How the search for a modbus TCP/IP motor controller led to a revolution in the design of our next-generation scanner
When I joined the company in September of 2019, our first scanner, from now on G1, was already certified and all the heavy lifting on development was already done. Some of the choices were made having in consideration the time constraints to have the scanner certified and those decisions led to use what was available rather than what we could find that better suited our needs.
With the first iteration of the scanner certified and with all that work behind our backs, we started focusing on the second-generation scanner - from now on it will mention as G2. That is where I came in. I was assigned the task to look for a MODBUS TCP/IP motor controller, so that we could have the possibility to see, via Ethernet connection, the status of the steppers on the scanner. Due to the constraints that we had on the G1, we ended up using some motor controllers that could only be accessed locally. That meant programming, debugging and general status would have to be done through an USB cable.
Going back to the constraints for the G1, it was made in a way that, to access the motor controllers, you had work at least an hour disassemble the scanner just to reach them. Even then, the way their software, general quality and robustness of their components (motor controller and stepper) meant that you could spend a few hours debugging the scanner, just to realise that the stepper was in error mode because the friction on the components was too much for it to start moving.
With that in mind, we set on finding a new motor controller to see if we could change how we debug our equipment. Bear in mind that at this point we hadn’t work with anything MODBUS TCP/IP related, and like in most R&D processes, in paper everything works like it should, until you use it on the field and then nothing works. You must assume that it will not work as intended and be open to the possibility that you will have to look for a plan-b.
I then started to delve into MODBUS TCP/IP motor controllers, also looked at the same brand we had, and looked everywhere for those controllers. I’ve noticed that in automation, MODBUS TCP/IP is not the preferred communication protocol and since we are not using fully fledge industrial automation components, things are harder to come by.
We then came across a brand in Denmark that sells NEMA23 size steppers with an integrated PLC and motor controller as well has MODBUS TCP/IP protocol. It was not something that we were considering at first, but it seemed a good solution and it could open a few possibilities with the PLC for future control of the scanner.
After an initial internal selection, we decided to first contact that company so that we could later decide if we would change only the controller or go for the integrated solution. As it turned out, the sales representative for that company was due to come to Portugal a few weeks later. We were a bit apprehensive for the quote we got, which was about 3 times as much for the same solution we had on the G1, but we were still open to have a look at that system.
A few weeks passed and, we were introduced to that system, and I have got to say that it was worth 3 times more.
Sometimes, it’s not everything about the price. Yes, it’s true that the system was more expensive, but the perks were much more interesting in the long run.
We had a quantity discount, to the point it could even be at par with the current solution since the current supplier did not offer any discount.
We could finally have dedicated and working customer service, at this time we can still talk with that sales representative that came to the office and really have our questions answered - again something we did not have for the current supplier.
The stepper was ready to put on our equipment with pre-made cables, our current supplier had to assemble the stepper with the encoder and send universal cables, that we later had to make an in-house cable adapter to connect the stepper to their cable. Makes perfect sense that a company sells you a stepper and a cable, but that cable does not connect to the stepper.
This new stepper had a much better software interface, that we could use correctly without much issue. We now could define the movement profiles, and change and monitor the parameters of the stepper with a graphical interface that was very intuitive.
Firmware and interface software that was regularly updated, as well as being able to update the stepper firmware without touching the equipment other than connecting it to the Ethernet.
The cable connection on the stepper is made with M23 screw terminal connectors, instead of fragile clip connectors that break very often, making us scrap that unit altogether.
The labour associated with this new solution greatly decreased compared to the G1, making it in the end, very cost-effective.
With all that, we went with it. The result is that we ended up redesigning all the scanners in terms of mechanical components and electronical solutions. We no longer needed a big electronics box for all the controllers, we changed the stepper for the socket gripper movement for one with an integrated controller as well.
Since the stepper is much more efficient than the previous one, it can apply more torque to the vertical movement linear table. That exposed another issue from another supplier. We ended up designing our own linear table for laser movement. That allowed us to gain more accuracy for the G2 scanner and overall reliability and repeatability of scans.
Lessons that we have learned:
Be open to all options, until you test them or have a definitive excluding factor, they can be valid for your solution.
Avoid looking, within reason, to price as a first excluding factor for your project. Sometimes the purchase price is high, but labour reduction and features are also a factor that most times overweight the purchase price.
Leave your comfort zone. Do not be afraid to ask for quotes and samples and ask for discounts for several units as well. You would be surprised by the quantity discounts on some parts. Remember that at some point you could be required to produce several units and those discounts come into consideration when choosing the components.
Embrace change. In R&D you must be ready to change your designs drastically until you have a solid and reliable product. Be wary that your plans might not work outside of the paper and plan for those occasions.
Test, test, test and test again. Have products for testing only and test as much as possible, as soon as possible. There are always things to test and improve, there are always scenarios you did not account for that change how you do things, either assembly or using the product.
Study your bible. Always make sure you know what is expected of your product and what standards it must comply with. The worst thing you can do is redesign your product because you were following the wrong standards or none.
Plan for the future. Be aware of how many units per year you are expected to produce and how, if at all, is the product going to be serviced. These are major factors to consider. You must consider the labour and time needed for producing it. Look for maintenance-free components if you have a sealed or difficult-to-service product.
