PLC programming languages

[Frank Buss]

I've posted this to a german newsgroup, but not much response so far. Maybe
people using PLC's don't read newsgroups at weekend or not at all

Maybe next year, when I have less client projects, I want to build a small
microprocessor system for simple DAQ tasks, but it should work standalone,
too. Using the PLC standard IEC 61131-3 for the programming language
environment looks like a good idea. I've found a summary in this document:

http://157.158.12.7/kss/dydaktyka/materialy/PC/PLC_IEC61131-3.pdf

My idea is, that even hobby programmers can use this system. The hardware
should be extendable with terminal blocks, like this one:

http://media.digikey.com/photos/Phoenix Photos/1711039.jpg

for I2C, SPI, digital IO and analog IO. For DAQ, transfering PLC programs
and debugging it should use USB.

The most work would be to create the PC side software for a useful
development system. Do you think that IEC 61131-3 standard is a good
starting point for it? The software could compile all 5 defined languages
to byte code, which is then interpreted on the device, and it could even be
extended with other languages, like Basic, for the casual programmer. Time
critical parts could be written in assembler.

I expect at least a year development time. Maybe it could be started as a
open source community project on sourceforge.net.

 

[Jamie]

I've posted this to a german newsgroup, but not much response so far. Maybe
people using PLC's don't read newsgroups at weekend or not at all

Maybe next year, when I have less client projects, I want to build a small
microprocessor system for simple DAQ tasks, but it should work standalone,
too. Using the PLC standard IEC 61131-3 for the programming language
environment looks like a good idea. I've found a summary in this document:

http://157.158.12.7/kss/dydaktyka/materialy/PC/PLC_IEC61131-3.pdf

My idea is, that even hobby programmers can use this system. The hardware
should be extendable with terminal blocks, like this one:

http://media.digikey.com/photos/Phoenix Photos/1711039.jpg

for I2C, SPI, digital IO and analog IO. For DAQ, transfering PLC programs
and debugging it should use USB.

The most work would be to create the PC side software for a useful
development system. Do you think that IEC 61131-3 standard is a good
starting point for it? The software could compile all 5 defined languages
to byte code, which is then interpreted on the device, and it could even be
extended with other languages, like Basic, for the casual programmer. Time
critical parts could be written in assembler.

I expect at least a year development time. Maybe it could be started as a
open source community project on sourceforge.net.

you're a little behind.. Things like this already exist how ever, if you
think you can do better then do so.

I would how ever, brush up on the format use of the PLC ladder that
you have posted in the link you provided. Some parts are not user
friendly much like the OR/AND logic for example.
Most PLC's at the start of a scan, will take a snap shot of all
inputs ,outputs and at the end of the scan will then write back all of
this to
the output area's.

So, putting that in context.

|-------][---+---------------()--|
| |
|-------][---+

The above could be treated as a OR sequence.
|-------][-----][------------()--|

And that as a AND sequence of control.

Now if you want to perform MATH functions, you then
would have a CMD that actually has the "OR" / "AND"
statements on memory etc..
For others that are interested in the subject matter.
here is a short but simple pictation of a PLC.

http://webpages.charter.net/jamie_5"

 

[Frank Buss]

you're a little behind.. Things like this already exist how ever, if you
think you can do better then do so.

Thanks for the hint, looks like there are already some interesting open
source projects, like this one:

http://www.beremiz.org/

But other projects and websites are dead, like PuffinPLC.org or
www.linuxplc.org.

I would how ever, brush up on the format use of the PLC ladder that
you have posted in the link you provided. Some parts are not user
friendly much like the OR/AND logic for example.
Most PLC's at the start of a scan, will take a snap shot of all
inputs ,outputs and at the end of the scan will then write back all of
this to the output area's.

So, putting that in context.

|-------][---+---------------()--|
| |
|-------][---+

The above could be treated as a OR sequence.
|-------][-----][------------()--|

And that as a AND sequence of control.

I didn't use it so far, but doesn't look complicated or not user friendly,
at least not for simple controlling tasks. Where do you think is the
problem?

Now if you want to perform MATH functions, you then
would have a CMD that actually has the "OR" / "AND"
statements on memory etc..

This sounds like a good idea, something like in LabView. Meanwhile I've
bought the standard document and looks like you can define your own
function blocks, with typed input and outputs, which can be combined like
the predefined boolean types. This concept sounds very powerful:
Mathematical calculations, or e.g. a filter, can be implemented in ST,
which then are the building blocks of a schematic like diagram in LD.

For others that are interested in the subject matter.
here is a short but simple pictation of a PLC.

Thanks, nice example.

 

[Jamie]

Jamie wrote:

you're a little behind.. Things like this already exist how ever, if you
think you can do better then do so.

Thanks for the hint, looks like there are already some interesting open
source projects, like this one:

http://www.beremiz.org/

But other projects and websites are dead, like PuffinPLC.org or
www.linuxplc.org.

I would how ever, brush up on the format use of the PLC ladder that
you have posted in the link you provided. Some parts are not user
friendly much like the OR/AND logic for example.
Most PLC's at the start of a scan, will take a snap shot of all
inputs ,outputs and at the end of the scan will then write back all of
this to the output area's.

So, putting that in context.

|-------][---+---------------()--|
| |
|-------][---+

The above could be treated as a OR sequence.
|-------][-----][------------()--|

And that as a AND sequence of control.

I didn't use it so far, but doesn't look complicated or not user friendly,
at least not for simple controlling tasks. Where do you think is the
problem?

Now if you want to perform MATH functions, you then
would have a CMD that actually has the "OR" / "AND"
statements on memory etc..

This sounds like a good idea, something like in LabView. Meanwhile I've
bought the standard document and looks like you can define your own
function blocks, with typed input and outputs, which can be combined like
the predefined boolean types. This concept sounds very powerful:
Mathematical calculations, or e.g. a filter, can be implemented in ST,
which then are the building blocks of a schematic like diagram in LD.

For others that are interested in the subject matter.
here is a short but simple pictation of a PLC.

Thanks, nice example.

each vender of PLC's like to come up with their own little
variety of complexities.
For example in the Keyences families. the advanced logic may look like
this.

|--------[LDA #100]--[STA TM100]------|

and in Alan Bradleys like this.

using a whole block box with full descriptions of the operations.

|------ MOVE -----+
| Source xxxxx |
| DEST xxxxx |
|-----------------+

The operand values could be memory,timers, ADC, DAC or immediate
values etc////


Addressing inputs/outputs are also different among venders///

some use actual numbers that would look like PORT numbers..
0500, 501, 502 etc... for outputs and 0,1,2,3, etc.. for inputs..

they normally designate a max range for input numbers and the values
above that are for outputs..

You can find this in units like Keyence..

In Alan Bradley's they're using I = input, O = output for external
references and normally specify a unit point..
for example
I0:2 means an input on module 0 which is the CPU in this case of
#2 point.
and
O1:0 is OUT on module 1 at point 0.
normally each module device has no more than 16 points of reference
on it.

So you can now see how things can get a little mixed up in the PLC
world when you're trying to support all..

We deal with DirectLogic, Keyence, Omron, Siemans, Alan Bradley of
various generations along with On panel view systems using things like
Wonder Ware etc... So touch base with others at times.

We also deal with micro controller chips etc..

It's gets tough on some days.
http://webpages.charter.net/jamie_5"

 

[Frank Buss]

each vender of PLC's like to come up with their own little
variety of complexities.
For example in the Keyences families. the advanced logic may look like
this.

|--------[LDA #100]--[STA TM100]------|

This doesn't look very useful, because I think in diagrams it is not useful
to use IL code without blocks.

and in Alan Bradleys like this.

using a whole block box with full descriptions of the operations.

|------ MOVE -----+
| Source xxxxx |
| DEST xxxxx |
|-----------------+

The operand values could be memory,timers, ADC, DAC or immediate
values etc////

This looks better.

Addressing inputs/outputs are also different among venders///

some use actual numbers that would look like PORT numbers..
0500, 501, 502 etc... for outputs and 0,1,2,3, etc.. for inputs..

they normally designate a max range for input numbers and the values
above that are for outputs..

You can find this in units like Keyence..

In Alan Bradley's they're using I = input, O = output for external
references and normally specify a unit point..
for example
I0:2 means an input on module 0 which is the CPU in this case of
#2 point.
and
O1:0 is OUT on module 1 at point 0.
normally each module device has no more than 16 points of reference
on it.

But I hope you can define some names for it? Would be much better to use
shutdownReactor=1 instead of guessing that O1:42 is the right port, but it
was assigned to a motor for adding the next fuel rod.