SPOTLIGHT ON: SPEAKER LINE MONITORING

One of the major differences between a Public Address system and a Voice Alarm system is the need to monitor every part of the system for potential failure.  The most substantial and the most vulnerable part of a system is the loudspeaker circuits themselves.  There are a number of methods used by different manufacturers to monitor the health of the 100V loudspeaker circuits.

DC Monitoring

The DC monitoring method is considered the simplest to implement and may be familiar to those used to fire alarm sounder circuit monitoring.  It works by introducing a DC voltage across the circuit via an internal pull-up resistor.  A terminating resistor is installed across the circuit at the end of the circuit.  The presence of the resistor pulls the voltage down to a set voltage that is measured  by a monitoring circuit in the central equipment.  If the terminating resistor is disconnected or a short circuit develops across the line then the monitoring circuit will detect a change in the measured voltage.  The simplicity of this method and the fact that resistors are pretty much bulletproof means that DC monitoring is very reliable.  It also has the added advantage that it can support multiple circuit branches, by simply re-calibrating the expected measurement voltage.

However, it’s not all good news.  The major problem with the DC method is that every loudspeaker on the line must be fitted with a capacitor.  The capacitor must be fitted in series with the speakers transformer.  At DC the coil of the transformer represents a low resistance, hence if installed without a capacitor it would appear as a short-circuit across the line.

End Of Line Monitoring

The method of End Of Line (EOL) Monitoring varies slightly depending on manufacturer, but they all depend on installing an active EOL device at the end of each circuit.  The device is usually powered by either a sub-sonic (30Hz) or super-sonic (20kHz) tone that is injected on the line.  The tone can either be continuous or pulsed.  If pulsed it must be frequent enough to satisfy the requirements of the standards.  BS5839 pt 8 states that a new fault must be reported within 100 seconds of the fault occurring.  The tone is usually mixed with the input signal of each amplifier.  When the tone is present on the circuit, the EOL device is powered and signals back to the central equipment that it received the monitoring tone.  Again, this can be done in a number of ways.  Some systems require that another signal pair be wired back to the central equipment, but this isn’t ideal due to the installation costs involved, so most systems inject a signal back onto the line that is recognised by a detection circuit.  Some systems present a common mode voltage between the circuit and earth, which is not only used for detecting short circuits to earth but can also be loaded to earth by the EOL device to signal that the monitoring tone was detected.   Similarly, some devices generate a current between the line and earth that is detected by the central equipment.

So, why would we use EOL monitoring instead of DC?  Firstly, there is no need for capacitors to be installed in each loudspeaker.  It can also be argued that it is a more thorough test of the circuit.  While DC monitoring effectively tests the integrity of the cable itself, EOL monitoring tests the circuit for its ability to pass an audio signal.  It also tests the amplifier that is driving the circuit and can give an indication of if the amplifier is overloaded or not.  If the load is too high then the amplifier will not be able to produce a tone that is sufficient to power the EOL device.

There are disadvantages.  Each loudspeaker circuit must be wired as a radial, so only one EOL per circuit.  Any branches off the main circuit will not be monitored.  A common problem is noise.  The monitoring tone used must be a pure tone, as any harmonics in the signal will most likely be heard as noise.  So it is imperative that the amplification is up to the job of producing the tone accurately and without distortion.

Addressable End Of Line Monitoring

Addressable EOL monitoring is very similar to standard EOL monitoring as it will still be powered by a monitoring tone (most likely a continuous one).  The difference being that each device has a unique address (obviously).  The central equipment will poll the circuits requesting an acknowledgement for each address by imposing a data signal onto the existing line.  This data signal is usually modulated at somewhere between 40-60kHz.  This is not part of the amplified monitoring tone but usually added after the amplification stage.  If the device recognises its own address then it will impose its own signal response on to the line back to the central equipment.

The advantage of this method over regular EOL monitoring is that you are no longer restricted to radial circuits and can therefore have multiple branches each with an addressable EOL.  The same noise issues can still occur however.  Also it is important that installation is well organised and that the addresses are well documented to avoid duplicate addressing.  This is especially important if the EOLs are to be fitted in hard to reach positions!

Return Loop Monitoring

The return loop method is quite self-explanatory.  Instead of each circuit being wired as a radial, they would be wired back to the central equipment in a loop topology.  This isn’t particularly common due to the obvious additional cost and availability of DC/EOL solutions.  The method is simple in that the system produces a monitoring tone onto the line and a tone detector is fitted to the return leg of the circuit.  If the detector stops receiving the tone then it knows that there is an open or short circuit and will report an issue to the fault system.

There is a major advantage that is associated with this method.  It doesn’t apply to how it monitors but what it can do once a fault is detected.  Some systems can implement a ‘Class A’ loop.  This means that if an open-circuit is detected the system can switch the amplifier output so that it drives both the send AND return legs.  This allows the system to maintain coverage to the entire area even if there is a break in the circuit.

Impedance Monitoring

Impedance monitoring involves the constant measurement of the load on each speaker line.  In simplistic terms, a tone is produced by the amplification which passes through an internal resistance.  The voltage drop across this resistance indicates the load present on the circuit.  A high load (low impedance) will produce a high voltage drop while a low load (high impedance) will produce a low voltage drop.  The load reading (or voltage drop) is calibrated as a reference window and if the measured load goes outside this window then a fault is reported.  The good thing is that it doesn’t matter how many branches  there are attached to a particular circuit and it doesn’t care whether capacitors are fitted to each speaker on not.  It also saves on fitting EOL devices to each circuit.

However, there’s a reason why this method is at the bottom of the list.  Its not very accurate.  The problem is that the long lengths of cable and speakers themselves are subject to environmental effects.  So a change in temperature can affect the characteristics of these elements and therefore the load, taking it outside of its reference window and causing an incorrect fault.  The easy thing to do is to widen the window, but this is a bad option.  A 10% window means that you will need to lose 10% of your circuit before a fault is reported, which is unacceptable.  There is a workaround to this issue whereby a ‘dummy’ load can be fitted to the end of the circuit.  This dummy load essentially consists of a resistive-capacitive network that presents say a 20W load at the tone frequency (usually 20kHz).  With this fitted it allows the reference window to be increased  due to there being a significant load change when there is a break at any point in the circuit.  Unfortunately, this isn’t the end of the problem.  When the system is in use and a signal is present on the circuit, the perceived load can fluctuate wildly and take it outside even the widest of reference windows.  For this reason, many systems halt impedance measurement during announcements, which is far from ideal as it may take a possible fault report outside of the 100s time limit.  Impedance monitoring is not recommended for systems broadcasting background music, for the same reason.  Having said all this, it is possible to make this solution work effectively under the right circumstances.

So… which is best?

As always, there is no one-size-fits-all answer.  Every method has its advantages and disadvantages, and very much depends on a particular situation.  For example, if a system is being replaced and existing circuits are being retained then it probably would not be financially viable to fit capacitors to every loudspeaker to allow DC monitoring.  Similarly, you may have existing radial circuits, in which case installing EOLs will probably be the best option.  If you have circuits with an unknown number of tee-offs and no opportunity to survey then you may have no option but to use impedance monitoring.

For newly installed circuits, DC monitoring is attractive due to its simplicity, but your choices may be limited by the need to use a manufacturer that offers a particular system facility.

If you have a question then please feel free to comment or contact us.