Ten mistakes technicians make when deploying a Multisite Capacity Plus system.

Although the below tips refer to Multisite Capacity Plus, many of them could apply to IP Site Connect too.

At work, I'm not involved that much with technical support (i.e. helping customers fix something that doesn't work). I do however get to hear about common problems with configuration and use this knowledge when helping my customers with setting a system up.

So here is a list of the top ten things I see, that people get caught out on:

1. No Router.

There's no two ways about it: Multisite Capacity Plus requires a Router at each site, period. Some people say that you can use a layer-3 switch but I've never seen that work.

2. Bad RF planning.

2 and 3 are kind of related but the typical problems I see with system designs are:

Choosing VHF for indoor and/or urban coverage. For indoor applications, UHF (or 800MHz if you can use it) works way better. VHF is a better choice in rural and/or suburban environments.

Using frequencies that are not free of intermodulation. Before installing repeaters on a site, make sure there is no/little risk of intermodulation by running a calculation using the frequencies already in use on site, as well as the frequencies of the repeaters to be installed.

Using frequencies that are being used by someone else. In this case, either the telecommunications regulator didn't do a very good job (rare) or the other user is using more RF power then they are allowed to.

I have a number of posts sharing anecdotal experience on the above.

Understanding the RF Path from Commscope is also a good source of general knowledge and a starting point for anyone starting out in this industry.

3. Problems with RF filtering – intermodulation and desense.

I've seen many systems; where there is one or more repeaters, and they either use multiple antennas (i.e. two antennas per repeater) or use a duplexer for each repeater.

A (compact notch) duplexer provides no RF filtering other than to allow the repeater receiver and transmitter to the same antenna. Other types of duplexers will provide some RF filtering but this too is not sufficient to prevent intermodulation or desense.

Intermodulation occurs when two or more signals (A and B) mix to produce a third signal (C). In most cases this third signal does not produce any symptoms but in some cases, it can land on the receive frequency of an existing system.

As a practical example, lets say you had a candidate site in which you wanted to install a two repeaters. Currently, there are three repeaters in place and using the following frequencies:


TX1 438.75000 MHz RX1 431.15000 MHz
TX2 172.98750 MHz RX2 164.58750 MHz
TX3 158.40000 MHz RX3 163.80000 MHz
TX6 213.8000 MHz

Your two repeaters will use:

TX4 407,91250 MHz RX4 417,91250 MHz
TX5 406,32500 MHz RX5 416,32500 MHz

Running the above through my intermodulation calculator gives the following results*:

Total number of IMs for Rx 1: 6
IM01 @ 431,03750 MHz 7th order -f1+f2-f3+4*f6
IM02 @ 431,13750 MHz 7th order -f1+3*f2+f3+f5-f6
IM03 @ 431,02500 MHz 8th order 3*f2+2*f3+f4-2*f5
IM04 @ 431,10000 MHz 8th order f1+f2+4*f3-f4-f5
IM05 @ 431,11250 MHz 9th order -f2+f3-2*f4+f5+4*f6
IM06 @ 431,21250 MHz 9th order f2+3*f3-2*f4+2*f5-f6

Total number of IMs for Rx 2: 7
IM01 @ 164,60000 MHz 5th order f1+2*f2-f5-f6
IM02 @ 164,46250 MHz 6th order -f2-3*f3+2*f5
IM03 @ 164,67500 MHz 7th order 2*f1+2*f3-2*f4-f6
IM04 @ 164,61250 MHz 8th order -3*f3-f4+f5+3*f6
IM05 @ 164,68750 MHz 8th order -5*f2+2*f4+f6
IM06 @ 164,71250 MHz 8th order 2*f2-f3-f4+2*f5-2*f6
IM07 @ 164,50000 MHz 9th order f1-2*f3-2*f5+4*f6

Total number of IMs for Rx 3: 3
IM01 @ 163,67500 MHz 8th order -f1+4*f2+2*f3-f5
IM02 @ 163,75000 MHz 8th order 2*f2+4*f3-2*f4
IM03 @ 163,82500 MHz 9th order 4*f1-f3-3*f5-f6

Total number of IMs for Rx 4: 5
IM01 @ 417,88750 MHz 5th order -f1-f2+2*f4+f6
IM02 @ 417,96250 MHz 7th order -3*f2+2*f3+f5+f6
IM03 @ 418,03750 MHz 7th order -f1+f4-f5+4*f6
IM04 @ 417,80000 MHz 8th order 6*f2-f5-f6
IM05 @ 418,00000 MHz 8th order -2*f1-f2-f3+f4+3*f5

Total number of IMs for Rx 5: 7
IM01 @ 416,30000 MHz 5th order -f1-f2+f4+f5+f6
IM02 @ 416,45000 MHz 5th order -f1+4*f6
IM03 @ 416,21250 MHz 8th order 6*f2-f4-f6
IM04 @ 416,41250 MHz 8th order -2*f1-f2-f3+4*f5
IM05 @ 416,43750 MHz 8th order 2*f2+3*f3+f4-2*f5
IM06 @ 416,22500 MHz 9th order -2*f1+f2-2*f3+3*f4+f6
IM07 @ 416,37500 MHz 9th order -3*f2+2*f3-f4+2*f5+f6

12,5kHz observation window and run to the 9th order.

In most of Europe 213,8MHz would be used by DAB+. Most of these intermodulation products would be fairly low but TX6 (f6) could be as much as 100kW so even if suppressed by 90dB, you could still get 2µV at the receiver input of one of your repeaters - which, in the right (wrong) circumstances, could override a good proportion of user's signals.

Intermodulation can be minimized by ensuring that there are galvanic connections between the antenna and repeater. Also, in some cases, it is necessary to install a circulator between the repeater transmit output and the rest of the system.

Also the coaxial flyleads between the various elements in the RF subsystem (e.g. cable between TX and duplexer) need to be a specific length such that the voltage node occurs on the cable ends.

A galvanic connection occurs when two dissimilar metals come into contact or when one part of an electrical connection becomes rust or contaminated. This connection acts like a kind of a diode and when subjected to RF energy can produce harmonics (i.e. TX freq. * 2 and * 3).

Generally speaking, for sites where there are two or more repeaters from the same system or in the same part of the frequency band, you should use a combiner. Yes, this is more costly and the insertion loss seems high, but the benefits far outweigh these two downsides.

In some cases you can use two duplexers and two antennas but I've only ever seen one system where this way used - and it was only because of the frequencies allocated to this customer.

If there is a fault with the duplexer - or any other RF filtering component. Or, the antennas from different systems are not placed far enough apart from each other, desense could occur. Desense can also occur if you use too much power.

Basically, desense occurs when a large enough off-frequency signal reaches the RF front-end amplifier of a receiver. The interfering signal does not cause the receiver to pick up anything and instead will simply make it seem deaf (i.e. receives with reduced sensitivity) while the interfering signal is present.

Related to desense is co-channel interference. This is where a strong signal is present a few kilohertz off the receive frequency. This may cause the receiver to pick up noise or could interfere with the receivers ability to produce a clean signal. To check whether your system is suffering from desense, have a look at this post.

I've seen this more than once: If you have a trunked system and use separate antennas for each repeater, the radios will behave strangely. This is becuase the repeater with the highest antenna will give better coverage than the other(s). Here is a post on this very topic.

There are a good many readers of my blog who work for one or other of the well-known RF filtering-hardware manufacturers. I'm sure they'll be able to add to the above so have a look at the comments (quite often they're more useful than the post itself).

4. Router setup (NAT).

From when Motorola Solutions launched Multisite capacity Plus in 2011ish, I was taught that the Routers used at each site shall support and be configured for NAT.
NAT loopback was initially required but this requirement was dropped in R2.2.

The point is: I have always used NAT and it has always worked in every situation.

The Router you choose should support these features which potentially means you could use just any model. My recommendation here is to NOT use any Router primarily intended for domestic use. Like I always say "You've spent €000's on MOTOTRBO repeaters and peripherals so why let your system down with a CPC Router you got for €50 on Wish.

If you're tight on budget, I've had good results so far with Mikrotik RB2011 and they go for around €116 on Amazon. Okay, they're not not everyone's cup of tea but hey, they do the job!
In addition to NAT, the Router attached to the Master repeater must have a static IP address. If you will use DNS, you can possibly use dynamic IP addresses.
I've never given much attention to DNS (partially because I don't have a DNS server I can fiddle with or maybe I'm too lazy-arsed to setup a DNS server to try, or both).

I prefer to use static IP addresses for all the repeaters - that way you don't loose communications while the DHCP pool gets refreshed.

5. Not being careful about UDP ports.

To save yourself some head scratching, use my cheat sheet and set your repeaters according to it. I have a video coming up on this topic so keep your eyes peeled.

As far as port numbers go, the Master repeater needs its own UDP port. The rest channel host needs to have the same IP address on each site and should use the same UDP port all-round. Each peer should have its own UDP port.

As far as choosing port numbers go, I use this simple rule: Radio ID+49999. So in this case, the master gets 50000; the 1st peer gets 50001 and so forth. As far as a Rest Channel Site Port goes, I always use the default value of 55000.

As far as IP addresses go, the repeater with the lowest Radio ID on each site gets IP address 0.0.0.1; the next repeater gets 0.0.0.2 and so forth. The rest channel site IP can either be 0.0.0.100 or 0.0.0.200. (0.0.0.x refers to your subnet e.g. 192.168.1.1)

If you're using IP Remote Programming, remember that the Device programmer uses TCP ports 50000 to 50100.
Also remember to unblock all of the above ports in Windows Defender Firewall, on the RM Device programmer host PC. However, don't disable Windows Defender Firewall (permanently) or you'll expose that PC to all the badness out there.

6. RSSI Threshold in the Radio and/or Repeater.

There is a RSSI Threshold setting in both the radio and repeater. In some of my hands-on presentations I've shown this set to -40dBm in the repeaters without explaining why I did so. Mea Culpa!

The RSSI Threshold in the Repeater should be set to -40dBm only while you're setting up and testing the system. Once deployed, you should set this to no less than the mean on-site noise floor plus three standard deviations. Or, to save you the calculation 10dB above the noise floor.

This value is used by the repeater to determine whether the channel is busy or not. It also helps the repeater to determine whether it should ignore a weak signal. So using this setting you can reduce the inbound range of your system if needed (this is useful when frequencies are shared).

Setting this value to -40dBm on a live system means that radios that have a signal below 2,2mV will be ignored! This value is okay if you're standing 5m away from the repeater but wont work in real life.

So how do you determine what the noise floor is? Using RDAC or the repeater's web page, read the RSSI while there is no on-frequency traffic. Do this at several times during the day and note the value each time to get some reliable data.

On the radios, if you choose to set TX Admit Criteria to Channel Free, the default value is -124dBm. Many people miss this setting and the result is that the radio sometimes fails to transmit. Here is an old post on this topic.

7. Capacity planning.

Multisite Capacity Plus uses all-start for talkgroup calls. This means that there must be a free slot; on all sites to which the talkgroup is streamed, in order for the call to go ahead.

As an example, you have a three Multisite Capacity Plus system with two repeaters at site 1 and 2 and one repeater at site 3. All talkgroups are streamed to all three sites. If the system is currently passing two talkgroup calls, if any other radio - not party to those two calls - makes a third talkgroup call on the system, it will fail, even though one repeater at site 1 is idle.
Why? Site 3 only has two timeslots (one repeater) and both of these are free. The requested talkgroup is streamed on all sites and so in order for this call to go ahead, one timeslot on site 3 must become free.

Talkgroup calls are streamed statically to all sites, no matter if there is a call party at a site or not. There is no trunking controller so there is no real means to manage mobility information.
In the above example, it would have made sense to either restrict some talkgroups to only site 1 and 2 or to increase the number of repeaters on site 3 to three.

For the Erlang fans, Multisite Capacity Plus uses Erlang B. (Capacity Max however uses Erlang C due to the trunking controller and call queuing/pre-emption capabilities).

8. Reserved Wide Area slots too high.

Multisite Capacity Plus has a Reserved Wide Area Slots setting that will reserve a certain number of slots on each site purely for talkgroup calls that are streamed on more than one site. Talkgroup calls tat only go out on one site will not be allowed to use these reserved slots and if all other slots are busy, the call will fail - even when those reserved slots are free.

9. ARS on in the radios but no application.

Turning on ARS in the radio without having an application to acknowledge and process them will just create a data storm. With ARS, when the radio turns on, it sends a short burst of data to the defined ARS Radio ID. The radio (or MNIS) receiving this will send an acknowledgement.
If no acknowledgement is received, the radio will try again in a short while and while there is a free slot. This will continue until the acknowledgement is received.

Apart from creating a data storm, the batteries on all portable radios will go flat sooner as the radio is transmitting more often (without the user knowing).

10. IP backhaul bandwidth/latency problems.

Motorola Solutions produces a very useful System Design tool that can calculate the needed bandwidth at each site - among many other things.

In terms of latency, your design goal should be less than 90ms.

Jitter should ideally be zero. If the variation in jitter exceeds 90ms, you could experience call setup failures or garbled audio.

Packet loss should ideally be zero. A higher packet loss could result in audio holes or distortion.