DataSocket PSP data buffering

Similar Questions

• Binds a shared variable to a NOR-PSP data object does not work

  Hello

  I want to share data between a RT-target and one or several hosts (LV 8.6.1). Shared network variables are deployed to the target RT.  According to NOR to access shared another project or host variables must be done by defining a variable shared on the host computer and the aliasing they NEITHER-PSP data object on the target.

  

  I did and the shared host variable was an error (0x8BBB0011) during execution.

  

  Next I have an alias to a shared variable, deployed on the host to another project. It worked.

  Another thing I tried was to bind the variable of the RT-target to a display element:

  

  This works! And as you can see that the path of the NOR-PSP data object is exactly the same thing! So, what is the difference between linking a data object to a shared variable and a display element?

  Is there a bug in the EVS, or am I missing something here?

  Home project:

  

  The VI editor

  

  Hope someone has an answer.

  Concerning

  Matthias Quade

  Dear Mr. Quade,

  Thanks for posting on the Forum of National Instruments. There is a known issue with the path of the linked with LabVIEW 8.6.1 variable

  Please download the patch for LabVIEW 8.6.1, it should solve your problem:

  http://Joule.NI.com/nidu/CDs/view/p/ID/1255/lang/de

  Best regards from Munich

• cache of data buffers have cancellation information...?

  data in sga buffer has cancellation information...?

  If your transaction is one that makes several changes, these changes will be on the same block.
  See my answer in the other thread I showed you posted: March 16, 2012 10:17 in response to: 920092
  REDO the INSIDE info

• Data buffers Cache vs. data Retreival

  Hello all-
  
  I want to know how the increase / decrease in the effects of data cache, the size of the size of the buffer for data retreival?
  
  I was going through the SER60 and he said:
  
  * "When you retrieve data in the Essbase Add-in for Excel worksheet or use the report generator to retrieve data, Essbase uses the extraction buffer to optimize recovery" * "
  
  There is an interdependence between cache data and memory data retreival buffer size?
  
  Also in SER60 I found in the section "activating dynamic recovery buffer Sizing" as follows:
  
  If a block of very large database and recoveries include a large percentage of cells in each block through several blocks, consider setting the option VLBREPORT in the essbase.cfg of Essbase configuration file.
  
  Someone at - he used this command in the past and can made me know the advantages and disadvantages of it?
  
  Thank you!

  Important notes around the use of VLBREPORT

  The parameter does not apply to databases of global storage.
  The VLBREPORT configuration setting applies only to databases that contain no members dynamic Calc, attribute or dynamic time series.

  Also the buffer cache and the recovery of the buffer are completely separated, because as far as I know

  See you soon

  John
  http://John-Goodwin.blogspot.com/

• Add data in the header of a .wav file

  Hello guys,.

  I seek your help for a small problem that I can't spend during my internship.

  Here's what I want to do: I acquire signals that I write a .wav file. What I would do then is to reopen this .wav file and add data in the header, such as the sensitivity of the microphone for example. I am doing this is because it's one of our program of works, it can resume the pressure vs time signal using this information. I already have the hex code I need to add, but I was not able to determine if such a thing is possible or not after a day of research ideas would be greatly appreciated. I wonder if Matlab might do the trick but I prefer to stay with the 'normal '... screws

  Thank you very much for your help.

  To insert data into the middle of a file, you will need to rewrite the entire file after that point, as Gerd said in the linked post.  I would do the following:

  1. Search for the location you want to put your new piece
  2. Read 65 000 bytes from forward and store in a local cache (shift registers work well for this, or just a wire at this point).
  3. Write your new piece, taking note of how long it is
  4. Give your data buffered less the length of your new song
  5. Read a 65 000 other bytes and put it in a new buffer
  6. Write the end of the old buffer and a part of the new buffer until you get to 65 000 bytes
  7. Repeat 4 to 6 until you reach the end of the file
  8. Update the offsets in the header to match the new locations of piece, since you have moved things

  You can probably do it with both data pads in a pair of registers at offset in a loop.  The 65 000 bytes is chosen for best performance.  You can use other sizes, but your speed can suffer.  Let us know if you encounter any problems.

• Producer/consumer or master/slave, acquisition, control, analysis and recording of data

  Hi all

  I have NEITHER cDaq-9188 and its modules NI 9214 (thermocouple), NI 9203 (HAVE current) and NI 9263 (AO tension). I need to measure the temperature and pressure, and according to these data, I need a frequency converter to drive a compressor control. I've developed acquisition, analysis and recording program by the use of the state machine, the design of producer/consumer model and I have no doubt about their. They all work very well. Now, I need to add a control program in their. The problem is that I need to check the compressor with feedback from temperature data permanently, but I do not necessarily all the data acquired and permanently from the results of the analyses. For example, I need to record data once in 30 seconds. Which model desing should I use for this task?

  Edit: I was not able to find examples of advanced program relevant to the design of master-slave model. If you know where to find them, could you write in this thread please?

  The declarant buffer not data.  Do only the queues.  A new notification replaces the old value if she is ever read or not.  You would be loss of data with the notifiers if the loop of consumers with a notification utility cannot keep up with the data that they fade.

  What I show all the data buffers through the queue in the loop of the producer.  The queue of the consumer removes all data and do stuff with him (such as logging) if the elapsed time is over.  What is the problem with that? That is what for (imstuck) described in his message.

  The other alternative is that you put the timer into the loop of the producer.  Have the Boolean value TimehasElapsed drive a deal structure this place only the data if the time is over.  In this case, only some data gets queued, and consumer loop removes all data as it gets and it opens a session.

• read into the buffer variable published network shared

  

  I use a shared variable that is published in the network buffer to write data to an RT target and read data buffered by the host application.

  I followed a different behavior regarding data buffering using Variable vs reading read the shared Variable node.

  There is no problem the shared variable node read but using the results of reading variable function of loss of data, such as access without alkaline reserve. I can read only some of my data values.

  I see no difference in access in the buffer in LV2009 help.

  against

  Any help is appreciated

  Christian

  Hi Christian,

  The behavior you describe is because the client side PSP buffering is not applied to the Variable API shared in LabVIEW 2009.  It was a design decision, but eventually, we hope to include this feature in a future version of LabVIEW.  This bug here as you pointed out is the lack of documentation for it.  Corrective Action Request # 202159 has been filed for this issue.

  Let me know if you need assistance working around this issue.

  See you soon,.

• LabWindows CVI RS-232 is not fast enough

  Hello

  Is it possible to write a C program which takes RS-232 data buffers and prints it on the screen?

  I ask this question because I write a program in LabWindows CVI and the data is not correct. I used SimpleTerm Gold to see what data are supposed to look like, but CVI misses some. I believe it is not fast enough and therefore lack of data.

  A C program (do not use the functions or reminders CVI) would be possible and better?

  Is it possible to make the CVI to work faster? I changed the environmental policy of the previous "do not sleep". There are two ComRdTerm because we're not getting all the other packages.

  Rate 115200 baud, no parity, 8 data bits and 1 stop bit. Most of the data I receive are 313 bytes and the smallest is 17.

  Any help would be MUCH appreciated as I am completely puzzled.
  Thank you

  Code is below.

  OpenComConfig (com, "COM9", 115 200 Baud, 8, 1, 0, 1024, 1024); / /< opens="" serial="" ports="">
  SetComTime (com, 0); / /< sets="" serial="" port="" to="" never="" time="" out="">
  T0 = clock(); / /< calculate="" how="" fast="" comrdterm="" is="">
  bytes_read0 = ComRdTerm (com, & buffer0 [0], 314, 10); / / read com and store it in the buffer
  sprintf (p_check0,"%.4s" & buffer0 [0]);
  T0 = clock() - t0;
  Double time_taken0 = ((double) t0) / CLOCKS_PER_SEC; / / in seconds
  printf("%d\t%f\t%s\n",bytes_read0,time_taken0,&buffer0[0]);
  Memset (buffer0, 0, strlen (buffer0)); / / Clears the buffer

  T1 = clock();
  bytes_read1 = ComRdTerm (com, & buf1 [0], 314, 10);
  sprintf (p_check1,"%.4s" & buffer1 [0]);
  T1 = clock() - t1;
  Double time_taken1 = ((double) t1) / CLOCKS_PER_SEC;
  printf("%d\t%f\t%s\n",bytes_read1,time_taken1,&Buffer1[0]);
  Memset (buf1, 0, strlen (buffer1));

  Thanks for all the help guys. I finally found the solution. The program was running too slow because I kept opening and configure ports (not shown in the code) I simply kept open ports and it works fine.

  I appreciate all the comments and insight. Each of them will definitely help me still to develop with CVI.

  Kind regards
  Keegan.R

• Questions about serial port read and write at the same time

  Hi I create a user interface for the communication serial port, where there are essentially 2 front panels, where the user enters commands one and the other where the prints of UART is delivered. I thought initially using a state machine but the reading and writing may be independent sometimes and so I can't rely on States. I searched a bit on the forum and he left me even more confused. Help, please.

  (1) in a thread that sessions visa duplicated has been used for writing and reading at the same time, is it recommendable? How will this affect performance?

  (2) essentially when the vi is reading data are it must constantly view as well, however, someone said that it takes too much memory to use shift registers, so how do I go about this? If using a State in queue after the loop of reading it affects the playback loop and be sequential?

  In addition anyway is to move the cursor to the latest data from the indicator

  (3) for the control of the user input, assumes that the user has entered an order in the control and press ENTER, then writing visa is launched, but if it comes in another string and press enter then write must be called again... is - it possible? will detect the previous commands in the control of compensation?

  (4) according to my understand the expectation for the event do not monopolize resources and writing can go in parallel, am I right?

  Thank you. I have attached a very basic vi which took me to the point, but I want to make it more robust. Please help especially in the part of the user interface.

  su_a,

  (1) you can have only one session to a port. Several UART can handle full duplex so performance is not affected.  At flow rates of high data and large amounts of data, buffering and latencies of BONE can become a problem.

  (2) who told you that shift registers using too much memory? Shift registers are usually the best way to transfer data from one iteration to another. String concatenation inside a loop (registry to offset or not) causes the chain to develop and may require re-allocation of memory.  Your VI never clears the string so its cold length become very large.

  Generally, you do not have an active cursor on an indicator.  If you want to always display the most recently received characters and turn on the vertical scroll bar use a property node to keep scrolls to the bottom.  This can be annoying for users if they attempt to manually move the scroll bar and find that the program continues to move it back automatically.

  (3) if the user has changed the value in the chain of command, when he hits enter the modified value event fires.  Simply strike brace does not change the value and does not trigger the event.  Not control need to be erased, but the value that he has to change.  If you want to send the same command again, have a button send a command may be a better choice.

  (4) write is a case of the event.  It is not in parallel with anything.  The structure of the event do not monopolize resources.  The other loop will run while it waits.

  The event loop will not stop when you press the STOP button. Probably it wll take two command: change events of value after JUDGMENT before any loop stops.  Replace the Timeout event (which never expires) with a STOP: value change event and a real wire of this judgment to the Terminal endpoint.  Remove the local variable.  Make mechanical locking when released.

  Lynn

• Delay a table

  Hello

  I have a table 29 x 41 that is the output of a datasocket. The datasocket collects data from a wavefront sensor and each number of the interval is a measure of a microlense (there are 29 x 41 microlenses). The problem is that when I change the settings of the Wavefront sensor to use only a part of the microlenses (exodus 29 x 29) values that correspond to unused lenses continue to be their last measured values and I need the values of unused lenses from scratch.

  In other words, I have a table of 29 x 41 with certain values that are constantly evolving and some that are stable. I need to change the ones that are stable at zero.

  I thought that I could remove the table with this same table but 1 second later. Any ideas how to do this or other suggestions?

  Thank you

  Max

  You can use a shift register or a feedback node so that you will keep the values in the previous table and can compare it.

• Control GPIB SR830 lock-in: read/write problem

  Hello

  I use Labview 2009 SP1 and Visa drivers to control a function 6221 Keithley generator and a lock-in SR830 amp. Basically my program sets the generator to a certain frequency and amplitude, sets things on the lock-in, asks a few values, then reads the data from the lock-in for a minute or two. It is then repeated on several different frequencies. It works well, but it hangs at random time at time, after doing several frequencies with no problems whatsoever. Today, I discovered the spy tool OR wonder, so I was able to locate the bug! Here is an exit of last orders before the bug (there nothing else after the line 63).

  

  As you can see on line 53 I query frequency lock-in. However the response (5623,46 Hz) comes later, in line 57, which has degraded the program. Following this section of the program, I send some more orders for the lock-in, which are ignored, and then the program enters a while loop which is supposed to acquire a number of data points. It works on infinitely because there is no data to acquire, and that is how I noticed this bug at the start.

  So my questions:

  (1) why this happens to you, and how can I avoid it? I have also contacted the manufacturer of the SR830, but I have a lot of faith in big people on this forum!

  (2) after the line 63, in my program there are several other commands that are supposed to be sent to the lock-in, and then the while loop. Orders are not shipped, but Labview must pass by these to the while loop (which she can't seem to). So what Labview does exactly and how/why does it go through the commands of writing without sending the data?

  Thank you!

  Hi Albert

  I think I have it solved. I use the 'FAST2' mode to transfer data from the lock to the PC. This example uses buffers out of the lock-in to transfer data, IE the same buffer used to pass error messages, or replies to queries.

  Between the moment I read the last set of data and stop the transfer of data, the lock-in has the time to write a little more data in its output buffer. This reading of the error mess, but only has a negative effect when the LIA status byte is set to 8 (ie the reference unlock). When this happens, it shifts all the query error responses and the first 'LIAS?"reads the answer to the query"TRE"and 'FREQ?', read the answer to the second query"NE?". Unlock event occurs when the function generator changes its frequency, but can't not every time, which is what made it seem random.

  So, this is easily solved by just read everything that is left on the pad before proceeding. Strangely, it works that when it is followed by the 'ABORT' driver, which basically just sends the command "rest." This erases the data buffers, but from my understanding of what is written in the manual, they are not involved in the mode of FAST2.

  

• Generation of C++ with LabVIEW DLL error

  Hi, I am a student at the UC Irvine, learn how to use LabVIEW.  I'm working on a practice project for which I use a PID system to control the temperature of an object.  Things are starting to be a little awkward with the control statements nested, loops, etc., so I thought I have exported a LabVIEW DLL with all the difficult to encode things like interfacing with my instruments and tracing of data or something else and do the calculation and follow-up of the commands in C++.

  The DLL file I made in LabVIEW exploded without a hitch, as far as I can tell, but when it a link to my C++ test program (which at this point is just including the .h file of the LabVIEW for the DLL and an empty main() function), I get a lot of build errors.  I'm using Dev - C++ that uses g ++ to compile the code.

  Here is a picture of my build errors:

  Normal windows applications are not so forgiving on where you can put your DLL under LabVIEW. You must have a DLL or in the Active Directory where the executable file itself (this is your build directory, not the source directory), windows or directory system or any what directory defined in the PATH environment variable. Anythin else probably won't work, or at least not reliably.

  No need to add the cintools directory to your include paths or paths of the library file. You must only that if you decided to leave native LabVIEW DLL export data types, in which case you would need Manager of LabVIEW to prepare the correct data buffers.

  Rolf Kalbermatter

• Programmatically determine if the forcing is enabled on target in real time

  A HMI application that I'm working, I want to inform the user at startup if the forcing is enabled on the target if all I/O's today obliged.  This information is readily available in the Distributed System Manager, but I can't seem to find a way to get this information programmatically.  Is this possible?  I hope there is a simple solution that I forget.

  PSP data items that are automatically published by the engine of analysis for each variable are displayed in the dialog box browse that you have attached to your previous post.  As long as you have activated the publication for the network i/o Variable, you should be able to access the information of forcings by reading a PSP Variable with the appropriate URL (for example "ni.var.psp:///mod1/AI0/ForceEnabled").  Similarly, you might find the value of forcing by reading "ni.var.psp:///mod1/AI0/ForceValue.  That's how the Browse dialog box is to get this information.  It's simply display all PSP items present on the network.

• NEITHER 9205 always returns the noise from rail to rail.

  Hi all

  I'm completely stumped on this one. I went through the procedure of implementation of the NI 9205 module in my project. Setting the channel to the appropriate type. Placed the node IO in my FPGA code, read and all I get is noise from rail to rail, that is, whatever the measurement range is, I'm getting noise across the range. I ran it as a measure of CSR, as a measure differential, differential with COM updated wrt my diet, differential with COM Earth and the mass of the chassis connected to my powersupply ground.

  Rail to rail noise ringing bells for anyone?

  I checked with an oscilloscope that the appropriate signal is seen on the pins in all the configurations above. I even unplugged and plugged a 3 v battery through the pines and still noise from rail to rail get.

  Help!

  HM, I used the 9205 before, but didn't have never seen something like this. I am more fan of data buffering in a FIFO, rather than reading a control value, but still, you shouldn't see this kind of securities for a reason that I don't see. I had problems with the modules where something is messed up and randomly, the signal will drop out then go back (in fact, I am about to send an or because it is still under warranty, but it's in a portable box, so I think that he may have had problems with shock or something of the sort). I wish I had a good answer for you, but you seem you know what you're doing, so it could be one of those cases where there is a hardware problem.  You have any other module HAVE you can share and see if those working in this niche? You can try it in a different location to maybe see there is a connection from bad backplane? You can also check if your local rep OR has a loan that they can indulge for a day, just to see if another 9205 works. Usually, these are the steps I take when I think that there may be a hardware problem. Unfortunately, except that I don't have a lot of input for you.

• Shared a Variable trigger

  Hello all, I have two projects (one is 64 bit LV, the other is in 32 bit LV) I run by-side on the same machine. The 64-bit project control a camera, and the 32-bit calls code to perform processing. I would like to send a trigger to make the 64-bit program wait until some of the raw data buffers are processed before sending more data for the 32-bit project. Unfortuantely cannot recomplie the code for 64-bit processing.

  Is there a good examplse to do? I was thinking about using some kind of shared variable configuration, but I don't really like shared variables.

  Thank you

  You can send commands via tcp/ip for different processes, e.g.

  Shared variable network is probably a little easier, but not much.

  /Y