Last update: 4.Sep 2008
Beginning of 2006 I was lucky to acquire HP8562A spectrum analyzer. Step by step I found it has several problems and start to restore it back to 100% conditions. Here is the story.
Bad new: When powered on, sometimes there was clear picture, sometimes just mesh of lines. Eh, intermittent problem, night mare.
Good new: After several on/off cycles there was just meshed screen without ability to read anything. Intermittent problem became permanent, hopefully will be more easy to find the fault.
Thanks to Agilent's web pages I downloaded scan of Service manual and start to read. It was my first "HP standard" service manual and I was very impressed by excellent job what HP did in old good times. I followed step by step trouble shooting procedures and it was quite easy to find missing blanking signal.
But now, why is blanking signal missing? I tried to search A2 controller board for some "visible" trouble but without success. OK, definitely I need detailed diagram. After extensive searching on the net I found the web of Cristian, F1FPS at http://www.voilelec.com/ham/hp8562a.php and acquired CLIP (Component Level Information Packet) from him (Cristian, thanks again). Now with detailed diagrams it was already much more easy to proceed. With help of oscilloscope I located faulty resistor trimmer A2R262 (STOP BLANK ADJ). After trimmer replacement, screen scrambling disappeared.
Now another problem become evident - intermittent screen intensity jumps. I start to play with minimum intensity adjustment on CRT driver (board A17) and found there is corelation between those jumps and tuning of A17R55 MIN-INTEN trimmer. After trimmer and resistors around replacement, problem disappeared, yippee! BTW, you can see that also cleaning was more than necessary!
When CRT became useful, I could start to play with RF measurement. And found another problem. Calibration signal at 300MHz was shown 20dB lower than it should be. Quick check by HP436A power meter confirmed that calibration oscillator is ok, so problem will be in the input path. After quick look into service manual an answer was found again - faulty low-band part of A8 Dual Mixer. Damm.
There are no details about mixer even in CLIP so I had to start my own troubleshooting. Network analyzer showed that low_band input impedance is only around 2 ohms, it means that something is blowen. Visual inspection under magnifying glass didn't show any burns. Due to lack of tools to play with such small things I decided to try one local company what uses more spectrum analyzers in their production. And that was the goal! Sometimes they also have problems with burned mixers. They cooperate with another company what has tools for micro bonding and it became just another mixer to be repaired. First they checked input protection PIN diodes but both have been fine. Then they found there is shorted beam lead diode (black "dot" next to AP2 label). As soon as diode was replaced (BTW by original HP part from mid 80's), mixer start to work excellent again.
Note: HP's P/N of Dual Mixer used in HP8562A was 5086-7749. Now you can still order from Agilent spare part with the same P/N. BUT!! It is different mixer for current SA's without low-band segment. Just somebody forgot to assign new P/N. Don't ask how I found this information.
Here are more detailed pictures of mixer: Pic2 Pic3 Pic4 Pic5
December 2007: Once during measurement I have noticed 40dB(!) drift in displayed noise level. After observation I discovered relation between turn-on time and noise flickering. It looked that some capacitor is formating after power on or is temperature sensitive. I dismantled power supply board and checked all electrolytic capacitors with ESR meter. One of them was fluctuant. After replacement noise flicering was gone. Other capacitors looked good but anyway I decided to replace all of them by modern Nippon Chemicon types.
EEPROM replacement
Based on several notes on excellent hp_agilent_equipment Yahoo group I decided to replace original EEPROM 2816 (A2U501) by modern 28C16 substitute. It was not necessary but could avoid potential troubles in future.
Fan replacement
After all those years original fan has already become quite noisy. It is standard 12V type so replacement was easy.
Because I spent already so long time inside of SA, I also downloaded firmware from EPROMs. Here you can find Rev 910415 which is last available firmware version for HP8562A. In this archive you can also find firmware history of more 85xx SA's.
4.Sep 2008: Added more detailed pictures of mixer
27.Dec 2007: Power supply restoration
2.May 2007: Added firmware download part.
