Analog Trunks

Plain old telephone service (POTS) trunks are two- or four-wire analog service types that you can use to connect a voice gateway to a PBX. They are typically used where the call volume is low, because each trunk is capable of carrying only a single call at a time.

The analog trunk types that you can use for PBX integration include FXS, FXO, and E&M trunks.

Configuring FXO/FXS Connections

FXO and FXS ports are complementary interfaces. That is, they are meant to be connected to each other. When the PSTN supplies a POTS line to a location, it is supplying an FXS interface. To operate properly, the FXS interface must be connected to an FXO interface, such as that on analog station equipment (telephone, fax machine, modem, and so on) or to an FXO port on a voice gateway, as discussed in Chapter 6, "Connecting to the PSTN."

PBX analog trunk connections are typically FXO interfaces. They normally connect to a PSTN POTS trunk, which is an FXS interface. A PBX analog FXO trunk must be connected to an FXS interface on a Cisco voice gateway to work properly.

PBX systems usually also have FXS ports to connect to analog station equipment. This is called a station-side connection to the PBX. In some cases, you might want to connect a voice gateway to a PBX on the station side. Station-side PBX connections use FXO interfaces on Cisco voice gateways. You can configure these FXO ports exactly as described in Chapter 6. Station-side connections to the PBX can be useful when you are using the PBX as a pass-through device to get to the PSTN.

An example of where this might occur is during a migration from a PBX to an IP telephony solution. In the early stages of a migration, the PSTN trunks might still be connected to the PBX. The voice gateway can route calls between the PBX or PSTN and the IP voice network using station-side PBX connections if ports are available and the call volume is low enough.

FXS ports support two methods of supervisory signaling to detect on-hook or off-hook conditions and line seizure: loop-start or ground-start.

Loop-start is the most common signaling method used on PSTN trunks. It has two significant disadvantages: No mechanism exists to prevent glare from occurring, and no switch-side disconnect supervision exists to indicate the end of a call. These limitations make ground-start the preferred signaling method for tie trunks between a PBX and a voice gateway.

For ground-start signaling to function properly, the physical wiring must be correct. The connection is polarity sensitive, so you need to wire it with tip connected to tip and ring connected to ring.

After you have determined the proper connection port type and signaling, configuration is straightforward. Start by entering voice port configuration mode using the voice-port slot/subunit/port command. After you are there, you can configure signaling using the signal type subcommand, where type is either groundstart or loopstart. Loop-start signaling is the default. For installations outside of the United States, you can set up the locale using the cptone locale command.

To use caller ID, you must use the caller-id enable command on the voice port. When you are using FXO and FXS ports to set up connections to a PBX, it is important to remember that FXS ports send caller ID and FXO ports receive caller ID. Because of this relationship, caller ID is sent only one way when used on this type of connectionfrom FXS to FXO.

Example 7-1 shows the configuration of both an FXS and an FXO port. A station name and number is assigned to the FXS port using the station-id name | number word command. When station ID is coded, the interface sends that static information as the caller name and number.

You can verify the configured signaling and status of an FXO or FXS port using the show voice port summary command, as shown in Example 7-1. Output from the show voice port summary command shows that ground-start signaling is configured for these two ports.

You can display more specific information by using the show voice port slot/subunit/port command. Example 7-2 shows the caller ID-specific information displayed by this command. Note that the FXS port clearly shows that it will send the caller ID, whereas the FXO port will receive the caller ID.

MGCP

You can place both FXO and FXS ports under Media Gateway Control Protocol (MGCP) control. When defining FXO ports to CallManager for MGCP control, you must configure the signaling type. You can also control whether the port is used for inbound calls, outbound calls, or both. No signaling type is defined for FXS ports. However, when using an FXS port as a trunk, you must define the number of digits expected on an inbound call for proper call routing to occur.

Configuring E&M Trunks

The enhanced mechanisms for signaling answer and disconnect supervision available make E&M trunks a good choice for connections between a Cisco voice gateway and a PBX. Five discrete interface types are defined for E&M connections. Cisco voice gateways support Types I, II, III, and V. Cisco platforms do not support E&M Type IV.

E&M signaling defines a trunk side, which is usually the PBX, and a signaling side, which is usually the voice gateway. Cisco E&M interfaces are fixed as the signaling unit side of the interface. This might make it necessary to change the E&M trunk settings on the PBX to operate as the trunk circuit side depending on the interface type in use.

Note

For more information about E&M circuit wiring, operation, and troubleshooting, see Chapter 5, "Circuit Options."

Before beginning to configure an E&M trunk, you need to do the following:

• Determine whether the trunk will be using two-wire or four-wire operation Clearly understand the wiring arrangement in use by the PBX and verify that the wiring to the E&M port on the voice gateway is correct. Even though these connections are described as "two-wire" or "four-wire," it is common to have six or eight wires in actual use. Incorrect wiring is one of the most common problems encountered when installing an E&M trunk.
• Determine which interface type to use Cisco voice gateways support Types I, II, III, and V. Note that Type I is the most common in North America, whereas Type V is the most common throughout the rest of the world.
• Identify the correct start dial supervision signaling to use Immediate start, wink-start, and delay dial are the different options that can be selected. This setting defines the signaling format to expect prior to delivering addressing information, such as dialed number identification service (DNIS).

After you have obtained the necessary information, have verified the wiring requirements, and have the port physically connected, you can begin configuration.

Enter voice port configuration mode using the voice-port slot/subunit/port command. If necessary, configure the country-specific locale parameters using the cptone locale. The default locale is us.

Specify how many wires are used for voice transmission using the operation {2-wire | 4-wire} command. Note that this defines the wires used for the audio path only, not for signaling. The default is two-wire operation.

Specify the E&M interface type to which this port is connected using the type {1 | 2 | 3 | 5} command. On the VIC-2E/M or VIC2-2E/M interface module, you must set both ports to the same interface type.

Example 7-3 shows an E&M trunk configured for interface Type I, four-wire operation, and wink-start dial signaling.

Note

The current version of CallManager MGCP does not support analog E&M trunks.

Verifying and Troubleshooting

To verify that the port is configured correctly, use the show voice port slot/subunit/port command. This command shows you the current configuration of the parameters that were already discussed in addition to many others. Example 7-4 reflects the port that was configured in the preceding example. Notice that the port was configured for four-wire operation, interface Type I, and wink-start signaling.

If the port configuration is correct and you are still having problems placing or receiving calls on the trunk, the first place to look is at the wiring between the PBX and the voice gateway. As mentioned, the most common problem encountered with E&M trunk is incorrect wiring.

Note

You can find detailed information about E&M interface types and wiring arrangements on Cisco.com at www.cisco.com/en/US/products/hw/gatecont/ps2250/products_tech_note09186a008009452e.shtml.

• E&M Type I and Type V Uses two leads for supervisory signaling (on-hook/off-hook signaling). Cisco voice gateways expect to see off-hook conditions on the M-lead and signal off-hook to the remote device on the E-lead.
• E&M Type II and Type III Uses four leads (E, M, Signal Battery [SB], and Signal Ground [SG]) for supervisory signaling. The PBX connects the M-lead to the SB-lead to indicate the off-hook condition. The voice gateway connects the E-lead to the SG-lead to indicate the off-hook condition.
• Voice path operation The two-wire or four-wire operation for the audio path is independent of the signaling type. For example, a four-wire voice operation E&M circuit has six physical wires if it is configured for Type I or Type V. It has eight physical wires if it is configured for Type II or Type III.
• Voice path wiring In the four-wire mode, some PBX and key system products reverse the normal usage of the tip and ring (T&R) and Tip1 and Ring1 (T1&R1) pairs. In that case, to match up the PBX voice pairs with the voice gateway E&M voice pairs, you might need to connect T&R on the PBX side to T1&R1 on the voice gateway side, and T1&R1 on the PBX side to T&R on the voice gateway side.

The next step in troubleshooting is to look for problems with supervision signaling. You can do this by using the debug vpm signal command. Enable this debug mode and place a call from the PBX to the voice gateway. Verify that the gateway can see the on-hook/off-hook signaling. If you do not see the supervisory signaling in the debug output, a mismatch might exist between the interface type on the PBX and that of the Cisco voice gateway.

Finally, validate that the gateway sends and receives the correct digits to and from the PBX. You can do this by using the debug vpm signaling and debug voip vtsp session debugging commands. The debug voip vtsp session command displays the digits received or sent. If digits are passing but are not the expected digits, double-check the call control (dial peers, CallManager route patterns, and so on).