Can I read the dd file without the device

[_Bryan]

Hello,

Is it possible to read the DD file in Configurator 3.1 without connecting to the device? 

The short version of the story......  I have only used the configurator a few times and was walked through the process.  Now I need to figure out how to control a positioner (DVC6000f), I just need to stroke a valve open, closed and position it somewhere between the two.  I am not able to connect to a positioner right now, so I would like to load the DD file into the configurator and look around to see if I can figure out how to give it a position command signal.  Once I am able to stroke the valve I will need to write a simple LabView program to make it easier than the configurator program.

  What is the best way to figure out which block I need to position the valve?  I am hoping when I actually connect to the positioner it will be obvious, but life is usually not that easy.

Bryan

[_Bryan]

Hello,

  I figured out how to create a offline project and load the DD.  As I thought earlier, I do not have a clue at what I am looking for.  I have attached a jpg screenshot with the parameters.  Looking at a Dresser file (not in the screen shot) it looks like something is wrong with the Fisher DD.  I guess I will try to fix it before asking more questions.

Bryan

[_Bryan]

Ok, here is the latest.  The attached file is the latest DD files for the Fisher DVC6000f from the fieldbus foundation web site.  http://www.fieldbus.org/index.php?option=com_mtree&task=viewlink&link_id=1010&ffbstatus=Registered&I...  I tried to open it with Configurator 3.1, and 3.1.1, I even downloaded Configurator 3.2 and tried it in evaluation mode to look at the blocks.  But it does not seem to work, in 3.1 and 3.1.1 if I try to import the files it always gets an error saying that it can not read a file.  If I copy the directory into the 005100 data folder the earlier jpg image is what results.  Can you check the file?  Or do you know why it will not import into the configurator program?

Bryan

[fieldbus]

Hello Bryan,

NI-FBUS 3.1.1 and the earlier version cannot interpret this DD file. There are EDD elements in the transducer block of DVC6000 (R010205). So far, only NI-FBUS 3.2 supports the EDDL.

Please install the NI-FBUS 3.2 evaluation version, and then run Interface Configuration Utility (version 3.2) to import your aforementioned DD/CFF files (I've tested it; these files could be successfully imported).

After the import, please add an offline interface and then run NI-FBUS Configurator 3.2. Now the DVC6000 (R010205) will be listed in the Device Type Tree and you can add it into the project (under the offline link). Please double-click the corresponding block (typically it should be transducer block or function block) to open the Block Configuration Window. If the "Stroke Valve" is a method, you can find it in the menu of the Block Configuration Window; if it is a parameter, please locate this parameter in the list of this window, change its value and then click "Write Changes" ("Write Changes" takes effect only in online mode).

Please let me know if I could do more help for you!

Best Regards,
--Josiane

[_Bryan]

Hello Josiane,

  Thank you for the information, I was able to load the DD and look around the configuration blocks.  Initially engineering only asked about simple control of the valve, today things have changed.  I found out that there are to be many tests performed on the valve including sine wave response, step testing, Hysterisis, linearity, dead band, and some others. 

So my question now is how well can I control the positioner by writing a Vi in LabView?  I can perform the above tests on a positioner with a 4-20mA output signal from my DAQ.  Would it be possible to write the 4-20mA output signal into an input block (instead of an analog output channel) and control the valve in real time?  I could just change the mA signal into whatever units the positioner wants.  Or what would be the best approach to control the valve for testing?

  Is there a number I can reach for support?  It might be easier than email.

Bryan

[Ryan.S]

After you install the FBUS driver, you will have some VIs in LabVIEW to read/write values to the blocks. See the attachment. 

The parameter "tag" is the element that you want to read/write. The format is Blocktag.Struct.Element. The input type can be changed by right-clicking the VI and "select type".

You first need to figure out which block/element is the one you need to write to do the test.  TransducerBlock1(DVC).FINAL_VALUE.VALUE  or FunctionBlock1(AI).OUT.VALUE?   I have no idea....

There are also some examples in \National Instruments\LabVIEW x.x\examples\Ff  for your reference.

FBUS support now is only Email support. You can send the question to fieldbus.support@ni.com, with your phone#. If necessary, we will call you.

[_Bryan]

Hello,

  I finally have some hardware to play with  I was able to modify an example file to send a control signal to the actuator, read the position and chart both.  It works reasonably well, at lest I think it does.  But it brings up a few more questions.

  I notice there is quite a bit of dead time (.5-1 sec approx) between when I change the signal and the actuator starts moving.  What is a reasonable time to expect?  How could I reduce the dead time?

  I have only tried this with the positioner in manual mode.  Will it work the same if it is in automatic? (not enough time to try today)  Would I need to change which block I am feeding my signal into if I am in automatic mode?

  One program I need to write is a stroke speed program.  If the valve is closed, give it an open signal and time how long it takes to reach position.  What would be the most efficient way to do this, with minimal deadtime in the system.  I do not know how fast/slow the fieldbus system is.  Can I do it with my setup?  Or would something else work better?

(Edit)  The dead time I am talking about is in LabView/Fieldbus.  I noticed when I looking at the chart the position signal coming back from the positioner is choppy and not smooth.  Which I think is because of the communication rate of the fieldbus system.  In my Vi I tried from a 10 ms to a 100 ms wait in my while loop with no difference.  As a side note this is not a DVC like I thought I was getting.

Bryan

Message Edited by CCI_Bryan on 04-11-2007 08:39 PM

[_Bryan]

I can't seem to edit my last message again.....

I am running LV 8.2.1 and using a PCMCIA-FBUS/2 interface with the NI power supply and cable.

Bryan

[Ryan.S]

Dear Bryan,

Glad to know the project is moving forward!

According to our test, the average time for a device to read from the bus lies between 50ms to 150ms. It depends on the device's response time. On the other hand, the valve will need some time to react and perform after it get the signal.

Please bear in mind that, the FOUNDATION Fieldbus network runs at 31.25kbps. It will be disappointing if we want a "VERY fast" and "real-time" control.

[_Bryan]

Hello,

  Here is a screenshot of a vi I made from one of the Fieldbus examples.  I found that it takes 200-300ms for each iteration of my loop, so the ramp is about 30 seconds.  I do not expect "real-time" I just figured that it would be better than this.  In the VI pictured I have a 10ms wait, I have tried up to a 500 ms wait and the position started to smooth out.  But if I am trying to get a stroke speed and I am sampling every 500 ms, it is not going to work so well.  For slower tests it should not be that big of a problem. 

  I have the positioner in manual mode, would it make a difference in normal (automatic) mode?  I send my signal to AO.OUT.VALUE and read position from AO.PV.VALUE and it works in manual mode but not norma/automatic.  Do I need to read/write to different blocks for the normal mode?  I am learning field bus control as I go, so I still have many questions on how everything works.

Thanks

Bryan