Question:

What does it mean when I get Error 33 or 34, Tuned RF Level Mode, on the HP 8902A?

Answer:

Part 1. SUMMARY OF "ERROR 34" INVESTIGATION AND FIX

Error 33 or 34 results from the IF frequency slewing after RF range switching while performing RF range-to-range calibration in Tuned RF Level mode.

Here's an explanation of the frequency shift and error:

 

  1. The switches that change RF ranges are driven by 20 ms 15 V pulses. Being mechanical, the switches consume a significant amount of current and draw down the 15 V supply slightly.
  2. The 40 V supply, which is referenced off the 15 V supply, glitches even more slightly due to the 15 V glitch. A10C10 decouples the 40 V supply from most of the effects of 15 V noise by combining with A10R33 to create a 250 ms time constant. Since the decoupling is not totally effective, the small 20 ms disturbance in the 15 V supply translates into a small overshoot and recovery transient on the 40 V line lasting a little longer than 1000 ms (1 second).
  3. The 40 volts is filtered again on the A20 LO Control board by a power supply decoupling circuit with a time constant of 100 ms. This 40 V is then used to supply the LF VCXO Tune Amplifier and Filter. A small glitch on the 40 V translates to a small frequency change in the LFVCXO (A22 Assembly). The change with a HP 5371A Frequency/Time Interval Analyzer has been measured . . . almost 1 Hz, and taking over a second to return to its original (nominal) 2MHz frequency. The actual change in the 40 V was never measurable.
  4. This 1 Hz shift in LFVCXO frequency translates to a 600 Hz change in the Agilent 8902's LO frequency when at frequencies around 1200 MHz, since the HFVCO is phase locked to the LFVCXO's 300th harmonic and the LO frequency is the HFVCO doubled. [1 Hz X 300 (harmonic#) X 2 (doubler) = 600 Hz].
  5. After RF range switching, the firmware expects some interruption in the IF and has a short delay, but the frequency slewing occurs well into the period when the instrument is trying to compare the IF level to what was measured in the previous RF range. In some cases, the RF ranges may line up at these frequencies at a critical point where the IF power dropped below the power limit for calibration as the frequency moved out of the IF bandwidth. The instrument would have made a good calibration if it would have returned from the "Is the IF there?" routine successfully, as the power measurement is actually made and averaged after even this long transient.

    To get rid of the transient, reference the 40 V off a 6.2 V temperature compensated Zener diode. While the 15 V still biases the diode, there is no IF frequency shift after switching RF ranges. The 20 ms glitch appears on the IF, due to coupling of the 15 V into the oscillators, but the firmware expects this and deals with it accordingly.

Part 2. Agilent 8902 ERROR 33 or 34 MODIFICATION PROCEDURE

So, you have an 8902A that returns Error 33 or 34 when in the TRFL mode?

Install the following modification and you'll eliminate this failure mechanism. Note that this procedure requires the selection of a 6.2 V Zener diode that will set the output voltage of the +40Vdc power supply within a specific window. Also, it is required to perform three adjustments and one performance test after the modification.

  1. Turn off the unit and remove the A10 assembly. Place it on a static-free work surface.
  2. Modify the A10 as follows: 
     Reference Current New New


 Designator Value Value P/N


 R33 5.11 k ohm 1.21 k ohm 0757-0274


 R37 5.11 k ohm 1.96 k ohm 0698-0083


 R42 17.8 k ohm 28.7 k ohm 0698-3449


 R43 10 k ohm 5 k ohm 0698-6320


 C10 6.8 uF REMOVE-REMOVE-REMOVE-REMOVE


 VR13 None 1902-0680


 Note: Install VR13 in place of A10C10


 so that its cathode is connected


 to the junction of R33 and the


 base of Q13. VR13 will be rev-


 erse biased.


 NOTE: All resistors are 0.125W/1% EXCEPT R43 which


 is 0.125W/0.1%. The Zener is 0.25W/5%.
  3. Re-install the modified A10 assembly back into the 8902A. Turn it on and let it warm up (with the top cover on) for at least 15 minutes. Measure the +40 V power supply. If the reading is not 41.0 +/- 1.0 Vdc (+40.0 to +42.0 Volts) then it will necessary to select another zener diode. Don't be alarmed if takes a few tries, but be careful not to lift the pads on the PC board.
  4. If you have an 8902 Operation and Calibration manual, perform Adjustment 4 [Sampler Efficiency and Offset], Adjustment 5 [LO Doubler Output Power and Balance], and Adjustment 24 [IF Synchronous Detector].
  5. If you have an 8902 Operation and Calibration manual, perform Performance Test 10 [Tuned RF Level (Using a Step Attenuator)].
  6. If you have an 8902A Service manual, refer to Service Sheet 32 for power supply schematic. Modify the schematic accordingly.

(c) Copyright Agilent Technologies 1998, 2001. All Rights Reserved.

This information is subject to change without notice and is provided "as is" with no warranty. Agilent Technologies shall not be liable for any direct, indirect, special, incidental or consequential damages in connection with the use of this material.