Easy expansion of I/O using Modbus

[DICKEYBIRD]

....Centroid went with it for Hickory.

Name dropper!

[Richards]

DannyB,
I'm not going to convince you or anyone else that Modbus is SUPERIOR to whatever it is that you prefer to use. It is a viable alternative to a lack of I/O when combining multiple machines to act as a unit. The question of how many bits constitute as word has been a matter of debate for decades. My first computer was an 1977 IMSAI, built from a kit. It was an Altair clone, using the same 8080 chip that Altair used. It used bits, nibbles and 8-bit bytes. I built interfaces to use that machine with automated photo printers to service my professional photographers. Next, in 1978, I bought an Alpha-Micro that was a 16-bit computer using six Western Digital chips to make its CPU. I still programmed that machine in assembly language to have tight control of industrial processes. During that time, I wrote a multi-user, multi-tasking operating system for a Z80 so that I could run distributed computing using multiple CPUs to run time sensitive algorithms in the printing machines. Those machines allowed me to print 1,500 color negatives per hour compared to 250 negatives per hour using a standard Kodak-S printer.

I'm not bragging and I'm not talking down to anyone. I'm 70 years old and have been designing process control computers since 1977. I've had to deal with 8-bit word, 16-bit words, 32-bit words and 64-bit words. I've have to deal with big-endian and little-endian data representation. I've had to learn to use all types of intercomputer communication starting with token-ring. I've learned to read the manual and not try to reinvent the wheel everytime a need arises to use off-the-shelf components. Most of all, I've learned not to blame a protocol when the error was caused by programmers who ignored the protocol standards.

Would I use Modbus in sophisticated systems? Of course I would. I use it every day. Millions of gallons of oil run through pipe-lines 24/7, pipe-lines that are totally controlled by Modbus devices, pipelines that cost millions of dollars per day if shutdown. They run. They run reliably. They run because those people who chose the devices and wrote the routines that connect the devices knew how to make Modbus work. Believe me, the owners of those pipelines have tried many other protocols. They stick with Modbus because it works over the hundres of miles of pipeline that they own.

Most protocols take shortcuts so that they can handle a shortcoming of a "standard". The ISO/OSI standard uses seven layers in its ethernet stack. My MACs, my Linux machines, my Windows machines and the various microcontrollers that use ethernet all adhere to that seven layer stack. EtherCAT only uses three of the seven layers in its stack. That's probably a discussion for another time, but when speed comes at the expense of following the ISO/OSI standard, I wonder what others corners were cut.

What CentroidCNC chooses to do is fine with me. I have full confidence in what they have released and I have confidence that whatever they release in the future will do exactly what they say it will do; but, please remember that the experience of many people who choose to share ideas are not an attack on what others may choose to share.

[DannyB]

You don't really seem interested in *actually* listening to what anyone else has to say, or even listening to the feedback that others are giving you about how you are participating, so i'm not going to continue this discussion.
Have a good one!

[Richards]

Thank you DannyB.

[eng199]

It looks like I've upset some people.

Not me. However, I don't think you are looking at it from the same perspective as Centroid or its hardcore supporters. This results in a "not so warm" reaction.

Minimizing support is a large part of Centroid decisions. An impression I get from your concept is that Centroid would be supporting a multitude of Modbus products from other manufacturers without having a new product of its own to pay for support. This may not sound reasonable, but experience shows that Centroid often ends up supporting other manufacturer's products. Step and direction is one example of an open protocol ACORN supports. This is so simple it feels odd calling it a protocol. Even so, Centroid has to borrow or buy dozens of drives from other companies and spend hundreds of hours testing and documenting them.

The alternative is designing our own product. Questions are usually easy to answer because we are familiar with the product. We get some money for it, so we might still be in business to support it. The product is tested and known to work in our system.


Also, you're a little late to the party...the Modbus, CAN, etc. meetings were over a year ago.

[Richards]

Eng199,

Thanks for responding. I kind of live in my own little world where I develop solutions for (mostly perceived) problems. Because I'm getting old and moving around without a cane or sometimes a walker can be a problem, I prefer to have the status of various things on an HMI so I can minimize getting up and down. Before the fall frost, I had a PLC out in the garden to monitor temperature, humidity and hydroponic timing. Other PLCs monitored various electronic projects that I'm working on for the photo industry. Another PLC is my interface to an oil refinery. One PLC acts as a Client to those other PLC Servers. All of the statuses that I need to monitor are right in front of me on a $125 Android tablet. One of the "what if" questions that keeps circling through my mind is: What if I wanted to run multiple CNC routers (or lathes or mills) where each job needed frequent attention, but not constant attention. How could I monitor those machines? How could I control those machines remotely?

The thread about a turret tool changer seemed to mesh closely enough with the extensions that I hoped to see to ask the question about Modbus.

I fully understand the cost of support. A good portion of my oil pipeline check goes towards buying parts and pieces that might help solve other problems. Sometimes I'm lucky and that new part or pieces opens up a new revenue stream. Other times, I just chalk it up the school of hard knocks.

Bottom line. You guys make great stuff and you've always done more than expected to help us customers use your products. I appreciate that. Hopefully, others do too.