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

Back to Subject List

Old thread has been locked -- no new posts accepted in this thread
???
09/25/10 14:51
Read: times


 		 #178718 - Brute Force - The IC pins would curl up before finish
Responding to: ???'s previous message
Andy Neil said:
Neil Kurzman said:
try every pattern on the chip under test

How long would that take?!

How many combinations of input, output, clock, etc are there for such a "brute-force" attack on even a 14-pin chip...?!
And that's before you even think of stored state...


I am rather sure that most of the "chip identifiers" used a template scheme for each "known" type and would then simply cycle through a few hundred of these templates to see if the chip matched one of them. Trying to write an intelligent discovery process would quickly become prohibitive in terms of development cost, run time and resource usage.

Michael Karas



List of 103 messages in thread
TopicAuthorDate
Very nice for third countries but...            01/01/70 00:00      
   My Advice            01/01/70 00:00      
   Today's top award for Stating The Blindingly Obvious...            01/01/70 00:00      
   Doesn't make sense            01/01/70 00:00      
      A microcontroller is a nice logic-chip tester            01/01/70 00:00      
         device in question is not for identifying unknown chips            01/01/70 00:00      
            Wasn't talking about linked device            01/01/70 00:00      
         IC logic still popular here(used)            01/01/70 00:00      
            sure, but            01/01/70 00:00      
               sure, but, but            01/01/70 00:00      
                  sure, but, but, but            01/01/70 00:00      
            Limitations            01/01/70 00:00      
               Areas of Doubt and Uncertainty            01/01/70 00:00      
                  I can imagine one case            01/01/70 00:00      
                     That would have different pinout            01/01/70 00:00      
                     what is all this about identifying chips            01/01/70 00:00      
                        See Richard's post            01/01/70 00:00      
                  Brute Force - The IC pins would curl up before finish            01/01/70 00:00      
                     Actually not too ahrd/expensive to detect chips            01/01/70 00:00      
                        But            01/01/70 00:00      
                           Of course            01/01/70 00:00      
                              Again, it depends on your goals            01/01/70 00:00      
                                 Scope of project is important            01/01/70 00:00      
   It's more interesting            01/01/70 00:00      
   both of my old portable programmers do this            01/01/70 00:00      
      Is there codes?            01/01/70 00:00      
         What, exactly, are you looking for?            01/01/70 00:00      
            unknown is hard....known is easy?            01/01/70 00:00      
               it's possible, but not reliable to check for unknowns            01/01/70 00:00      
                  xeltex superpro            01/01/70 00:00      
                     be careful ... that was several revisions ago ...            01/01/70 00:00      
               RE: known IC logic for test is easy?            01/01/70 00:00      
                  unknown....            01/01/70 00:00      
                     Pardon??            01/01/70 00:00      
                     IF you can't recognize the part, THEN discard it!            01/01/70 00:00      
                     The most likely reason you'd want to do that ...            01/01/70 00:00      
         The code was compiled and in a library            01/01/70 00:00      
            Place..The code was compiled and in a library            01/01/70 00:00      
               Will that help?            01/01/70 00:00      
                  in the future...            01/01/70 00:00      
                     It's just a small-C compiler with intrinsics for SuperPro            01/01/70 00:00      
                        help??            01/01/70 00:00      
                           Can't you just go from first principles?            01/01/70 00:00      
                           If you want that software, you should contact XELTEK            01/01/70 00:00      
                               The code was compiled and in a library???            01/01/70 00:00      
                                 Well, obvisouly!            01/01/70 00:00      
                                 The hardware for this sort of thing is not rocket science            01/01/70 00:00      
                                    library binary            01/01/70 00:00      
                                       I can't purchase the expensive SuperPro            01/01/70 00:00      
                                          homemade            01/01/70 00:00      
                                             this makes no sense            01/01/70 00:00      
                                                missing??            01/01/70 00:00      
                                                   no, I am not            01/01/70 00:00      
                                                      not so fast ...            01/01/70 00:00      
                                                         while all that is valid - to an extent            01/01/70 00:00      
                                                            kitchen table...            01/01/70 00:00      
                                                               Kitchen-table = home built with lacking tools/knowledge            01/01/70 00:00      
                                                                  understand....            01/01/70 00:00      
                                                         Can often be solder problems            01/01/70 00:00      
                                                      multiply            01/01/70 00:00      
                                                         if you use that many ....            01/01/70 00:00      
                                                            work's            01/01/70 00:00      
                                                               because            01/01/70 00:00      
                                                                  You're right in large part, but ...            01/01/70 00:00      
                                                                     did NOT bring this up for any other reason            01/01/70 00:00      
                                                                  back...            01/01/70 00:00      
                                                                     Can have uC or PC as master - just different advantages            01/01/70 00:00      
                                                                        You'll need lots of storage.            01/01/70 00:00      
                                                                           For what?            01/01/70 00:00      
                                                                              You need storage for the 250K test vector sets            01/01/70 00:00      
                                                                     Yes, you have to do the work            01/01/70 00:00      
                                                               "works (sic) flawlessly (sic)"            01/01/70 00:00      
                                                               You haven't the means to determine that it works flawlessly            01/01/70 00:00      
                                                                  XOR normally easier than trying high-end drive + sampling            01/01/70 00:00      
                                                                     That does work if you have one functional part for "seed"            01/01/70 00:00      
                                                                        Can perform time calibrations without specific seed            01/01/70 00:00      
                                                                           None of this works with just an XOR            01/01/70 00:00      
                                                                              Just an XOR?            01/01/70 00:00      
                                                                           These assertions look entirely wrong to me.            01/01/70 00:00      
                                                                              Read up on phase detectors and separate from lgic detect            01/01/70 00:00      
                                                                                 Yes, I know about phase detectors, and in this case...            01/01/70 00:00      
                                                                                    Learn, and it will not be irrelevant anymoer            01/01/70 00:00      
                                                                                       This stuff is still irrelevant until you know the type            01/01/70 00:00      
                                                                                          And the type covered again and again            01/01/70 00:00      
                                                                                             For the purposes described by the O/P it's not relevant            01/01/70 00:00      
                                                                                                Already covered            01/01/70 00:00      
                                                                                                   Until you have a way of ascertaining the ins and outs ...            01/01/70 00:00      
                                                                                                      Obvious, but _you_ singled out propagation delay            01/01/70 00:00      
                                                                                                         various but still MCS-51            01/01/70 00:00      
                                                                                                            Difference hobby or commercial use            01/01/70 00:00      
                                                                                                               you caught a VERY important issue - and did not know            01/01/70 00:00      
                                                                                 avoid kicthen table            01/01/70 00:00      
                                                               there is no way in hades ....            01/01/70 00:00      
                                                                  Not even commercial testers are really commercial grade            01/01/70 00:00      
                                                                     emphasizing the above good post            01/01/70 00:00      
                                                                     I'd guess the application is a bit different            01/01/70 00:00      
                                                                        ah ...            01/01/70 00:00      
                                                                           Yes ...            01/01/70 00:00      
                                       It's always a tradeoff between time and money            01/01/70 00:00      
                                          e-bay??            01/01/70 00:00      
                                             If you find one and buy it ...            01/01/70 00:00      
   If you REALLY want one of these ... but one that works ...            01/01/70 00:00      
   not a defective chip, but ...            01/01/70 00:00      

Back to Subject List