AnsweredAssumed Answered

8720D Repair - Intermitent Failure..

Question asked by najockey on Mar 31, 2010
Latest reply on Mar 31, 2010 by najockey
Hi All,

    I am working on the repair of an 8720D. Initial symptoms were sometimes failure to start sweeping after passing startup tests, inacurate source level (not noticed initially) , and instability if it did make it past the initial startup. The Power Supply voltages were within spec, except for the 22 v, which was 22.9 (versus the 22.7 v max ). The configuration file memory was reset and the battery is right on 3.0 volts. I noticed that small vibration would trip out the unit. So, I soldered up all the open via's in the CPU board (there were about 30 of em). Apparent good operation (Op Ck 21 and 22 successful) resulted, and vibrations applied to the CPU board would not reset the unit as before. The unit then failed Op Ck 21 at the -50 dBm level test. This was traced to the source level being wrong. What was apparent on the screen at startup was that the S11 sweep was centered between -10 and -20 dB (mostly, above 2 GHz), rather than 0 and -10 db when the level is correct. At the -45 dBm switch when turning the level down manually (in auto mode, uncalibrated), the level changed back to between 0 and -10 dB. Traced this to a stuck 20 dB attenuator stage that was not being initiallized off, or reset off when at the -45 dBm switch point. A loose ribbon connection to the Test Set interface was found. And the level has been correct since. Now all the attenuator stages are acting properly and the 20 dB stage that was stuck is getting switched in and out as commanded. At that time I could not make the analyzer stop working.

    So, I let it run to see if it would fail again, and put the covers back on. I did a calibration of an operational congifuration file and let it sit and sweep measuring a cable assembly. About 1 1/2 days into this test the screen blinked off, and came back with the restart screen with the message "Please Stand By - Error : ADR" . It would then reset about once every few minutes after flashing that message, and sometimes the message would end in "INST".

     Pulling off the lower cover, I found the Pulse Generator assembly was very warm. If I grabbed it, I could hold my fingers on it for about 20 seconds before the discomfort caused me to release the heat sink of the Pulse Generator. With the cover off the Pulse Generator cooled down to a holdable temperature, and the once every few minutes resetting ceased. Although, it did reset once during this time. Op Ck 21 and 22 were successfully passed, as well as all other tests.

     So, I guess my question is, is this nomral Pulse Generator operation? Do I need a CPU board too ? Is there something else wrong?