Interpretation of CDMA Tx Analyzer Measurement Results
This section contains expected result values and possible causes of error if the expected results are not met.
Frequency Error
The standards specify very tight Frequency Error performance. In the PCS bands, the 0.05 parts-per-million specification translates to only 99 Hz at a carrier frequency of 1980 MHz or 40 Hz at 850 MHz cellular frequencies. Frequency Error shows up as uncorrelated power that adds to the noise floor or shows up in other Walsh codes.
If a particular site loses its reference to GPS time, its reference signals will begin to drift over time. Phones already using the site can remain on the air because they derive their timing from the signals transmitted by the base station. However, phones using other sites/sectors may be prevented from using the site because they are confused by the error in frequency. This creates what is known as the "island cell effect". By itself, the cell is still functional. To the rest of the system, it's inaccessible.
This island cell effect can be caused by a failure in the site's GPS receiver and timebase distribution network. Using the test set's Internal GPS receiver provides an independent time reference that will allow you to determine if this cell site is out of sync with the rest of the network ("island cell effect").
PN Offset
Verify that the PN Offset is correct. If you are in Manual mode, the PN Offset will display the value you entered. Make sure the Time Offset is small (less than 3 µs). If the Time Offset is greater than expected, see the section on Time Offset below. If you are in Auto mode, the test set will tune to the PN Offset with the smallest Time Offset. If an incorrect PN Offset is displayed, the Time Offset will likely be very large.
Time Offset
The CDMA standards specify a maximum offset of 10 µs; generally 5 µs is a recommended maximum. If the Time Offset is large enough, an "island cell" can occur. The "island cell effect" was mentioned already when discussing Frequency Error.
A phone moving outward toward the limit of its cell will need to acquire the adjacent cell in order to hand off. If the time offset of the target cell is too far from that of the current site, the handoff may not happen.
The dependence of the phone on correct system time limits the physical distance to a target cell's antenna. Each PN Offset is 52.08 µs; if the propagation delay is too long, the received PN Offset may be different from the value designated on the Sync channel. This difference can confuse the phone, causing the handoff to fail.
Another cause of timing error is a bad GPS receiver or timing distribution network within the base station. To test your base station's GPS and timing distribution system, you can use the internal GPS receiver to generate an accurate timing reference for the test set to use while performing base station tests.
Channel Power
When making channel power measurements, make sure you have accounted for the loss of the test cable and high power attenuator you are using. If you do not know the loss of your test cable and high power attenuator configuration, perform an insertion loss measurement and enter the insertion loss as the RF In Loss value. For more information see "Two Port Insertion Loss"
If the channel power is lower than expected, verify you have a good connection to the RF output of the base station. Also, verify the cable you are using to connect to the base station is not faulty. A low channel power may also indicate a bad power amplifier.
Inaccurate channel power (high or low) may indicate an incorrect power setting at the base station.
Estimated Rho
The CDMA base station standard specifies that Rho must be greater than 0.912. Typical values for a healthy base station are greater than 0.94.
Rho failures can indicate problems in:
Because the uncorrelated power appears as interference to the phones, poor Rho performance will affect the sector's capacity. The added interference can require that the signal on traffic channels be raised to overcome the interference. This may, in turn, be seen as further interference. At some point, the site will have to shed calls in order to supply the remaining calls with enough signal versus the interference in the system.
- Compression in linear amplifiers
- Magnitude and phase errors in the IQ modulator
- Phase non-linearity (group delay)
- Spurious signals in the transmission path
- Carrier feedthrough
Carrier Feedthrough
A good carrier feedthrough level is less than -25 dB. The IS-97 standard does not specify carrier feedthrough; however, this measurement provides an additional tool to troubleshoot the base station's transmitter.
In the frequency domain, carrier feedthrough can show up as an uncorrelated energy spike that can be seen on the spectrum analyzer by closely spanning into the top of the CDMA signal. The effects of carrier feedthrough can also show up as higher noise levels on the Code Domain Power screen. The inactive Walsh codes will be pushing the -27 dB specification for noise. Carrier feedthrough can be caused by the lack of isolation across the mixer and cavity of the transmitter's I/Q modulator. Shielding can help reduce carrier feedthrough.
Noise Floor
A good noise floor value is between -40 dB and -45 dB for modern base stations. If the noise floor is above these values (for example, -20 dB), the modulator may be the source of the added noise. Another possibility is a problem with the base station's channel card. The channel card generates the individual Walsh codes.
Pilot Power
If the measured value of pilot power is more than +/-0.5 dB different from the intended setting, it may indicate the following:
- There has been a change made to the digital gain setting of the pilot channel.
- The base station power amplifier has a problem.
Delta Page Power
If the measured value of Delta Page Power is more than +/-0.5 dB different from the intended setting, it may indicate that the digital gain setting for the Paging Channel has been changed.
Delta Sync Power
If the measured value of Delta Sync Power is more than +/- 0.5 dB different from the intended setting, it may indicate that the digital gain setting for the sync channel has been changed.
Delta Quick Page Channel Power
If the measured value of Delta QPCH Power is more than +/-0.5 dB different from the intended setting, it may indicate that the digital gain setting for the QPCH has been changed.