Email: Password: Remember Me | Create Account (Free)

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
02/02/11 22:00
Read: times


 
#180972 - Doesn't help to send at exact time - PC receive will jitter
Responding to: ???'s previous message
Neil Kurzman said:
Assuming No Time stamps. The samples need to be to exact same time apart. That means interrupts for sampling and transmitting

Above only works if the PC is able to process and time-stamp the received characters exact enough. The PC can't, so there are then only three working alternatives:
- the PC knows the sampling rate and steps a time variable for every sample it processes.
- the data is sent with a time stamp and the PC processes the data whenever it has time.
- the data is sent without a timestamp and the PC checks the current time and the data is presented with whatever jitter there will be.

1) requires that the PC never loses enough data to get one sample off. And the sample rate must be kept constant.

2) is most robust, and you have the bandwidth so there is no problem sending timestamps too.

3) will result in jitter. The load on the PC will affect the amount of jitter.


List of 40 messages in thread
TopicAuthorDate
Transmitting a 100 Hz signal through 8051            01/01/70 00:00      
   What is your problem?            01/01/70 00:00      
   RE: "the 8051 has transmission speed upto 19200"            01/01/70 00:00      
      the 8051 has transmission speed upto 19200            01/01/70 00:00      
         Incorrect analysis            01/01/70 00:00      
            Timing            01/01/70 00:00      
               Doesn't help to send at exact time - PC receive will jitter            01/01/70 00:00      
                  Windows For Sure Will...            01/01/70 00:00      
                  A Labview Limitation?            01/01/70 00:00      
         What makes you think that?            01/01/70 00:00      
            What makes you think that?            01/01/70 00:00      
               How fast are you sampling            01/01/70 00:00      
               Insufficient analysis            01/01/70 00:00      
                  20x            01/01/70 00:00      
         you can (I have) easily get 460kb.            01/01/70 00:00      
            No need to look for high baudrates            01/01/70 00:00      
               It doesn't help            01/01/70 00:00      
                  At least ...            01/01/70 00:00      
   Haven't you done this before...?            01/01/70 00:00      
   Here are the results in the form of images            01/01/70 00:00      
      Didn't bother to preview, did you?            01/01/70 00:00      
         How did you concluded?            01/01/70 00:00      
            Duplicate!            01/01/70 00:00      
         How did you concluded?            01/01/70 00:00      
            I think you're wrong            01/01/70 00:00      
            Looks close to 'as expected'            01/01/70 00:00      
               Seems like around 200Hz sample rate            01/01/70 00:00      
                  4 samples/period for the 50Hz signal            01/01/70 00:00      
            Not sure about your concept            01/01/70 00:00      
               Depends            01/01/70 00:00      
               Continuous monitoring?            01/01/70 00:00      
                  Ambiguity            01/01/70 00:00      
   Displaying the Signal at real time            01/01/70 00:00      
      Still not mentioned what the problem is            01/01/70 00:00      
         I do not think they know the problem            01/01/70 00:00      
            A Ring Buffer ...            01/01/70 00:00      
               A double-buffered solution also possible            01/01/70 00:00      
            What is the Problem.            01/01/70 00:00      
         real RS232?            01/01/70 00:00      
            Huge FIFO in USB-to-serial adapter            01/01/70 00:00      

Back to Subject List