9237 & 9205 with 9178 chassis sampling rate problem

Similar Questions

• With the NI 9205 module Max sampling rate - problems

  Dear friends,

  I develop a project of lv, which makes and control system of engine dyno. The material is CRio-9022 with other cards and also 9205 for AI. There is an encoder for angle attached to the motor shaft with 3600 chatted by Tower as well as an index to indicate the end of a revolution. the output of the encoder is measured by card 9411. The speed of the motor is 1500 rpm. I measure pressure data and couple when I receive a 'tick' of the wheel. This means my sampling rate for pressure and torque each is 90KO/s.

  but I was not successful to lead it. The program is great and I can show them, but I believe that there is a problem in the choice of material for the task. With the data of pressure and torque of the 9205, I also measure other channels for the controller output mass flow and temperatures. So in all I use 8 channels of the 32 available. But only the pressure and torque are acquired at the wheel-driven sampling rate. the rest are acquired about 5 times per second.

  Since the 9025 is a multiplexing ADC, 250K sampling frequency is divided by the number of channels accessed = 250 K/8 = 31 K samples/channel. With this in mind, I decided to acquire data of pressure and torque with each beat 3rd rotary encoder, essentially on 30K samples/s sampling. However, I see a large amount of noise.

  So I decide to average more than 1 second cycles (so the engine runs at about 25 cycles/sec, I averaged over this issue). The resulting pressure and torque graphics do not match with those measured by an oscilloscope in terms of amplitude but the frequency and shape is correct.

  I noticed an interesting feature in the charts. When I pass interpolation between the points, I see several curves made by points instead of a continuous locus of points. Accordingly, I find that the acquisition is slower than necessary, and so there are less number of points sampled as required. These points are not synchronized 25 cycles I have on average and therefore the separate "curves". It is because of the possibility that some points receive a higher number of 'contributions' several times (when you add), the neighbouring points.

  so I conculde that the 9205 is not fast enough to do the job. also noise, perhaps due to crosstalk or gosting when the mux changes channels. the impdences output pressure and the couple are of the order of 10 K ohms.

  the Labview code outline: well, there is a vi FPGA, which takes the rotary encoder ticks and sends a signal to the case of each 3rd tick. The signal contains a 16-bit integer, indicating the number of ticks. This signal is sent to a 1 element FIFO. This fifo is read in a parallel while loop, where it remains awaiting a new element. The while loop bed fifo, where data are available, takes a measure of pressure channel. A node memory of the method is called to provide data according to contained in the index number equal to the number of ticks to signal fifo. Then he adds the current pressure reading to the reading of the memory and stores the sum in the same memory location.  Thus, an array of elements of 1200 is formed, where each elemnt is a sum of the values taken of more than 25 cycles. This memory is transferd to a dma fifo and reading side host. is done similarly to involved couple. host-side the fifo is read and divided by 25 to get the average. This average is displayed on a waveform graph.

  Please check the attached file to get an idea of the problem. Sorry for the long post.

  Please suggest if you understand the problem and suggesstions or solutions.

  
  

• CANOpen - sample rate problem

  Hello

  I use the CANOpen toolkit to communicate with a sensor that can be set at a sampling frequency of 1 kHz. Before I used the standard frame API and there I received my PDO objects to each MS because I want to make programming easier I would use CANOpen. But I can't set the sample rate of 1ms. When you use 2ms data is not transmitted with the correct timestamp interval.

  I have attached the message error code when you use 1ms as sampling frequency and the part of the diagram block with the initialization of the PDO.

  Maybe someone can give me a hint.

  Best regards

  --
  Joachim

  It is good, your sensor begins to send pictures as soon as you start communication.

  You might remove the remote setting function and then it would behave like your old application framework API, except for one detail.

  Your old approx. api framework its mult read read all available images.

  The CANopen Read is only a single point read. Thus, he read only a single image. This means that you need to run the loop in ms rate, that is really risky with windows.

  You can enable buffering, but it can lead to errors of overflow, or turn off the buffering is now and read the most recent value.

  See the example updated the.

  DirkW

• After starting hearing sound becomes distorted (sample rate problem?)

  Hello!!

  Just when I start the program (Audition 2), I noticed that all the sounds that play inside or outside the program are distorted. He

  kind is sound that occurs during playback of a file to the frequency of sampling, kind of digital noise.

  When I stop the program and start any sound or a mp3 file on the computer, he continues to play distorted. It happened immediately and did not have before.

  I have Windows XP. The audio driver is "Soundmax Digital Audio". The audio driver is "Audition Windows Sound".

  At the hearing, in the Control Panel, hardware Audio Setup, "directsound output ports" are defined Soundmax Digital Audio with the size of buffer 2048; audio channels: 2 ; Bits per sample: 16. Said of the driver properties: 44100 Hz sampling frequency of; clock source internal, samples of size 2048 buffer

  There is no problem with my drivers or anything. I have no idea why this has happened.

  Any ideas would be greatly appreciated! I hope it's a sort of hearing that must be adjusted.

  I had this happen when I used a card firewire motu back in the day.   It worked fine for a few days and all the sudd

  en I'd get the loud pop and the line in my record.  I sold it after finding that it was not compatible with my windows upgrade.   Here are a few ideas that might help you

  • Your sound card has a windows update and threw it out of your drivers. Try to re install your drivers again now that your pc has updated.

  Hope that helped.

• Block in the amplifier with high material sampling rates no DSA

  Hello

  I intend to use the starter kit OR locking amplifier to detect a harmonic signal of 400 all about. I intend to use the PCI-5105 digitizer (I don't him have not bought until now, but I will soon). The start kit lock, I downloaded and unpacked the files and was hoping to give it a try with an NI USB-6363 however the vi does not open because it says that there are a lot of sub - VI lack (Subvi read.vi I HAVE... clear.vi, etc.). The LockinDemo.vi also seems to use NI 4472 as a default material. My questions are if the lock-in-amplifier starter kit OR 1) can be used with PCI-5105 and 2) cannot vi missing of the be downloaded from somewhere or be replaced by other blocks on mine?

  Thanks in advance for any help

  Benoit

  Hi Benoit,.

  First to start development with LabVIEW Real-time, you can check if your system supports LVRT. You can make use of RT disk utility that will test your system as shown in this article: http://digital.ni.com/public.nsf/allkb/9209361E17708D548625744A007FF353

  We have lists of equipment belonging to systems tested here: https://decibel.ni.com/content/docs/DOC-10692 and you can always check for the other hardware compatibility here: http://zone.ni.com/devzone/cda/tut/p/id/8239

  We do not have a driver available for the PCI of pledge cards but if these cards use VISA calls, you can develop and LVRT recognizes your card by following the KBs:

  (1) acknowledge the third peripheral part with LVRT: http://zone.ni.com/devzone/cda/tut/p/id/3142

  (2) expand LVRT peripheral Driver PCI: http://zone.ni.com/devzone/cda/tut/p/id/3251

  I hope this info helps!

• Problem with the audio sampling during export

  I'm having a problem with my audio sampling rate. For some reason, it is set at 8 khz by default when I export. It wasn't a problem until the last update (2 weeks ago). I'll try and give all the info I can to the front. I also tried customer support... which was a nightmare.

  OK, the question that I address is that when exporting audio sampling rate is 8 kHz by default. Which to my knowledge has NO relevant use what so ever. Even when I set the file format to the h.264 format, he insists on the reset to 8 kHz. Even if I manually set all look rendering queue and click on output module again once it back the sampling frequency to 8 khz automatically of changes that I need to set it again. Why can it not only be fixed to 48000 like all other adobe products is and as it was two weeks ago. Does anyone have an answer?

  This problem is resolved by update fixes after effects CC (12.2.1), which is now available:

  http://Adobe.LY/AE_CC_1221

  Note the part at the end of this page on a crucial update for the creative cloud desktop application, which addresses serious problems with the SOUL, first Pro and After Effects.

• Bad analog output help Every_N_Samples-NI-9263 cDAQ-9172 chassis (works with cDAQ-9178 chassis)

  Hello

  The NOR-9263 analog output voltage geberation works correctly with the cDAQ-9178 chassis but gives wrong result using the chassis NOR cDAQ-9172.

  In the attached code example, a single cycle of a sine wave is composed of 40000 samples and came out in the background using Every_N_Samples at a rate of production of 5000 samples per second.

  The output buffer size is set to 10000 samples.

  Prepare us the buffer writing 10000 samples 1, then write the remaining data in the background using the Every_N_Samples callback.
   
  Bug: Using the cDAQ-9172 chassis, to the 5000 s/s sampling rate with the help of an external field (or through closure to another HAVE), we observed that 1 10000 samples came out twice, followed by the rest of the waveform. The last 10000 samples are never exits. If you are working properly, we would expect to see 1 full cycle of a sine wave.

  The bug does not occur with the chassis NOR cDAQ-9178. I use the driver NIDAQmx v9.2.1f0 on Windows XP
  The bug does not happen with simulation devices, so you will need to use harwdare real to reproduce.

  Please find attached an example of code C based on the example program OR "ContGen - IntClk.c" to reproduce this bug.

  Thank you

  whemdan,

  The MathWorks

  Hi whemdan,

  By default, DAQmx regenerate old samples if no new data is available.  To give the correct behavior, you can:

  Use DAQmxSetWriteRegenMode to disable the regeneration (DAQmx_Val_DoNotAllowRegen).  In most cases, this is recommended if new data are written continuously in the buffer as the build is in progress.

  If you just need to generate 40 k samples, you can write them just all at once, rather than in 10 pieces of k (the code you attached probably is just an example, so I'll assume that you have a reason to write the data into segments in your actual code).

  I think the difference in behavior between 9172 and 9178 can if explained by the different way, buffering is set up on each product.  The 9172 uses a buffer of 8 k (on the STC2) in all cases (source).  The 9178 uses an 8 k of memory buffer (on the STC3) If you use regeneration shipped, but uses the 127 samples FIFO cartridge, if you use no on-board regeneration (source).

  Then... on the 9172 8191 samples are immediately transferred to the FIFO.  By default, the hardware is going to request new data when the FIFO is less to fill (this is configurable with DAQmxSetAODataXferReqCond).  I'm not sure what the transfer data request size is in your case (you can set the maximum value with DAQmxSetAOUsbXferReqSize), but obviously it is bigger than the other 1809 samples that you have not yet sent to the Board of Directors of your first entry.  At this point, the pilot will regenerate 10 existing k samples so that sufficient data will be available to meet the demand of data transfer.

  The 9178 however use the FIFO of 127 smaller samples so you will not have the same behavior in your case.

  In summary, the behavior is explainable by the difference of material.  If you want to avoid to regenerate old samples, you should ban the regeneration using DAQmxSetWriteRegenMode.

  Best regards

• Different sampling rate with the same connector AIO, Labview FPGA

  Hello

  I use LV 2009 with the new Toolbox FPGA and an NI PXI 7854R. I acquire an analog signal with a sampling frequency of 600kS / s. I need as the sampling rate for the processing of the data, but I also need the signal sampled with a much smaller, variable sampling frequency to a FFT.

  I've attached a picture to clarify, in a simple example, I'm looking for.

  

  I tried with the structure case only take each ' iht iteration, but did not get the expected results.

  Does anyone have another idea how to solve my problem? Of the, "Resampling" express VI in the funtion FPGA palette does not help me.

  Thanks in advance,

  Concerning

  Hello

  the connector for the analog input is a "shared resource", so you should he alone in your FPGA Code.

  Find attached an example that shows how to perform this task of analysis.

  Concerning

  Ulrich

  AE OR-CER

• NI 9237 sampling rate

  Hi all

  I use a module 9237 for certain measures of the load. My experiences last over time and so I'm generating a lot of data due to the minimum sampling frequency.  I can't define an external time base so I can lower my sampling rate to something easier to manage?  Even just a sample rate of 500 s/s would make a huge difference.

  Thank you

  Hi cannisbellum,

  9237 specifications frequency range of minimum data (fs) using the internal master time base is 1,613 kech. / s and external use master timebase is 391 s/s. The simplest would be to sample at 2kS/s and decimate your data by 4. This can be done by using 'Decimate 1 table D' or ".vi Decimate (continuous).

  Rates valid for the NI 9233 OR 9234 sampling and NI 9237 - http://digital.ni.com/public.nsf/allkb/593CC07F76B1405A862570DE005F6836?OpenDocument

  Best,

  CARISA

• How to create a waveform from an array with arbitrary sampling rates

  Hi I know that sounds a little silly,

  Suppose I created a simple table of figures DBL with a structure For, Say size 16. now, I want to create a waveform DBL with these 16 numbers at an arbitrary sample rate. so if I use 1 kech. / s to the sampling frequency, I want to have a waveform with a duration of 16 milliseconds.

  Please help me, I need it too

  TNX

  Hello

  You must use the wave to build function as shown here: http://zone.ni.com/reference/en-XX/help/371361G-01/lvwave/build_waveform/ . Wire you your Board at the entrance Y and then wire the dt of entry in your sampling rate.

  -Zach

• acquisition of data with different sampling rates high

  I have a few questions on the use of the OMB-DAQ-3005 with different sampling rates high.

  For our application, we have 8 analog inputs. Which two are a quick response and should be sampled frequently.  We have an encoder quadrature (CPR 1000 running at 1800 rpm).  We plan to sample X 4 encoder.  For the analog inputs for the quick response, we want to trigger a sample of each pulse or each a few pulses, thus creating a timestamp with the position of the encoder with respect to position index as well as two fast analog inputs.  We have data correlating the analog inputs with the position of the encoder.   Other analog inputs, we want to measure relatively slowly (for example once every 5 dry or similar).

  How can I go on the configuration of the two (or more) sampling rates different wherein I can taste entered at different frequencies?  Also, is there a way to reset the encoder count after outbreak of the index as I have the position of the encoder with respect to the index?

  Maybe you'll find someone here who uses the OMB-DAQ-3005, but this forum is really more designed for LabVIEW programming issues.

  I've never used the OMB-DAQ-3005, but out of curiosity, I took a glance at the Manual of OMB-DAQ-3005.  The answer to both your questions are:

  1. you cannot run a hardware DQA Multiplex (like this one) at independent rates by channel.

  2. the OMB-DAQ-3005 supports an Index Z feature to reset the counter - look for documentation on how to configure any software interface you are using.  If you get stuck, you can try to discover media appropriate for instrument channel.

  Best regards

• Problems with the LCD - samples.war 2.6 LCD screens

  Hello!
  We are trying to compile a project with screens LCD - samples.war, using LCDS 2.6 Beta 1, and it does not work. It generates this error:
  
  AddChannel (String) method is undefined for the type DataDestination LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 55
  
  CreateDestination (DestinationSettings) in DataService type method is not applicable for the arguments (String) LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 54
  
  CreateDestination (DestinationSettings) in the MessageService type method is not applicable for the arguments (String) LCD-samples/src/flex/samples/runtimeconfig
  ChatRoomService.java
  line 42
  
  CreateDestination (DestinationSettings) in the RemotingService type method is not applicable for the arguments (String) LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeRemotingDestination.java
  line 18
  
  The createUUID (boolean) in the UUIDUtils type method is not applicable for the arguments () lcds-samples/src/flex/samples/feed
  Feed.Java
  line 32
  
  The createUUID (boolean) in the UUIDUtils type method is not applicable for the arguments () lcds-samples/src/flex/samples/feed
  Feed.Java
  line 53
  
  The createUUID (boolean) in the UUIDUtils type method is not applicable for the arguments (lcds-samples/src/flex/samples/marketdata)
  Feed.Java
  line 42
  
  The createUUID (boolean) in the UUIDUtils type method is not applicable for the arguments (lcds-samples/src/flex/samples/marketdata)
  Feed.Java
  line 67
  
  The createUUID (boolean) in the UUIDUtils type method is not applicable for the arguments (lcds-samples/src/flex/samples/pdfgen)
  PDFService.java
  line 32
  
  The createUUID (boolean) in the UUIDUtils type method is not applicable for the arguments (lcds-samples/src/flex/samples/pdfgen)
  XDPXFAService.java
  line 32
  
  The method getFlexClient() is undefined for the type FlexContext LCD-samples/src/flex/samples/qos FlexClientConfigService.java line 15
  
  The method getFlexClient() is undefined for the type FlexContext LCD-samples/src/flex/samples/qos FlexClientConfigService.java line 21
  
  Method getService (String) is undefined for the type MessageBroker LCD-samples/src/flex/samples/runtimeconfig
  ChatRoomService.java
  line 41
  
  Method getService (String) is undefined for the type MessageBroker LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 53
  
  Method getService (String) is undefined for the type MessageBroker LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeRemotingDestination.java
  line 17
  
  The method isStarted() is undefined for the type MessageService LCD-samples/src/flex/samples/runtimeconfig
  ChatRoomService.java
  line 44
  
  SetDestination (DataDestination) method is undefined for the type JavaAdapter LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 78
  
  The setId (String) method is undefined for the type JavaAdapter LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 76
  
  SetMetadataSettings (MetadataSettings) method is undefined for the type DataDestination LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 61
  
  SetNetworkSettings (DataNetworkSettings) method is undefined for the type DataDestination LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 73
  
  The setScope (String) method is undefined for the type DataDestination LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 57
  
  The setSource (String) method is undefined for the type DataDestination LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 56
  
  The setSource (String) method is undefined for the type RemotingDestination LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeRemotingDestination.java
  line 19
  
  SetSubscriptionTimeoutMinutes (int) method is undefined for the type DataNetworkSettings LCD-samples/src/flex/samples/runtimeconfig
  EmployeeRuntimeDataDestination.java
  line 64
  
  The start() method is undefined for the type MessageDestination LCD-samples/src/flex/samples/runtimeconfig
  ChatRoomService.java
  line 46
  
  The DatabaseCheckService type must implement the inherited abstract method Service.getSettings () DatabaseCheckService.java LCD-samples/src/flex/samples 9 line
  
  The EmployeeRuntimeDataDestination type must implement the inherited abstract method Service.getSettings (LCD-samples/src/flex/samples/runtimeconfig)
  EmployeeRuntimeDataDestination.java
  line 44
  
  The EmployeeRuntimeRemotingDestination type must implement the inherited abstract method Service.getSettings (LCD-samples/src/flex/samples/runtimeconfig)
  EmployeeRuntimeRemotingDestination.java
  line 8
  
  Do you know something? There is a updated version of screens LCD - samples.war to LCDs 2.6?
  

  Hi Rafael,.

  LCDS 2.6 is not available on Adobe Labs. Please address all questions and issues the appropriate preliminary Forum.

• High sampling rate for 18 strings in a single task - DAQmx

  Hello

  My current experiences require the acquisition of 18 channels (6 HAVE custom voltage with excitement, voltage AI 12) data at a sampling frequency of 50 kHz. My question is if it is faesible using a single cDAQ-9172 chassis. Used modules are 2 x NI 9237 and 4 x NI 9215.

  The previous researcher on this project used 2 different Renault to do this, but I was hoping that I could reduce it to a single DAQ to simplify the synchronization of all channels.

  I wrote a *.vi (attached) to do this and write to a TDMS file using a structure of producer-consumer; However, when I run the * .vi, I can run for minutes, but the number of samples recorded for the PDM is rarely more than 100 k. Subsequently a buffer size error (attempted to read from the samples that have been overwritten) with less than 200 k samples ever record. I checked the max sampling rate (using a property timing node) with the configured task as in the * .vi and it shows a maximum of 235kHz.

  I can't say if I make just the structure of the producer consumer incorrectly or if I ask too much of the cDAQ-9172 unique?

  Any help would be much appreciated.

  See you soon

  Bart

  Hi bart.s,

  TiTou speaks sampling aggregate not simultaneous, so sampling rate 50kS/s is the same for all channels.

  I see the problem in the loop of the producer. If your sampling frequency is 50kS/s/ch and you read that a single sample/ch you will lose data because the producer loop cannot run so fast. You should read more than one sample. I recommend you also to move your tracing to the consumption loop code and work with larger amounts of data.

  The second problem may be with the error handling in your loop of consumer. Merge the mistakes of loop of consumer and producer and also add a few general for two loops off if the error occurs (for example, using local variable).

  Best regards

  CaravagGIO

• Sync module NI 9205 with NOR-9234

  Hello

  I have a cDAQ-9178 chassis with 4 x accelerometers IEPE PCB 16 absorbent of NOR-9234 (DSA). I have a NOR-9205 I intend to use to acquire the power of 4 problems. The NI - 9234 s are synchronized through their extensions in a single task. I would add the NOR-9205 to the chassis and synchronize it with the NOR - 9234 s. I have the following questions:

  1. Calendar synchronization

  How to properly START and the NOR-9205 CLOCK synchronization with the NOR - 9234 s? PXI I export the master St DSA. WHICH device (OR-9234 or NOR-9205) should export WHICH signals to WHOM?

  2. even or parallel task

  Since I produce a TDMS file in its task of NOR - 9234, it would be nice if I could simply to ADD the OR-9205 to this task to develop this unique TDMS with extra 4 channels of the NOR-9205 file. Is this possible?

  Thanks in advance,

  Hi golubovski,

  > Are you saying my task of canal expansion can fire them OR - s and NEITHER-9205 9234 with its 32 channels of multuplexed HAVE? How to recognize which modules can be combined in this way convinient?

  Yes. With the cDAQ-9178, any set of AI modules can be included in the same task and in most cases that apply to the AO and DIO as well, but you cannot mix a, AO and DIO in the same spot. The NOR cDAQ-917 x manual goes into more details on what combinations of modules, the types of channels and the tasks are possible.

  > I am aware since I worked so far with the PXI - 4472 DSAS. How well can correlate between the NI 9234 - and the NI9205?

  Well, I wanted to make sure that you did not know that it is something that you'll probably need to take into account. On the top of my head, I don't know the answer to this question.

  > As far as I know, in which case I use dummy DAQ with the NOR - 9234 to offset 40 clock cycles I guess I'll need to separate the tasks - how to integrate viewers into a single stream of TDMS?

  If you use two separate tasks and you are using the built-in support of DAQmx for TDMS logging, then you will get flow TDMS separated for each task. Note that a dummy read will still record to the PDM file, then maybe it will not work in your case.

  If you use the TDMS API to write to the PDM file, you can do the preliminary processing of data before it is written in a single stream, but there is some trade-off between ease of use and efficiency: either you write data point floating in the file (easier but less effective), or you have to deal with scaling (most effective raw data (, but harder to do).

  Another approach is to put all the modules in the same task and compensate for the delay of the filter after reading the data from the PDM file.

  > Also, for fixed 3.2 µs I have to calculate how many cycles given the actual sampling frequency or is there another method?

  If you filter delay compensation by removing a whole number of data samples, then Yes. However, there are other methods of data processing that can compensate for a number that is not part of the sample clock periods. I don't have a recommendation, but maybe someone else does?

  Brad

• new cDAQ-9178 chassis was smaller than the cDAQ-9172 old buffer?

  Digital waveform features:

  Waveform acquisition (DI) FIFO for cDAQ-9172: 2047 samples

  Waveform acquisition (DI) FIFO for cDAQ-9178: 127 samples per slot

  This means that the new HW (9178) has a much smaller than the old buffer (9172)!

  I want to run a correlation over generation/acquisition at 2 MHz with 2000 samples. Does this mean that the new chassis will not be able to acquire the whole of the data (I use a NI 9401 module inside the chassis)?

  FIFO sizes are misleading and not a true indicator of the chassis supporting the streaming is not the single buffer.  We did some tests comparing the latest and cDAQ-9172 chassis.  You will not notice the difference, especially with your application to a finished task of 2 000 samples at 2 MHz.  With 2 000 samples, you can run your DI task on your chassis cDAQ-9178 at 10 MHz if you wanted without problem (don't try with continuous or even finished tasks).

  The only time where the FIFO size really comes into play is when you run a primary task at high rates, say > 5 MHz, even if it is system dependent.  At these speeds a FIFO that is deeper 2 000 samples you buy only tens or hundreds of additional microseconds when Windows jitter is measured in milliseconds.

  I would be very surprised if there is a practical application that worked on a cDAQ-9172 chassis which did not work on newer USB cDAQ chassis.