Agilent Technologies
Search: in
|
Products & Services
 | |
Industries
 | |
Customer Center
 | |
About Agilent
 |
+ +

Agilent HomeAgilent Home > Products & Services > Test & Measurement > RF & Microwave Instruments & Systems > Network Analyzers > ENA/PNA FAQs
  -

ENA & PNA Series Network Analyzers Frequently Asked Questions

 

If your question or concern  is not answered here, contact your local Agilent field office or on the Web,  contact Agilent  at: www.agilent.com/find/contactus .

PNA Overall-System FAQ

 

The PNA Service FAQ contains information regarding the PNA's operating system (Windows 2000/XP), hard drive, networking, USB drives, administration/security, firmware, LCD display setting, and in general, non-RF or measurement related subjects related to the PNA network analyzers. Measurement related topics are discussed on this page.  

ENA Measurements

  1. Can the ENA make non-linear/mixer measurements?
  2. Does the ENA-L support direct sampler measurements and user ratio measurements?
  3. Does the ENA-L support user-characterized ECal?
  4. The "set Zo" function does not appear to affect the displayed S-parameters of my network analyzer. How does this function work?
  5. Is it possible to improve the print speed of my ENA analyzer?
  6. Is there an outline for configuring my HP Basic to communicate with the Agilent ENA analyzer via SCPI-LAN?
  7. ENA True Differential Measurement Methods
  8. What are the benefits of the built-in VBA?
  9. What are the new features of ENA Revision 3.5?
  10. Blaster Worm FAQ for ENA series
  11. Balanced Cable Measurement using the 4-port ENA

PNA Measurements

  1. I am using a PNA network analyzer (E8362/3/4A or B). How do I know if the network analyzer receivers are compressed?
  2. I am making high-power measurements using the PNA network analyzers. The uncalibrated results seem reasonable, but the calibrated data appears incorrect. What could be the cause?
  3. What is the power level of different measurement channels of the PNA network analyzers at preset?
  4. What is the power of a PNA network analyzer at start-up or preset?
  5. Can different measurement channels on the PNA have different power levels?
  6. On a PNA network analyzer, what happens to the power level when RF power is turned off during a sweep?
  7. I am making high-power measurements. It there a power limitation on the mechanical components of a calibration kit?
  8. I am making high-power measurements. It there a power limitation on electronic calibration or ECal?
  9. What are the benefits of power-meter cal or source-power calibration?
  10. What is the optimum power level for calibration?
  11. On a PNA network analyzer, what happens to the power level at each port during various measurements?
  12. On a network analyzer, what happens to the two-port calibration if the source or receiver attenuation is changed?
  13. On a network analyzer, what does the error message “source unleveled” signify?
  14. What happens to the PNA output power during a re-trace?
  15. On a PNA, what happens to the RF power during frequency-band crossings?
  16. For the PNA network analyzers what is the recommended method to save an instrument state and the associated calibration conditions?
  17. The "set Zo" function does not appear to affect the displayed S-parameters of my network analyzer. How does this function work?
  18. What are the major differences between the PNA A-series and PNA B-series microwave network analyzers?
  19. What are the TTL input/output levels for the PNA rear panel material handler connection?
  20. Does the Agilent PNA family support programmatic recall of the user preset condition?
  21. How to do I control an external instrument connected to the GPIB of an Agilent PNA Vector Network Analyzer?
  22. What are my options regarding high-power measurements with the PNA products?

  23. What is Agilent’s recommended way of calibrating the system for 1 dB compression measurement at different frequencies, let’s say 900, 1800, 2114, 2117 MHz?

  24. How linear is the power sweep? Can we do a source power calibration as a function of frequency and assume that the power sweep is accurate and linear?

  25. I want to store more than 100 cal sets on my PNA, but the PNA does not allow me. What can I do?

  26. Does the 4-port 20 GHz PNA-L have TRL Calibration?

  27. Spreadsheet for calculating the error vector magnitude in intermodulation distortion measurements using the PNA-X

ENA Measurements

1. Can the ENA make non-linear/mixer measurements?
Yes. The ENA with frequency-offset mode can be used to make non-linear/mixer measurements. The ENA-L does not offer frequency-offset mode.

2. Does the ENA-L support direct sampler measurements and user ratio measurements?
The Agilent ENA-L family of network analyzers (E5061A and E5062A) does not support A, B, R direct sampler measurements.

3. Does the ENA-L support user-characterized ECal?
The Agilent ENA-L family of network analyzers (E5061A and E5062A) supports standard ECal procedures. User Characterized ECal is not supported on the E5061A, E5062A. However, the ENA family of analyzers, E5070B and E5071B (with firmware 3.5 and above) does support User Characterized ECal.

4. The "set Zo" function does not appear to affect the displayed S-parameters of my network analyzer. How does this function work?
If the characteristic impedance of the device under test (DUT) is not equivalent to the impedance of the analyzer in use, then the impedance value can be modified using the "set Zo" function.

An example would be performing 75-Ohm DUT measurements with 50-Ohm network analyzer. The effects of changing the value of the system impedance, "Zo", will be realized within the Smith chart format. Specifically the R+jX marker values in the Smith chart mode. Modifications within the analyzer, due to changing "set Zo" are mathematical. There is no change in the inherent physical characteristic impedance of the analyzer. The measured impedance is mathematically normalized to the user entry value of Zo. The center of the Smith chart is now adjusted to the new characteristic impedance value of Zo. Other S-parameter formats are not affected by the change of the characteristic impedance.

Note that when working with waveguide (W/G) based DUTs there exist a special requirement to "set Zo" to "1". This value is a flag to the analyzer to account for W/G dispersion effects when performing certain internal calculations. This is only true for analyzers that support waveguide calibrations. The "set Zo = 1 Ohm" should be entered prior to performing a calibration.

5. Is it possible to improve the print speed of my ENA analyzer?
To improve print speed, ensure that the trigger mode is set to "trigger hold", when performing a screen dump to the printer via either the USB or parallel printer ports. Trigger hold is configured via the Trigger hard key followed by the Hold soft key.

Print speed is improved because the network analyzer CPU is optimized and dedicated for measurements. When the analyzer is configured to a "trigger hold", the CPU resources become dedicated to the printer, offering a significant reduction in print time.

6. Is there an outline for configuring my HP Basic to communicate with the Agilent ENA analyzer via SCPI-LAN?
A) Configure the HP Basic AUTOST to recognize the ENA's LAN Configuration.

The following is a method to establish remote communications via HP Basic to the ENA SCPI-LAN interface on the ENA.

  1. On the PC with HP Basic installed, locate the file HPIBS.
  2. Make a copy of the HPIBS driver and rename as desired. For this example the driver was renamed HPIBS2 and stored in the same directory. HP Basic MUST have an individual copy, with a unique name, for each access of the HPIBS driver.
  3. Near line 330 of the HP Basic AUTOST program, add a new LOAD BIN. Within this LOAD BIN statement configure the HPIBS2 driver to access the ENA's SCPI-LAN. Below is a copy of the configuration utilized for this example.
    335 LOAD BIN "HPIBS2;DEV lan[AGILENT-E45B822]:hpib9 ISC 9 TIME 5"
    350 LOAD BIN "HPIBS;DEV hpib7 ISC 7" ! SICL driven GPIB boards
  4. Line 335 is the additional GPIB configuration added to communicate with the ENA via SCPI-LAN. Note the use of the following within this command line.
    i. Call to "HPIBS2" for the driver
    ii. "DEV" or device is "lan"
    iii. Use of the alias, or computer name, "AGILENT-E45B822". You may choose to substitute the ENA's IP address here. If your information technology group assigns IP addresses dynamically the use of a direct IP addresses may cause problems if the ENA's IP address is reassigned.
    iv. Reference to "hpib9" - This is a MUST. The default SCPI server in the ENA is GPIB2 and the SICL name is "gpib9". The entry in the HP Basic AUTOST LOAD BIN must match exactly. To check the ENA configuration select
    1. SYSTEM (hark key - HK)
    2. Misc Setup (soft key - SK)
    3. GPIB Setup (SK)
    4. System Controller Configuration (SK)
    This now results in the I/O Configuration window for the Agilent I/O Libraries (at the ENA).
    For VISA Name "GPIB2" the SICL Name is "hpib9", ensure this is true. Within the I/O Libraries configuration window at the ENA, select the GPIB2 interface and press the EDIT button. The title bar should indicate "Internal Instrument Configuration", Logical Unit "9", and SICL Interface Name "hpib9".
    Exit up two levels and ensure the Talker/Listener Address is 17.
    v. The "ISC" or interface select code in line 335 is "9", which matches the "Logical Unit" in the ENA's I/O configuration.
  5. Note that LINE 350, referenced above, is a fully functional call to the original HPIBS driver for communications to the Agilent 82350A PCI card embedded in the PC. It has no relation to the ENA configuration The reference is intended to illustrate that HP Basic can support multiple GPIB interfaces. In this example line 335 for a LAN based SCPI device, the ENA, and line 350, an embedded Agilent GPIB card.
  6. Ensure you "RE-SAVE" or "RE-STORE" your HP Basic AUTOST file so that the modifications are stored.

B) The sample below is an HP Basic program that communicates with both the configured ENA via SCPI-LAN and to and Agilent 8753ES via the embedded Agilent 82350 PCI-GPIB interface. The program is a test to ensure the LOAD BIN statements are defined correctly and GPIB communications exist between both configured interfaces.

10 RE-SAVE "ENA_SCPI_LAN"
20 CONFIGURE SAVE ASCII OFF
30 DIM Myid$[40]
40 OUTPUT 716;"outpiden"
50 ENTER 716;Myid$
60 PRINT "This is my ID 1 :",Myid$
70 OUTPUT 917;"*IDN?"
80 ENTER 917;Myid$
90 PRINT "This is my ID 1 :",Myid$
100 END

Note the call to the 8753 is via the PCI-GPIB at select code 7 and device address 16. The call to the ENA via SCPI-LAN is at select code 9 and device address 17.

Below is a screen snapshot of the resultant program execution.

results_ENA_SCPI_LAN.jpg

This document may also be downloaded as a PDF: How to configure HP Basic to communicate with the Agilent ENA analyzer via SCPI-LAN

7. ENA True Differential Measurement Methods
Please see the attached file for details.

8. What are the benefits of the built-in VBA?
1. Built-in VBA on the ENA Series improves your network measurement environment.The product note, "Evolution of Test Automation Using the built-in VBA with the ENA Series RF Network Analyzers", explains how the built-in VBA can improve your network measurement environment. Practical examples of the effective use of the built-in VBA are given in the product note. For more information

2. VBA program launcher allows you to enhance the ENA's capability with your own VBA programs.

For more detail about the VBA program launcher, please refer to the ENA FAQ, "What are the new features of ENA Revision 3.5?"

3. VBA sample program library saves your development time.

The VBA sample program library provides various useful programs such as a filter measurement program, data saving utility, calibration wizard and more. By downloading sample programs, you can modify a sample program to meet your specific needs and save your development time compared to making programs from scratch.

For specific ENA Series product detail regarding the VBA sample programs, visit the web links below:

For the E5070B
For the E5071B
For the E5070A
For the E5071A

For additional ENA Series product information visit the ENA Series of Network Anlayzers' Web page.

9. What are the new features of ENA Revision 3.5?
See the attached file for details.

10. Blaster Worm FAQ for ENA series
Question 1:
The E5070B/E5071B was infected with the virus. In order to return to a normal state, what should I do?

Answer 1:
Please remove a LAN cable immediately and perform "System Recovery". The procedure is described in the user's guide

After that, please run virus scan using an external PC.The procedure of virus scan is described in the appendix of the service manual.

If the operation of the E5070B and E5071B does not return to be normal, please contact to Agilent service center for the repair (charged).

Question 2:
Can I install anti-virus software on the E5070B/E5071B?

Answer 2:
No, you cannot. Since anti-virus software might affect the measurement performance, do not install any anti-virus software on the E5070B/E5071B, nor make it run. Instead, enable to access the hard disk drive (HDD) of the E5070B/E5071B from an external PC, and then scan the HDD with anti-virus software installed on the PC. Please refer to the user's guide for enabling the access to the HDD from the external PC.

Question 3:
If I install the patches dealing with Blaster Worm on operating system (OS) of the E5070B/E5071B, do the instruments work normally? Moreover, please let me know the installation procedure of security patches.

Answer 3:
Agilent tested and confirmed that the E5070B/E5071B work normally under the following environments a) or b).

a)Installing the following service pack and patches on the E5070B/E5071B on which the service pack 3 is installed.

- Windows 2000 Service Pack 4
- August 2003, Cumulative Patch for Internet Explorer 5.01 for Windows 2000 Service Pack 4 (822925) (q822925.exe)
- MS03-026: Security Update for Windows 2000 (823980) (Windows2000-KB823980-x86-ENU.exe)
- 823559: Security Update for Microsoft Windows (Windows2000-KB823559-x86-ENU.exe)
- 816093: Security Update Microsoft Virtual Machine (Microsoft VM) (msjavwu.exe)
- Security Update for Microsoft Data Access Components (823718) (Q823718_MDAC_SecurityPatch.exe)

b)Installing the following patches on the E5070B/E5071B on which the service pack 2 is installed.
-MS 03-026: Security Update for Windows 2000 (823980) (Windows2000-KB823980-x86-ENU.exe)

<Note> To install the service packs 3 or 4 on the E5070B/E5071B on which the service pack 2 is installed, hard disk drive exchange is needed. The exchange should be performed by the service center (charged).

Here is the procedure to install the above-mentioned security patches on the E5070B/E5071B. In addition, only IT engineers or persons based on IT skill should perform installation of patches.

Step 1. Perform System Recovery of the E5070B/E5071B.

Step 2. Check the version of the service pack installed on the E5070B/E5071B. Push the "Explorer" soft key in Save/Recall menu. Then, Windows Explorer starts. Click "About Windows" under the Help menu to display the version of the service pack. how to access the HDD from an external PC.

Step 3. Download the above service pack or patches from the Internet homepage of Microsoft Corp. Be sure to download them using an external PC.

Step 4. Copy the service pack or patches to the E5070B/E5071B via a floppy disk or LAN. See the user's guide for

Step 5. Close the measurement screen of the E5070B/E5071B before installing patches.

Step 6. Push the "Explorer" soft key in Save/Recall menu. After Windows Explorer starts, please install the service pack or patches on the E5070 B/E5071B.

Question 4:
Isn't the E5070A/E5071A influenced of Blaster Worm? Are any measures to Blaster Worm necessary?

Answer 4:
The E5070A/E5071A carries Windows 98 operating system. As Microsoft Corp. announced that Blaster Worm does not affect Windows 98, Agilent has judged that the measure to Blaster Worm is unnecessary for the E5070A/E5071A.


11. Balanced Cable Measurement using the 4-port ENA
See the attached file for details.

PNA Measurements

1. I am using a PNA network analyzer (E8362/3/4A or B). How do I know if the network analyzer receivers are compressed?
When testing active devices, especially amplifiers, users should pay attention to the output power levels of their devices and the power incident upon the network analyzer’s receivers. Using the receivers in compression can make it difficult to differentiate between device compression and test system error. The procedure below describes a method to determine whether the internal network analyzer receivers are compressed or not. This procedure requires that the network analyzer be equipped with receiver attenuators.

On the MW PNA analyzers, receiver attenuators are available with Option 016. Receiver attenuators are not available for the lower cost MW PNA-L models. Connect your test device between ports 1 and 2 and set up an S21 and B channel measurement. Then change the receiver attenuator settings and examine the S21 and B. If the receivers are not compressed, the traces should only vary by the amount of

attenuation, and not have other variations. If the receivers are compressed, you will see change other than the exact amount of attenuation. You can make the comparisons without calibration. Just make sure calibration is off in all cases. Markers can be helpful to determine if the values have decreased by the attenuation amount. Repeat this test for the R channel receiver also, since S21 and AM-PM are both ratioed measurements and thus both receivers need to be tested. In an intermodulation distortion measurement, you can make the same attenuation change, but monitor the difference between the fundamental tone and the mixing products (the dBc values). If the dBc values change with attenuator setting, you can suspect that the PNA receivers are compressed.

If the test shows that the network analyzer receivers are compressed with the original settings, increase the receiver attenuation levels until the point that the receivers are no longer compressed.

On the MW PNA, the receiver attenuators can be controlled from the Channel > Power menu.

2. I am making high-power measurements using the PNA network analyzers. The uncalibrated results seem reasonable, but the calibrated data appears incorrect. What could be the cause?
A two-port calibration calculation is based on all four S-parameters. One possible issue in high-power measurements is that the S12 measurement could have high uncertainty due to noise, if the port powers are not uncoupled. When measuring high-gain amplifiers, it is recommended that you take advantage of the Port Power Coupled feature to uncouple the power of ports 1 and 2. Drive the input or port 1 with a low power level as to not damage the output receivers. Drive the output or port 2 with a high power level, so the isolation or S12 measurement does not approach the noise floor of the network analyzer. An accurate S12 measurement is fundamental to an accurate 2-port calibration.

3. What is the power of a PNA network analyzer at start-up or preset?
At preset, the source power level of the MW PNA port 1 is set to a nominal level (see Table 3), with the internal source attenuator on port 1 set to 0 dB. The port 2 power is off. On the PNA analyzers, only one port is on at a time. If the amplifier under test could be damaged by this power level, or will be operating in its nonlinear region, do not connect the amplifier until you have set a desirable power level. On the MW PNAs, you can save a “user preset” with different initial power setting conditions. Upon preset, the MW PNA starts with the new power levels.

Nominal power levels (Preset power level at port 1)

Network analyzer

Standard

Option 014, UNL or 014 &UNL 

E8362B (20 GHz)

0 dBm

–5 dBm

E8363B and E8364B (40 and 50 GHz)

–12 dBm

–17 dBm
 

4. What is the power level of different measurement channels of the PNA network analyzers at preset?
 Each channel is initiated with the nominal power level, even if a User Preset is saved with a different power level for the starting channel. Therefore, if you save channel 1 with a nominal power level of –60 dBm as a User Preset, then start channel 2, channel 2 will start with the nominal power level (–17 dBm for an option loaded E8364B). Be careful if you set up a new channel. You could damage the components, if you did not anticipate the high power levels. If you had RF power Off at the User Preset level for channel 1 and then you started a channel 2, then RF power would be off on channel 2 also. RF Power is a global parameter, versus the power level setting, which is a channel parameter.

5. Can different measurement channels on the PNA have different power levels?
 Yes. Different PNA measurement channels can have different power levels. If you set up two channels with different enough power levels resulting in different attenuator settings, the PNA will automatically put one channel in trigger hold mode. This is to protect the attenuators from switching continuously.

6. On a PNA network analyzer, what happens to the power level when RF power is turned off during a sweep?
 The power level is turned off at the end of the sweep, so the sweep will continue with RF power on. The next sweep will start with RF power off.

7. I am making high-power measurements. It there a power limitation on the mechanical components of a calibration kit?
 The open or short standards do not have a power limitation, as they do not dissipate significant energy. Most Agilent calibration kit loads have a maximum average power rating of 2 Watts or +33 dBm.

8. I am making high-power measurements. It there a power limitation on electronic calibration or ECal?
 The maximum power rating for ECal modules is either +10 or +20 dBm (see Table below). The ECal module also has a minimum power requirement for auto-orientation (not calibration, but orientation). If the power level at the module is less than –18 dBm, the user has to tell the analyzer the orientation of the ECal module. Simply unselect the automatic orientation check box and manually indicate to the analyzer how the ECal module is connected. Auto-orientation means that the network analyzer determines how port 1 and 2 are connected to ports A and B of the ECal module.

ECal model

Minimum power

Maximum

Maximum DC

RF power

Voltage at test port

N469x (MW ECal)

No minimum power for calibration. See paragraph above for minimum power level for auto-orientation.

+10 dBm

±10 Volts

8509x (RF ECal)

+20 dBm

±20 Volts

9. What are the benefits of power-meter cal or source-power calibration?
 A source-power calibration transfers the accuracy of the power meter measurement to the network analyzer. The output power of the network analyzer is accurate to within 2-3 dB. For the MW PNAs, the specifications are listed in the table below. A power meter calibration can provide accuracy of better than 0.5 dB. 

MW PNA power level accuracy specification Variation from nominal power in range 0 (step attenuator at 0 dB setting)

Frequency range

Standard

Option 014

Option UNL

Option 014 & UNL

10 MHz to 45 MHz

±2 dB

45 MHz to 10 GHz

±1.5 dB

10 to 20 GHz

±2 dB

20 to 40 GHz

±3 dB

40 to 45 GHz

±3 dB

±3.5 dB

±3 dB

±3.5 dB

45 to 50 GHz

±3 dB

±4 dB

±3 dB

±4 dB

In a linear S-parameter measurement, where the amplifier is operating well within the linear range, 2-3 dB of power variation may not make a significant difference. But if you are testing and specifying gain compression and trying to find the 1 dB compression point, 2-3 dB is a significant difference and source-power calibration is necessary. Another instance where it is critical to perform a source-power calibration is in the case of high-power measurements, where a pre-amplifier is used. Also, a source power calibration is necessary prior to a receiver calibration (in order to establish the reference). Receiver calibrations are useful for absolute power measurements.

10. What is  the optimum power level for calibration?
 In general, a calibration should be performed under the same stimulus/response conditions as the measurement. Thus calibrating at one power level (without amplifier) and then measuring at a different power level (with amplifier) is not ideal. However, the dynamic accuracy of the MW PNA products is extremely good, so that calibrating at a different power level does not pose a significant error (See chart for frequencies less than 20 GHz). The choice the user does have is to stay within the same power range (same attenuator setting), but to calibrate at a higher power level (without the amplifier) and then reduce the power level during the measurement (with the amplifier). Since the hardware setting is essentially the same, the accuracy is hardly affected. It is always better to calibrate at a higher power level (staying below compression), to reduce the uncertainty due to noise.For best measurement accuracy, select the measurement and calibration power levels such that the test setup power levels remain in the relatively flat region.

11. On a PNA network analyzer, what happens to the power level at each port during various measurements?

Parameter

Port 1

Port 2

Notes

S11

On

Off

S21

On

Off

S22

Off

On

S12

Off

On

Any parameter with two port cal; S11, S21, S12 or S22

On/Off

On/Off

Power switches between the two ports, as a two-port cal requires all four S-parameters.

RF power off

Off

Off

RF power is a global parameter and is turned off for all ports and channels.

12. On a network analyzer, what happens to the two-port calibration if the source or receiver attenuation is changed?
 The calibration is invalidated if you change the attenuator settings. If you change attenuator settings after you have performed a calibration, you must perform another calibration.

13. On a network analyzer, what does the error message “source unleveled” signify?
An unleveled error message appears when the source power is set to value greater than the maximum specified power. Lower the power level to solve this problem. The unlevel message is combined with a LVL indication on the status bar. The unleveled error message can momentarily appear between attenuator settings. It does not affect the measurement accuracy and can be ignored.

14. What happens to the PNA output power during a re-trace?
 The power level is maintained during re-trace, unless a frequency band is crossed. See the next question for frequency-band crossings.

15. On a PNA, what happens to the RF power during frequency-band crossings?
The MW PNAs have over twenty frequency bands. During band-crossings, the firmware turns off the RF power. Beware that if you are testing a high-gain device with ALC, when the PNA switches bands, the power shuts down and the DUT’s ALC attempts to increase the gain. Microseconds later, the PNA power comes back on; however, in this short time frame, the DUT or the PNA can get damaged. The band-crossings are listed below.

Model

Band

Frequency range (GHz)

E8362/3/4B

0

0-0.045

1

0.045-0.748

2

0.748-1.5

3

1.5-3

4

3-3.8

5

4-4.5

6

4.5-4.8

7

4.8-6.0

8

6.0-6.4

9

6.4-7.6

10

7.6-10

11

10-12

12

12-12.8

13

12.8-15.2

14

15.2-16

15

 16-20

E8363/4B

16

20-22.8

17

22.8-25.6

18

25.6-30

19

30-32

20

32-36

21

36-38.4

22

38.4-40

E8364B

23

40-45.6

24

45.6-48

25

48-50

16. For the PNA network analyzers what is the recommended method to save an instrument state and the associated calibration conditions?

The following description applies to all PNA analyzers with firmware release greater than 2.0. The table below summarizes types of calibration and state files that can be saved on the PNA, as well as a short description.

Types of Calibration and State Files
File Extension or Type Notes
*.cal PNA_calibration_conditions.jpg
*.sta Each *.sta file saves everything except calibration information.

State files contain the following Instrument State data:
· Measurements including limit lines and markers
· Arrangement of windows
· All stimulus values, start frequency, stop frequency, power, sweep type, etc
*.cst The PNA *.cst file saves instrument state information and a pointer to a cal set. The actual calibration data remains in the instrument. Only a pointer to the cal set is included in the .cst file. The pointer is not actually to the cal set name ("CalSet_32", for example), but instead it points to the "GUID". The GUID is a long, unique hex identifier.

If a *.cst file is saved the related cal set is not protected and can be lost. For this reason the following is the recommended method to save an instrument state and calibration data.

  1. Perform the required calibration.
  2. When complete save the state file as a *.sta file.
  3. Immediately save the calibration file as *.cal file.
  4. The files can now be transferred to any other PNA or simply archived on the existing PNA for future use.

To recall a state file and calibration file perform the following sequence:

  1. Recall the state file *sta (a warning message will be generated and displayed on the PNA).
  2. Recall the cal file *.cal (a message indicating that the calibration must be turned on will appear).
  3. Activate the calibration.

For additional information on calibration sets within the PNA refer to the embedded PNA help file.

17. The "set Zo" function does not appear to affect the displayed S-parameters of my network analyzer. How does this function work?

If the characteristic impedance of the device under test (DUT) is not equivalent to the impedance of the analyzer in use, then the impedance value can be modified using the "set Zo" function.

An example would be performing 75-Ohm DUT measurements with 50-Ohm network analyzer. The effects of changing the value of the system impedance, "Zo", will be realized within the Smith chart format. Specifically the R+jX marker values in the Smith chart mode. Modifications within the analyzer, due to changing "set Zo" are mathematical. There is no change in the inherent physical characteristic impedance of the analyzer. The measured impedance is mathematically normalized to the user entry value of Zo. The center of the Smith chart is now adjusted to the new characteristic impedance value of Zo. Other S-parameter formats are not affected by the change of the characteristic impedance.

Note that when working with waveguide (W/G) based DUTs there exist a special requirement to "set Zo" to "1". This value is a flag to the analyzer to account for W/G dispersion effects when performing certain internal calculations. This is only true for analyzers that support waveguide calibrations. The "set Zo = 1 Ohm" should be entered prior to performing a calibration.

18. What are the major differences between the PNA A-series and PNA B-series microwave network analyzers?

The PNA A-series, E8362A (20GHz), E8363A (40GHz) and E8364A (50GHz) units are obsolete and have been replaced by the PNA B-series. The E8361A is a 67GHz PNA and is an active product. The following information is NOT applicable for the E8361A 67GHz PNA. 

The information below pertains to the PNA A-series, E8362A (20GHz), E8363A (40GHz) and E8364A (50GHz) analyzers.

  • None of the mixer test options of the PNA B-series are available on the PNA A-series (Option 080, frequency offset mode; Option 081, reference switch, Option 083, frequency converter application).
  • The minimum frequency of the PNA B-series is 10 MHz whereas it is 45 MHz for the PNA A-series.
  • New CPU board in the PNA B-series includes a faster processor and standard 256 MB RAM
  • PNA-B series added a 4-port USB hub on the rear panel
  • PNA-B series provides external edge triggers via rear-panel BNCs, supported by a remote interface only
  • PNA-B series support the new pulse and antenna options (H08/H11) which are not available on the A models

There were many firmware features introduced at the same time as the PNA B-series. All of the features that are not dependent upon the new hardware outlined above will work on the PNA A-series after upgrading the firmware.

19. What are the TTL input/output levels for the PNA rear panel material handler connection?

  • TTL output: high = 2.5 V minimum, 5.0 V maximum; low= 0.55 V maximum
  • TTL input: high = 2.0 V minimum, 5.0 V maximum; low = 0.80 V maximum, -0.5 V min

At the time of this publication the noted TTL values were based on the #74ABT16244 CMOSS chip set.

20. Does the Agilent PNA family support programmatic recall of the user preset condition?
The PNA supports recall of the user preset via the MMEM:LOAD *.STA file command.

For some instruments, there is a direct SCPI command in the form of SYST:PRES:TYPE [FACT|USER]. The PNA does NOT support this structure.

For the PNA, use the MMEM:LOAD *.STA to recall the "UserPreset.sta" file.

To determine where the user preset is stored, access the PNAs SYSTEM tab key followed by the USER PRESET selection. For the PNA tested, the user preset file "UserPreset.sta" was stored as follows:
C:\Program Files\Agilent\Network Analyzer\Documents\UserPreset.sta

Via the programming language of choice, send the command:
"MEMM:LOAD ´C:\Program Files\Agilent\Network Analyzer\Documents\UserPreset.sta´"
in order to recall the user preset via SCPI.

21. How to do I control an external instrument connected to the GPIB of an Agilent PNA Vector Network Analyzer?
The following outlines the procedures to control a GPIB device either locally via the PNA or from a remote PC.
PNA GPIB Control Configuration

Within the PNA System tab, choose Configure - SICL/GPIB. Configure the SICL/GPIB per the image below, selecting the "SYSTEM CONTROLLER" and SICL options.

Control an External Device via PNA_1

Minimize the PNA application (select: View - Minimize Application). Access the Agilent I/O Library Control (blue "IO" icon on the task bar). Select: Run I/O Config. The application should look similar to the image below.

Control an External Device via PNA_2

Note that "hpib7 GPIB1" is the SICL name of the PNA, "gpib0 GPIB0" is the internal PNA GPIB interface. To view additional details related to a configured interface, highlight the interface and select the "Edit" button. The "hpib7 GPIB1" should reference the "Internal Instrument" whereas the "gpib0 GPIB0" will reference the "GPIB Interface Connection" (a.k.a. "SICL on NI GPIB")

To confirm proper operation of the "gpib0 GPIB0" interface connect a GPIB instrument to the PNA. Next, via the I/O Libraries Control icon (on the PNA task bar) execute the "Run VISA Assistant" application. The application should locate and indicate the instrument(s) found on the PNAs GPIB.

The instrument in this configuration is an E4426B with address 18. Using the Formatted I/O tab you can send an *IDN? to identify the instrument (ensure the IEEE 488.2 radio button is selected). Note on the left column the SICL name related to the instrument (GPIB0).

Control an External Device via PNA_3

Once this communication is verified the PNA is correctly configured to communicate with any instrument connected to the GPIB interface.

Remote PC GPIB Control Configuration (via PC Lan to PNA)

The following image is an Agilent I/O Control Library configuration from the remote PC. The PC will access both the PNA internally and the PNA´s GPIB (both via the PNAs LAN).

Control an External Device via PNA_4

To communicate with the PNA and the GPIB bus of the PNA you must create two separate VISA LAN sessions, GPIB0 and GPIB1, on the remote PC.

GPIB0 is the session that it is able to communicate with the PNA GPIB bus. The remote hostname or IP address is the PNA IP address and the remote SICL name is Gpib0 in consequence of what we set in the PNA.

Control an External Device via PNA_5

GPIB1 is the VISA LAN session to communicate with the PNA application.
Please note the SICL name that is in this case hpib7 as set in the PNA I/O config.

Control an External Device via PNA_6

After configuring the remote PC´s I/O Configuration, run the Agilent VISA assistant (on the remote computer) you should be able to see both instruments (the PNA and the external instrument connected via GPIB).

Control an External Device via PNA_8

Where GPIB0::18::INSTR is the e4426B and GPIB1::16::INSTR is the PNA application. Refer to this syntax to open a VISA session in your programming language environment and communicate with the instruments.

23. What is Agilent’s recommended way of calibrating the system for 1 dB compression measurement at different frequencies, let’s say 900, 1800, 2114, 2117 MHz?

The most accurate method involves setting up four channels, with power sweep, and the CW frequency set to the above frequencies.  You can set up a S21 trace and B trace. Perform a source power calibration on all four channels. Perform receiver cal on all channels, for the B trace. Perform a 2-port cal on all channels, for the S21 trace. Now this is a lot of calibrations, but if you have ECal, it's pretty simple. It's really only three sets of connections. You connect the power sensor, source calibrate channel 1, then 2, then 3, and then 4. No need to disconnect or re-connect anything. Just click through the channels. You make the through connection, perform receiver cal, click through the channels. You connect ECal, click through the channels. 

Now all of them are accurately calibrated and you can make your 1 dB compression measurement using the S21 trace (read both x and y-axis), or read the value off the B receiver. The B receiver measurement is redundant, since we did a source cal and the x-axis of the S21 trace is accurate. So you can just add the input power from the x-axis to the gain from the S21 marker value and you have the output 1 dB compression point. But sometimes it is nice to see the B trace also. If measurement time was critical, then the B measurement could be eliminated.

If you are willing to slightly sacrifice accuracy for time savings, you can save time by using the copy channel feature of the PNA. You can perform a 2-port cal, frequency-sweep, pick a power in the power sweep range, and calibrate from 900 to 2117 MHz.  Then copy it three times. Your 2-port cal is copied. This is definitely the easiest method, if you have a mechanical cal kit, since it would be too time consuming to repeat a mechanical calibration four times. Similarly, you can do a source power cal at let's say 1800 MHz and copy it over. But in this case, you can always leave the sensor connected and just click through the channels. It's a time issue, and selecting the "use reference receiver" choice in the source power cal menu saves a lot of time.

24. How linear is the power sweep? Can we do a source power calibration as a function of frequency and assume that the power sweep is accurate and linear?

It depends on the power linearity specification of the network analyzer. For the N52300-020 PNA-L network analyzers, it is ±1.5 to 2 dB. So theoretically, you could be up to 2 dB off, which is significant.

Most analyzer, in a narrow frequency range, and narrow power range, will have better linearity. For example, a quick test of one of the analyzers at the factory showed that for the above frequency and power setting, the linearity error was less than 0.1 dB.  You can perform an experiment (described below) to determine the linearity in your setup.

Set the trace to S21, frequency sweep, with a power level in your power sweep range, let's say -15 dBm, 1800 to 2117 MHz. Perform a source power cal at -15 dBm. Then copy channel 1 to channel 2. On channel 1, change the sweep type to power sweep, with a CW frequency of 1800 MHz, and the desired power sweep range (let’s say –30 to 0 dBm).  On channel 2, change the sweep type to power sweep, set the sweep range to the same as channel 1 (-30 to 0 dBm), and perform another source cal. So now on channel 1, you have interpolated source power cal (*), and on channel 2, a fully calibrated source power cal for your power sweep range. Compare the results. This will tell you what the difference is between performing a source power cal over the entire power sweep range, versus calibrating at one point and interpolating over the rest of the range.

25. I want to store more than 100 cal sets on my PNA, but the PNA does not allow me. What can I do?

2 port PNA and PNA-L network analyzers with firmware revisions 4.86 or less have a 100 calset limit. This does not apply to the 4-port PNA-L network analyzer.

To address the 100 calset limit in the PNA, the user can use a workaround which entails managing a set of .cal files that correlate to .cst files.  You can use the "Save as" function to save .cst and .cal files. This is a description of the file types.

  • *.cst - saves both Instrument state and a reference to the Cal Set data
  • *.sta -  saves instrument state only (no calibration data)
  • *.cal - save Calibration data only (no Cal Set)
 The PNA Help System has a detailed description of these file types. Search on ".cst" in the help system to find the information.

When a .cal file is saved, it contains the set of currently active calsets.  It contains no other state.  So when a .cal file is recalled, it simply adds the calsets from the .cal file to the set of available calsets.  It does not affect the current instrument state.  The exception is that if any of the calsets being recalled have the same ID as a calset being used by the PNA at the time of recall, that calset will be overwritten.  This of course will have an affect on the corrected data for measurements that are using that particular calset.
 
To manage sets of .cal/.cst files, the user would:
 
a) Set up the desired instrument state.  Perform calibrations.  Save the calsets (eg:   foo.cal).  Save the instrument state (eg: foo.cst).
b) Set up the next set of states. Perform calibrations. Save the calsets (eg:  foo2.cal).  Save instrument state (eg: foo2.cst).
c) Perform step (a) for all possible states.
 
To perform measurements using the above:
a) Recall foo.cal.  Recall foo.cst.  (the order is important)
b) Make measurements
c) Recall foo2.call.  Recall foo2.cst.
d) make measurements.
e) and so on.
 
Keep in mind that every recall of a .cal file will attempt to append calsets to the working set of calets.  Thus you will eventually hit the 100 limit again.  So to avoid reaching the limit, precede the recall of .cal with a command to delete all calsets.  This way you won't hit the limit.

Last update:  June 28, 2007