Moseley Associates Inc.
News from the Moseley Archives
Information about Moseley
Professional Support, Updates, and Information
Let us hear from you!

Optimizing Radio Performance
For Hostile Environments

INTRODUCTION

For most installations the RF environments are fairly benign and the SL 9003Q factory defaults reflect this by optimizing sensitivity, spectral efficiency, and delay. However, cases often arise where the RF environment can be downright nasty such as with nearby paging transmitters, strong co-channel and adjacent channel interference, lightening, and the nearby unlicensed ISM band. Moseley prides itself on a reputation for building robust radio products, and the SL9003Q is no exception. The SL9003Q can be easily configured for optimal performance from the benign to the most brutal environments from the front panel. The following discussion will show the user how to configure the frequency, front-end attenuator, QAM mode, interleaver, and pre-selector for best results and tradeoffs that result.

FRONT-END ATTENUATOR

The easiest place to start is with the front-end attenuator. The receiver has a 20 dB variable pin-diode attenuator in front of the pre-amp to protect the receiver from overload when faced with strong in-band and out-of-band undesired signals that find their way past the pre-selector filter (Figure 1.0). This attenuator is controlled from the front panel under QAM RADIO – RX CONTROL to one of three modes, ON/ OFF/ AUTO.

AUTO: (Factory default) In this mode the front-end attenuation is controlled by a leveling loop that begins to insert attenuation in front of the pre-amp when the input signal exceeds –28 dBm. It continues to increase attenuation with increasing input signal up to –8 dBm. In general this mode insures that your receiver will operate with greatest sensitivity and yet provide protection against occasional interfering signals.

OFF: This mode disables the attenuator completely. Use this mode if strong bursty interfering signals are sporadically triggering the attenuator leveling control and causing errors (this is a fairly low likelihood).

ON: This mode forces the attenuator on essentially placing a 20 dB pad in front of the pre-amp. This mode provides the greatest continuous protection against interference but also eats up 20 dB of threshold and fade margin. Use this mode if your received signal exceeds –43 dBm or when strong continuous interferer(s) existing in-band cause bit errors.

Figure 1.1 depicts the typical shape of the BER curve with and without high-level signals in-band. It should be emphasized that it is not necessarily only high-level adjacent channels that cause interference. There are many combinations of signals that can give rise to intermodulation distortion, which cause the resultant product to fall within the desired passband.

ASSESSING INTERFERENCE

This method is very useful to assess interference at your STL receiver (especially if you do not have a spectrum analyzer available).

Turn OFF the STL transmitter at the studio. At the receiver from the front panel navigate to QAM RADIO – MODEM - STATUS. The first line entry "QAM Modem" will indicate the RSL (Received Signal Level) in dBm. With no interference present the RSL will be below –120 dBm, typically. If this is not the case and RSL is above this level then you are receiving undesired interference within your STL passband.

For the QAM data to be properly demodulated at the STL receiver the RSL must be greater than the interference noise floor by the following amounts:
21 dB for 16 QAM
24 dB for 32 QAM
27 dB for 64 QAM
(To determine your QAM mode navigate down 5 more menus under MODEM STATUS until you read "MODE".) For instance, if your STL is operating in 32 QAM mode (i.e., 32Q) and your RSL interference is –90 dBm, then the minimum signal that your STL receiver can acquire must be greater than –66 dBm. Add 10 dB more for fade margin then you will want to see an RSL of at least -56 dBm.

QAM RATE

Let’s say you’ve discovered there is interference within your passband and there is nothing you can do about. You can’t change frequency, you can’t install larger antennae, you’re just plain stuck with it.

From the previous example it is seen that 16 QAM provides a 6 dB improvement in threshold. Basically 16 QAM is more robust than 64 QAM because it spreads its information out more and is easier to decode (there are only 16 states to detect compared to 64). 16 QAM will work down to a 18 dB CNR versus 24 for 64 QAM. Along with this 6 dB sensitivity improvement 16 QAM is more robust against interference, microphonics, and impulse noise such as lightning.

The trade-off is bandwidth. The approximate RF bandwidth for 64 QAM is the DATA RATE divided by 6 versus 16 QAM for which the RF bandwidth is DATA RATE divided by 4. Stated another way, 16 QAM utilizes 33% more bandwidth than 64 QAM, which is a fair price to pay. Though this is a very effective way to increase the radio’s robustness, keep in mind that the IF bandwidth of the receiver is fixed at 300 kHz. The receiver can effectively decode modulation that exceeds this bandwidth but performance will begin to degrade as the transmission bandwidth increases. Changing the transmission bandwidth is left to the users discretion; exercise caution not to exceed Part 74 bandwidth allocation.

To change QAM rate navigate to QAM RADIO – CONFIGURE MODEM – Mode/Effic. Switch from 64Q/6 to 16Q/4. It is imperative to match the QAM mode on both transmitter and receiver or the system will not operate. Don’t forget to change both.

INTERLEAVER

Bit errors may also result from sources other than traditional RF interference. Some of these noise sources include microphonics, lightening bursts, ignition noise, and basically anything bursty in its nature. The problem with these sorts of burst errors is when too many errors appear together the Reed-Solomon error correction algorithm will be unable to correct for them. To combat this phenomenon an interleaver is used to spread out the error bursts (Figure 1.2) so the error correction algorithm doesn’t have to deal with all the errors at once.

The trade off here for using interleaving is added delay. Table 1 shows the correlation between the interleave setting and delay.

Table 1. Interleave Setting vs. Delay

Interleave

Delay* (ms)

1

2.6

2

3.7

3

5

4

6

6

8

12

14

* delay is for 1408 kbps data rate

To change interleave length navigate to QAM RADIO – CONFIGURE MODEM – Intrlv. The factory setting is 3 (5 ms). Just like with the QAM mode setting the user must change the interleave setting to match on both transmitter and receiver or the system will not operate.

PRE- & POST- BIT ERROR RATE MENU

The receiver BER status screen is the most important indicator to the health of the link. From the front panel navigate to QAM RADIO – MODEM STATUS. The first screen that is shown is the "BER POST" and RSL status. "Post" refers to post-error correction count, or the bit-error-rate after Reed-Soloman error correction. This is the actual error rate. It is a long-term error count which means it will reflect every error that has been accumulated since the last time it was reset by pressing enter on the front-panel. The system should be error free (displayed as 0.00E+0) under normal operating conditions but it is quite reasonable to expect occasional errors due to external or environmental conditions. For a healthy link the error rate should not exceed 1.0E-10 (about 1 error in 3 hours at 1 Mbps).

Navigate down one more screen to find "BER Pre". This is the pre-corrected error rate, or the error count before error correction has been applied. There will usually be some none-zero error rate before error correction due to errors caused by nonlinearities within the radio link itself. This is especially true for 64 QAM modulation, which is quite sensitive to amplifier linearity and amplitude and group delay variations. The 16 QAM modulation isn’t nearly so sensitive. Pre-BER is a good indicator of proper circuit operation such as whether the power amplifier is being driven too hard. An increase of only 1 dB above the factory-calibrated level can be enough to cause a substantial pre-corrected error increase. For this reason the power amplifier output level is accurately controlled and compensated over temperature.

CHANGING FREQUENCY

For some types of interference, such as strong co-channel and adjacent channel signals, the only remedy may be to move the carrier frequency away from the interference. This is also a good test to see where the interference lays.

The frequency is changed from the front panel. Navigate to QAM RADIO – TX/RX CONFIGURATION and set the frequency as needed.

Frequency may be programmed from the front panel to +/- 1 MHz of center (factory calibrated frequency) without adjustment. Frequency change beyond this range requires re-tuning the VCO on the up/down converter to the center of its range as follows:

  1. Disconnect AC power cord.
  2. Unscrew the two retaining thumbscrews from the Up/Down Converter Module, disconnect the RF semi-rigid cables from the module and pull the module from its slot.
  3. Plug the VME extension card into the Up/Down Converter slot and plug the Up/Down Converter into extension card. (There is no need to reconnect the RF cables at this point).
  4. Re-apply AC power. You should observe positive lock indications for TX LOCK and RX LOCK (green LED’s) on converter back panel.
  5. On the front panel navigate to QAM RADIO – TX/RX STATUS. Scroll down to the menu that shows AFC. It should be between 4.5 and 5.0 VDC.
  6. Navigate to QAM RADIO – TX/RX CONFIGURATION and set to desired frequency.
  7. Navigate back to QAM RADIO – TX/RX STATUS and scroll down to AFC.
  8. Locate respective tuning access hole on the module, RX AFC on receiver down converter and TX AFC on transmitter up converter and adjust with small tuning screwdriver. A metal jeweler’s screwdriver works fine for this. If the new frequency is higher than the old, rotate adjustment counter clockwise until the synthesizer locks again. If the new frequency is lower, rotate clockwise until lock is achieved. Once the synthesizer is locked, adjust AFC for 4.7 V.
  9. Unplug the radio, remove the VME extender card, and replace Up/Down Converter back into its slot. Reconnect semi-rigid RF cables.
  10. Plug in radio and you should be back in business. This procedure must be repeated at both ends of the link.

FRONT-END BANDPASS CONSIDERATIONS

The pre-selector filter that is shipped with the SL9003Q is a 5-pole inter-digital waveguide bandpass filter. It has been optimized for lowest loss, high ultimate selectivity, and reasonable cost. The bandpass is 20 MHz, which was designed to keep the loss consistent between the inside and outside channel allocations. For most applications this pre-selector should provide the best overall performance,  but for extremely powerful near band interference such as pagers this pre-selector may not provide adequate protection.

Celwave, EMR and Microwave Filter Company are just three of the many companies that manufacture bandpass filters for additional protection in antenna farm environs.

Moseley recently introduced the Starlink Bandpass Cavity for the 944-952 MHz STL band. As you can see by clicking here (Thanks to Dennis Martin at KZLA Los Angeles) the cavity offers over 65 dB protection from potential interference.

 

[ Home | News | Customer Service | Company | Contact ]
[ Radio Broadcast | Television Broadcast | Wireless | Systems ]


© Copyright 1999 Moseley. All rights reserved.