused by my decision to use this specific hardware.

If it happens with the PLC of a well known big name company then I can shrug and say: guys I have chosen the same PLC that everyone else uses. It is not my fault.

But if it happens to a hardware that is not very well known then if it fails then everyone will blame me for my decision.

For this reason it is very risky to apply a not very well known product in a real world application where high SLA is required but not that high that we build a redundant system. So a single HW failure can cause a few hours of downtime in a factory which costs a huge amount of money. (The replacing cost will be small compare to the downtime cost.)

If Controllino already has references to industrial projects at big factories where it is used then at least we have some facts to support our decision. So it would help us to choose your product.

If you are aiming to be used in the industry as a substitute of PLC based solutions then this question has to be targeted and a comforting answer has to be given. A few example projects where they use it and can be publicly referenced would be the best that I can imagine.

It is not about being a milquetoast: as engineers we have to make the best possible decision when we design a project. Factoring the risks is part of the decision. And the unknown is always higher risk then something widely known.

Disclaimer: I have also just ordered a Controllino that I will use the evaulate the product in a company internal demo application. And if our team likes it and we have a reassuring answer to the reference problem then we will put the same Controllino into an integration project for a factory.