Hi all,

My problem is quite longstanding and although I've spoken to many people in Siemens I have yet to have a solution, all this talk of PROFIBUS by Curt has made me look again.

Many of my control systems are built to be incorporated into larger production lines, and are interfaced using a PROFIBUS slave module usually an EM277 or a WAGO unit. I have built a PC with an ISA PROFIBUS master card and used this for a few years just to test that communication is ok. I have several SIMATIC PGs with CP5611 interface, I would like to use these for testing, I just need some _simple_ software to exchange a few bytes with, probably a VB application. Is there anything like this out there? everyone wants to sell me a 400 Euro card with 350 Euro software passworded and keycoded, time limited, serial numbered or token discs. any ideas?

--
Marc Sinclair
http://www.germainesystems.co.uk

 

With siemens CP you can use OPC server software.

With VB you have to put an OPC client control (control integrated in VB when you install the softnet software from SIEMENS) on the seet, give the OPC server name, add the data you need to exchange (for example MD200) and it's finish!

 

Hi Marc

I'm afraid that is the whole point of Profit^hbus. Since you must belong to the cartel, and everything must be blessed, there won't be any "rogue" implementations at low, unseemly prices. And we'll have none of that trust busting free stuff either.

Regards

cww

 

hi,

Sorry but softnet uses those awful yellow token discs which fail over and over again.
--
Marc Sinclair
http://www.germainesystems.co.uk

 

If you have a large enough install base to cover the cost, Siemens does provide a "keyless" authorization scheme. I know that Softnet and S7 authorizations are available in this method. Just prepare to sign a pretty ominous legal thing...

Chris

 

Well, I'm not quite as cynical as cww, but ProfiBus was NOT designed to be simple for a PC app to interface to. It allows token rotation at below 1 millisecond rates - which PC RS-485 serial ports just cannot handle. That is why everyone wants to sell you a hardware card with a co-processor dedicated to fast token-bus management.

I suppose if your PC is the only "Master" and you just connect a few passive slave-only devices you could mimic ProfiBus without token rotation, but that would be useful only in lab situations.

Maybe try ProfiNet on Ethernet... err, no wait; seems you'll want to get iRT-aware switches. Just any old $1200 Cisco switch won't do, so I guess the saga continues anew

Best regards
- Lynn

 

I'll cease being cynical when there is one (1) low buck way to make use of ProfiBus, especially in my own software.

Regards

cww

 

Hey, Curt! What would you consider "low buck" for a Profibus implementation?
--
Steve Myres, PE
Automation Solutions
(480) 813-1145

 

Wait a minute! (responding to Linse's comments)

Who ever uses PROFIBUS (they like it all-caps) in token-passing mode? The strength of PROFIBUS is that it is very easy to use in Master-Slave mode, and that makes it both fast and deterministic. In the Fieldbus committee we had countless hours arguing over the hole in the PROFIBUS token-passing logic which the PROFIBUS guys could ignore because they didn't use it except for master-master communications for backup purposes. For data acquisition and most control purposes, PROFIBUS is an excellent low cost solution. It is really simple since there is no Application Layer in PROFIBUS-DP, making it about as simple as Modbus.

PROFInet is also simple as long as you do not need low jitter timing as for motion control. iRT chips and switches are only needed if you require sub-millisecond timing. Otherwise, the SRT (Soft Real-Time) available with full duplex switched high speed Ethernet is perfectly OK.

Dick Caro
===========================================
Richard H. Caro, CEO
CMC Associates
2 Beth Circle, Acton, MA 01720
Tel: +1.978.635.9449 Mobile: +.978.764.4728
Fax: +1.978.246.1270
E-mail: [email protected]
Web: http://www.CMC.us

 

Well, less than the $1000 for a pc card with a stack. I can do ethernet for nearly no cost. A $100 solution that puts my PC on Profibus with a driver would begin to be interesting. Especially if paired with remote IO that is
reasonably priced. The rules are different for a new design, I can see absolutely no reason to pick profibus unless you are already locked in. There are many much less expensive alternatives. Cost/benefit is awful.

Regards

cww

 

Re: Curt Wuollet's comments - I went to a Siemens product seminar on Profinet (their new Ethernet based protocol) recently and I am now somewhat less sceptical about its chances of success. One of the points that might interest
you is that they envision Profinet replacing Profibus within the next 7 to 12 years. In other words, they will continue to sell and support Profibus, but the writing is on the wall for it and Profinet will be what customers will be expected to use for new projects as soon as enough Profinet field hardware is available. The Interbus trade association has also agreed to support Profinet as their ethernet field bus. "Legacy" networks (such as Profibus) will be supported through transparent proxies.

The Siemens marketing rep stated that the economics of Ethernet are unarguable. He also stated that the far greater size of the IT market means that technical innovation in that market will always outpace the much smaller industrial market. When the biggest Profibus vendor says this, I would suggest not wasting any time on how to do Profibus more cheaply.

With regards to your point on costs, I have looked into the cost of doing two different small projects involving a PC plus a small amount of field I/O. The Profibus card plus OPC server would have been the bulk of the I/O cost, while the actual field I/O would have been the minor part.

 

Hi,

Sorry, but my point is that I already have the CP5611 interface in the PG. All I wanted was some simple software to connect to it.

I use PROFIBUS because my customers want to, In a large factory project, with integrators from many countries, at least the interface is common, yes it is expensive, but it does work, I agree that a simpler open, TCP based comms system will eventually win out, but until then, people need feeding.

-- Marc Sinclair
http://www.germainesystems.co.uk

 

I agree that Ethernet as transport is of glaringly obvious advantage. But, I wish just one of the majors wouldn't burn the midnight oil when they adopt Ethernet to try to make ethernet more proprietary rather than making their protocols more Open. The only exception is Modicon with
Modbus/TCP. They did the obvious thing, the rational thing, and simply encapsulated their proto in TCP/IP. The others have gone to great length to totally eliminate any possible advantage to using Ethernet by ignoring the ubiquitous, standard, Internet Protocol Suite and visiting
their own weird and unique protocols on top of good old Enet. This is incredibly arrogant on their part as it changes nothing. The wire spec never was the problem. Nearly every brick of the Tower of Babel already was based on a known and popular wire spec, so they are changing the
part that wasn't an issue. We have CAN and RS232 and 485 etal. Then they built their own unique and non-standard middle layers on top. They absolutely have to know that what is good about EtherNet _is_ the standards that are run on top, not the lower layers. So Profinet is simply their closed, non-standard, unique layer replacing the truly useful, ubiquitous standards that are inextricably wed with the wire spec, that is universally accepted as Ethernet. And what you end up with doesn't solve any of the problems except perhaps the cost of wire. What a phenominal waste of time and effort.

Regards

cww

 

Hello Marc,

if you own a Siemens CP card, simplest way is Siemens Prodave: it uses the S7ONLINE software layer to communicate with the installed hardware (your CP5611).

Prodave pros:
- relatively cheap (Siemens sells also a "mini" version)
- no annoying keys/serials/dongles
- tiny (installs just a DLL over the "S7online" layer)
- can be used with all interfaces (CP cards or serial/usb converters) supported by "s7online"
- coexists and does not interfere with S7 programming package
- no extra configuration needed, all MPI parameters (address, slot nr, etc) can be set directly by your application executable.

Prodave cons:
- no support to bit-level variables (addressed var must be at least a byte)
- little/big endian byte order conversion must be handled by your application.

About endian issue, additional calls to komfort.dll (an utility DLL shipped with Prodave) provide conversion functions. Personally I suggest to read/write a plain stream of bytes, endian conversion is then easily obtained in PC application by "swap bytes" instruction for 16bit values, by BSWAP assembler instruction for 32bit values (including floating point vars).

If you own a Siemens RS232/MPI converter, have a look at Thomas Hergenhahn's Libnodave, it comes for free.
Try http://sourceforge.net/projects/libnodave/

best regards

Luca Gallina
http://www.runmode.com

 

Curt, it isn't that bad. PROFInet in two of its three forms is just vanilla Ethernet with the PROFInet application layer. You can
still use commercial Ethernet wire, switches, and interface cards. For their iRT version of PROFInet, they really go whole hog and
use a proprietary interface chip to force the time synchronization on the network to get the microsecond isochronous transfers they
think they need for motion control. Somehow, Jetter uses standard Ethernet for motion control. I don't know what SERCOS is using for
their new Ethernet-based specification. Foundation Fieldbus is not intended for motion control, but HSE is based on off-the-shelf
Ethernet components, and claims the ability to synchronize clocks down to 62.5 microseconds for first-out trip analysis.

However, I tend to agree with you - anyone disfiguring Ethernet loses most of the benefits of Ethernet.

Dick Caro
===========================================
Richard H. Caro, CEO
CMC Associates
2 Beth Circle, Acton, MA 01720
Tel: +1.978.635.9449 Mobile: +.978.764.4728
Fax: +1.978.246.1270
E-mail: [email protected]
Web: http://www.CMC.us
Buy my books:
Automation Network Selection http://www.isa.org/networkselect
Wireless Networks for Industrial Automation http://www.isa.org/wirelessnetworks
The Consumer's Guide to Fieldbus Network Equipment for Process Control
http://www.spitzerandboyes.com/Product/fbus.htm
===========================================

 

Hi Dick

Yeah, it isn't them so much specifically, although I'm saddled with a whole bunch of Siemens and the lack of info irritates me. But none of the big automation folks seem to grasp _why_ people want Ethernet. They, for the most part, are expecting that Ethernet will provide what it does everyplace else but the automation market. That is, universal connectivity and interoperability. And you, more than anyone, see the stupidity in throwing away already ubiquitous standardization. But it seemingly doesn't have anything to with practicality or advancement, it smells a lot like bait and switch. Offer the positive concept of
Ethernet, the household word, and deliver the same old baloney. It doesn't seem very respectful of the intelligence of the buyers, but it does seem to be fooling some for now. In the interest of brevity, I won't quote Lincoln about fooling folks.

Regards

cww

 

Actually, it's not just Jetter. I believe there are close to a dozen "motion-over-Ethernet" protocols that use standard hardware. Both ODVA/Rockwell and Modbus-IDA/Schneider Electric believe tight time-sync between nodes (IEEE 1588) allows for scheduled BEHAVIOR without reliance on using start-of-frame timing over Ethernet as the event trigger.

Since people have been whining for "scheduled Ethernet" since the late 1980's (MAP/TOP days), I'm still interested to see how the PNO/Siemens "experiment" of iRT flys (or not).

- LynnL

 

According to the explanation I received about Profinet, they can do motion control over standard Ethernet as long as the network is not heavily loaded. The proprietary real time mode reserves a proportion of bandwidth for real time tasks so that non-real time traffic can saturate the remainder without affecting the reserved channel communications.

Real time scheduling and guarranteed bandwidth (or "Quality of Service" - QOS) are subjects that also interest people working on communications and entertainment applications. It is quite possible that standards will emerge from this field in a few years which will render the special "industrial" versions obsolete.

 

Ever heard about OPTO22, totally hardware driven organisation already embracing and envisioning the use of totally non-proprietary, license-free, open standards protocols. Have a peek at opto22.com, doing things differently!

Regards,

PB
[email protected]