Author Topic: Advantest R3477 Digital IF Board repair  (Read 1128 times)

0 Members and 1 Guest are viewing this topic.

Offline copyright927Topic starter

  • Newbie
  • Posts: 4
  • Country: cn
Advantest R3477 Digital IF Board repair
« on: February 14, 2020, 05:27:42 am »
Just like what happened to nctnico here(45#  https://www.eevblog.com/forum/testgear/advantest-r3477-9khz-to-13ghz-spectrum-analyzer/25/).About a week ago, my R3477 reported "self-check failure" when it was started up, and checking the diagnostics info showed the Digital IF board failed.After replacing a short circuit 22uf ceramic capacitor on -5v rail, I still got a"fail".Further inspection revealed that two of the five ceramic capacitors next to the clock shielding box were defective, resulting in the output of the 5V->3.3V converter being only 1.xV (however, all parameters were normal when the removed capacitor was tested with LCR :-//).The digital IF board returned to normal after replacement.But, yesterday, I got another "Digital IF board not found" error |O. This time, it was a 22uf ceramic capacitor in the I-channel shielding box that short-circuited and burned the 5V rail fuse.I realized it seemed like a problem of poor capacitor manufacturing?!(like rifa) The capacitor failure rate is too high.So I decided to replace all this capacitors on the board (using 22uf/50V X7R  1812package to replace).
 

Offline DaJMasta

  • Super Contributor
  • ***
  • Posts: 2289
  • Country: us
    • medpants.com
Re: Advantest R3477 Digital IF Board repair
« Reply #1 on: February 14, 2020, 05:52:12 am »
Odd series of failures, do they all happen for a common 5V rail?  Maybe there's some kind of intermittent noise/overload failure that's just screwing with the rail and various caps are dying because of it.  Maybe the problem's actually being fixed by the act of inserting and removing the board in the process of testing - maybe next failure just take it out and put it in a few times and see if you can get it to come good?

Otherwise I'd consider just running a mod wire out for 5v and ground (and any other rails that have been problematic) and just put them into a scope on roll mode to look for anything bad looking in normal operation.  May be worth just repeatedly running the self test to see if the issue can "solve" itself with retries making it a possible other intermittent failure.
 

Offline copyright927Topic starter

  • Newbie
  • Posts: 4
  • Country: cn
Re: Advantest R3477 Digital IF Board repair
« Reply #2 on: February 17, 2020, 11:32:23 am »
Of the four MLCCS that failed, only one was actually directly connected to the +5V (the capacitor that burned the +5V fuse), and one of the remaining three was connected to the -5v and two to the base of the transistor for the 5->3.3(wondering why they used four transistors to independently power four 3.3v chips which in the same small shielding box)
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf