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

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
11/19/05 05:21
Read: times


 		 #103906 - This won't really help.
Responding to: ???'s previous message
Jacob Boyce said:
I had a similar problem until I ran the results from the pseudo-random generator into the pseudo-random generator again

Two PRNGs back-to-back is still just a PRNG. And given the same input, the result is the same output.

To avoid this problem you need to use either (or better both) of the techniques:

(a) Include some distinguishing data in the PRNG, like device address or serial number.
(b) Include some non-deterministic variation (real randomness), like the low order bit(s) of a high-speed timer taken during an unrelated interrupt.


List of 34 messages in thread
TopicAuthorDate
RS485 Question            01/01/70 00:00      
   drive capacity            01/01/70 00:00      
      3 resistor termination            01/01/70 00:00      
         Only if you have power to burn            01/01/70 00:00      
            15 milliwatts. A lot of power?            01/01/70 00:00      
               Depends on your application            01/01/70 00:00      
                  how are you going to terminate at that            01/01/70 00:00      
                  RS485 loops are hungry...            01/01/70 00:00      
      You will loose all the noise margin!            01/01/70 00:00      
         not if you use 3 resistor termination            01/01/70 00:00      
            Better, but...            01/01/70 00:00      
               works for me, what do you have that is b            01/01/70 00:00      
                  No disagree...look what weird is planned            01/01/70 00:00      
                     Kai, I did not think you did            01/01/70 00:00      
                        Clarification            01/01/70 00:00      
                           3 resistors or not            01/01/70 00:00      
                              Then is all fail-safe bad?            01/01/70 00:00      
                                 Can you see the difference?            01/01/70 00:00      
                                    Kai, you keep ignoring the fact that RS4            01/01/70 00:00      
                                       I ignore nothing            01/01/70 00:00      
                                          It's the best we have and it works when            01/01/70 00:00      
                                             Yes, but it can fail from time to time            01/01/70 00:00      
                                                I have handled that with isolated RS485            01/01/70 00:00      
                                                   We do already            01/01/70 00:00      
                                                      485-422            01/01/70 00:00      
                                                         I challenge you to find a RS422 driver            01/01/70 00:00      
                                                            In the past.            01/01/70 00:00      
                                 All fail safe is good and 0.25V is plent            01/01/70 00:00      
                           That is not RS485, that is J1708            01/01/70 00:00      
                              trick to get around pseudo non random            01/01/70 00:00      
                                 This won't really help.            01/01/70 00:00      
                              Helpful            01/01/70 00:00      
                                 I am doing that using the Dallas DS2401            01/01/70 00:00      
                                    Yes that is what I am planning            01/01/70 00:00      

Back to Subject List