Well, my name is Cassiano. I'm on the 4th turn of Computer Engineering and i'm about to start working on a research using FPGA's.
On my research i'm going use FPGA to simulate a AS-Interface chip on a FPGA and other components.

Looking for references on internet, i couldn't find anything specific or related with the Industrial Automation area - using fpga to simulate the ASi chip. I think there's a lack of researchs on this area.
As i'm starting to study this technology, i'd like to know if anyone here have already read something like what i've said above?!?

thanks, Cassiano.

 

Maybe the lack of research is due to a lack of interest. It has been around for nearly 20 years and I don't think it was ever very popular. Certainly as a low cost interface for digital I/O points it seems that an FPGA solution would be very unlikely to improve the economic prospects of this steaming pile. I'd suggest another topic.

__________________
We never have time to do it right; but we always have time to do it over.

 

What do you mean by "simulate" in FPGA. You can use a simulator program to simulate your degins, or you can program an FPGA to perform your design function in real systems.

__________________
You don't need a quadraphonic Blaupunkt -- you need a curve
ball.

 

I was really considering what you said. More than 15, 10 or 5 years and as you said, it was ever very popular.
So I think that I'll start to look for another topic. Now i reaaly see that I need to reasearch a better topic.
Thanks Papabravo.

 

Ciao Cassiano,

If you are looking for an Industrial Automation topic, I'd suggest looking at ways to specify and build a physical layer that will carry power and data, have a data rate of 1 Megabit/second or more, a maximum cable distance of 1000 meters, and support 64 nodes. Now the incremental cost for a single point I/O device must be less than $10.00 or the equivalent amount in euros or lire. You don't still use lire do you? I think not.

__________________
We never have time to do it right; but we always have time to do it over.

 

I'm sorry Papabravo, I forgot to say that I'm from Brazil. So, no...I don't use lire.
Very interesting your suggestion. For now, I don't have many ideas how to work on what you've suggested, but tomorrow I'm going to meet with my professor and I'll talk with him about it and listen his ideas too. After the meeting I'll post what we've decided.
About your idea. Would it be a way to improve the AS Interface or about "creating" a new kind of network?!

 

I would think that a more capable physical layer would form the basis for a new network technology based on an existing MAC and Data Link Layer. There is a space to fill in your location. So what currency is used in Brazil?

__________________
We never have time to do it right; but we always have time to do it over.

 

Well, here in Brazil we use "Real". It was a currency created in 1994(I guess) to control the levels of the inflation. During a time, while its value was really controlled by the Central Bank, it ran with dollar and occured of Real have the same value of Dollar. But later, they decided to let the currency 'walk by it own', whithout the 'protection' of the governament. Today they prefer to keep Real's value, lower than Dollar one. To buy a dollar we need 1,72 Reais(plural). It is good to export/import market.

So I got some ideas while I was thinking about what you've suggested and I want yout opinion.
I've been thinking about to research a new way of communication between the slaves and the master.
Using a sensor as example:

SENSOR -------> MASTER -------> SOFTWARE

- the sensor : every variation on its inputs, will generate a clock to send a sequence of bits to the Master analyses. Depending the received data, the master will send a answer to the sensor activate its outputs and an update to the visual software. For example, by a circuit to convert parallel/serial data and another one to send it.

- the master : receive every little information and work as much as possible on it and send answers. With this I think that the development of sensors(at first) would be too much cheaper than today.

- the software : can be used to show to the user what is going on the network. Can be used to program the logic of the network.

!!!....."On this example, of course, the factory plant must be pre-defined.".....!!!

What do you think about it?!?

 

In the case of a single point sensor(input) we like to think of what is known as a Change of State or Cyclic Connection. In a client-server model the Master is the client endpoint and the sensor is the server endpoint. In the general case, this connection "produces" whenever there is a change on the input, AND it produces periodically to indicate that it is still there. This can be important in a factory environment if there is a possibility of a fork-lift severing the network cable. We call these cyclic messages, "heartbeat". Of course, as options on this connection, it could be Cyclic Only or COS only.

The COS/Cyclic connection can be acknowledged or unacknowledged with the obvious increase in network bandwidth requirements for those acknowledgements.

On a device with outputs the COS/Cyclic connection has fewer obvious benefits, except in the case of unacknowledged production and consumption over an unreliable link. In that case the periodic component raises the probability that at least one message will get through given enough tries.

To show what is happening on the network is only the beginning, modifying the network parmaeters to do adaptive load management is the interesting application.

This is fascinating stuff and, I've spent nearly 30 years on it. I hope you and your advisor can find a suitable project.

__________________
We never have time to do it right; but we always have time to do it over.

 

Papabravo, I'm sorry man...Reading my last post I realized that what I wrote is a bit different of what I desired.
I had done a 'picture' showing better, bt it just disappeared. Sorry, the description was as a normal network!!!
But I got what you said above and again, thanks!!! Now I'm studing what you've told me, writing somethings and I'll show to my adviser.
When we get a idea, I'll write you and see what you think about...

thanks, Cassiano.