Agilent Technologies 8566B Hardware Problems
 
 Question:
 
 What are the potential problems and solutions for the 
 8566B spectrum analyzer?
 
 Answer:
 
 Symptoms of a bad A6A6 First Converter (low band only) 
 Displayed signal level is too low 
 High LO feedthrough 
 Fails Displayed Average Noise test 
 Low band flatness problems 
 The second harmonic of a signal is too high

 Low Band Flatness 
 A mis-match problem between the First Converter and the 
 3.6 GHz Low Pass filter exists which causes 10 MHz to 50 
 MHz flatness to fail. Solution in production was to add 
 an isolator A6A19 (p/n 0955-0214) and a cable A6W24 (p/n 
 85660-20392) between the First Converter and the 3.6 GHz 
 LPF. Service manuals have not been updated yet. 
 Ringing on top of noise floor only at band crossings 
 Solution: A11A1 CIA. Could also have been ACLU. 
 Random 40-50dB noise spikes in high band only 
 Appears that comb generator is connected to analyzer 
 input. Problem is a bad A6A8 YTX (diode bonds). 

 Noise sidebands 10kHz from carrier 
 On older boxes, A7A2 100MHz VCXO caps sometimes 
 touch casting. Solution is to loosen casting cover 
 screws, reposition board then tighten down screws 
 again. 
 If noise floor rise (phase noise) is causing the 
 problem, try changing the A21 FM Coil Driver.

 Intermittent signal amplitude or power supply output low 
 May cause Instrument Check light 1 to remain on at 
 Instrument Preset. Problem is in 85662A IF Display 
 section. Cause is contacts in "MOLEX" connectors 
 (usually the ground wire) on power supply cable 
 assemblies are faulty. The plastic body of the connector 
 is often discolored as a result of the contact arcing. 
 Examine the connector behind the A4A9 IF control board 
 which supplies power to the IF section, and the 
 connector on the A1A10 Display Motherboard which 
 connects the power transformer to the instrument''s 
 power supply section. Replacement contacts are p/n 
 1251-5917. Replacement connector body is p/n 1252-0071. 
 200 second sweep time too slow 
 Remove A16 Scan Generator and replace A16R2 with a 21.5 
 Meg ohm, Agilent part number 0699-0075. 

 Line related sidebands 
 At 360 HZ spacing, fan microphonics may be vibrating 
 A7A2 100 MHz VCXO. Solution is to remove screws from 
 A7A2 casting cover and push or pull board until problem 
 goes away, then install screws. Since the fan can cause 
 microphonics, unbolting the rear panel cooling fan, and 
 rotating it 90 degrees will sometimes fix line related 
 problems on early instruments. This changes the position 
 of the 120V power leads, and changes the microphonic 
 level to sensitive assemblies. The microphonics from the 
 fan are known to cause spurs which are 700 Hz offset 
 from the carrier. A quick check for fan microphonics is 
 to block the fan so it will not rotate, then check for 
 spurs and line related sidebands. 

 YTO unlock 
 A20R16 "big gold resistor" is noisy causing (sometimes) 
 intermittent problems. Will see noise floor rise 10 kHZ 
 to 100 kHz from carrier caused by gold resistor also. 
 M/N Unlock 
 A7A4A2C12 (usually a white "dog bone" capacitor in the 
 lower left of the board) touches a feedthru on the 
 board. Solution is to bend the cap so it no longer 
 shorts to the feedthru. 
 TOI, Low power, low LO feedthru, high noise floor 
 Some symptoms seen in both low and high bands. Problem 
 is that two resistors changed value on the last 
 converter. A6A3R9 and R41. 
 Amplitude variations (but no noise spikes appear) in 
 both low band (100 Hz to 2.5GHz), and high band (2.5 GHz 
 to 22 GHz) 
 Amplitude variations of .2-1 dB occur 5-10 minutes after 
 turn on from cold start. Displayed signal skirts wiggle 
 as if signal had slight modulation. Problem intermittent 
 so you may only see an amplitude jump every 2-3 minutes. 
 To check for spurs, go to frequency span 500 kHz and 
 press MAX HOLD button. Solution: A6A11 Slope Generator. 
 Seen on p/n 85660-60126. Has not appeared on 85660-60238 
 (latest). 

 Amplitude variations with noise spikes 
 If problem appears in both low and high bands, solution 
 may be A6A17 300 MHz Bandpass Filter. Bypass filter to 
 check that spikes and amplitude variations go away. If 
 problem appears only in low band, solution may be A6A4 
 Second Converter. 

 Display blank. No high voltage 
 Check A4A9 IF Control Assy edge connector, P1-3 CRT 
 DSBL, for ~146 mV when CRT is ON and ~3.7V when CRT is 
 OFF. If voltages are correct, troubleshoot A1A6 HV 
 oscillator or A1A3 HV Regulator Assy. 

 Fails Frequency Span Accuracy test or has slight center 
 frequency offset 
 Problem is usually in 1 MHz span setting. Determine if 
 the problem is in the top or bottom box. If the problem 
 is in top box, solution is to re-adjust or replace the 
 ADC board A3A8. If bottom box is the problem, adjust 
 20/30 loop (PLL2). Also Sweep, Dac and Main Coil Driver 
 adjustments can be checked and re adjusted, especially 
 if wider spans are out of spec.
 
 Additional Information