A digipeater, also abbreviated as "digi", is very similar to the FM analog repeaters most hams have used at one time or another. It hears a station's transmission and repeats it -- hopefully over an area wider than the originating station's simplex coverage.
Rather than using voice, both the originating station and the digipeater transmit a squawk that is really just two tones being modulated. Generally, this squawk (or packet) lasts for less than one second, but it conveys a lot of information. It contains the callsign (and SSID) of the station, its location, the software and/or device being used, the icon the station has chosen to represent it on maps, and the accuracy of the transmitted location. Very often, a brief message and/or the frequency the operator is monitoring, and the time the packet was sent (called a timestamp) is included.
HF operators are probably at least somewhat familiar with more sophisticated digital modes such as RTTY, PSK31, and FT8 (to name a few). These modes use a larger selection of tones for greater efficiency and a simple kind of error correction. The biggest difference, however, is that these modes operate on sideband, rather than FM.
APRS, for all the talk of digital modes, is really just standard analog FM using two tones instead of voice. There is no modulation in quadrature like C4FM. This also makes a digipeater very similar to the FM analog repeaters most are familiar with. What this means for the operator is that any standard FM mobile or hand-held rig can be used to access a digipeater. APRSDroid, the Android APRS app, has a setting where you can have your smart-phone speaker be the output. You can, in fact, just key the mike on your HT and let the app generate the squawk. This will transmit your packet in the same way as if you used a software modem (e.g. DireWolf or Mobilinkd TNC) to "translate" your text into a squawk, inserted directly into the mic jack of your radio. While the method of using your microphone and smart-phone speaker may be too cumbersome for everyday use, it does prove that APRS is just standard analog FM -- no more mysterious than using a PL tone or DTMF tones on your radio.
The difference
A digipeater differs from a standard FM analog repeater because it operates on simplex. Virtually all APRS activity (other than satellite digipeaters) operates on the simplex frequency 144.390. There is no offset, like on a standard FM analog repeater, and there is no PL tone used to access the digipeater. It stores the packets it hears, and re-transmits them when the frequency is clear (hopefully). A standard FM analog repeater transmits what it hears simultaneously. This is why a standard repeater needs a frequency pair, and (usually) a duplexer.
A digipeater will not re-transmit voice transmissions. If you use voice on 144.39, you will simply be producing interference. The only possible exception is using VoiceAlert, which should be kept very brief. Even this will not be re-transmitted by a digipeater.
Virtually all digipeaters use a TNC or some software equivalent to filter the audio of the received packets and provide the optimum audio levels for packet communications. In theory, at least, your originating packet should emerge from a digipeater even clearer than when it went in. If you decide to operate a digipeater, you should pay close attention to the filter parameters and audio levels of both tones. Otherwise, you will merely be producing a jumble of characters from the packets you receive. There are also cases when the operator of the originating station has not observed these practices, and it will be a case of "garbage in, garbage out".
This is probably the subject of a separate post, but I'll briefly touch on it now. Different radios have different audio characteristics. You've probably noticed some stations are "muddier" with more "lows", and other stations' audio has more treble. There is no "one size fits all" audio setting that will accommodate all rigs. In addition, the "mic gain" varies from radio to radio. Any APRS operator should take a minute or two to ensure their TX audio levels are correct, and this should be done separately for both the high and low tones to account for different audio passbands on different rigs. This is usually accomplished by listening to your own signal on a second receiver and raising the TX audio level until it no longer seems louder on the receiver. This is something that usually only needs to be done once.
Smart vs. Dumb digipeaters
I should note that the term "dumb" here is not derogatory, any more than the term "dumb terminal" is when discussing networks.
A "dumb" digipeater is one that repeats received packets without considering if that packet has been already gated. Without a connection to the internet, a digipeater can't check if your transmitted packet is a duplicate ("dupe") or the original. This can increase congestion dramatically in areas of high traffic, such as the metro. For example, your mobile station may use "smart beaconing" and has just transmitted a packet indicating that you made a right turn. That packet has already gone to a gateway and made available to operators world-wide. But a "dumb" digi has re-transmitted it, and another "dumb" digipeater has re-transmitted that repeat, and -- if the N-path allows -- a third "dumb" digi will echo that right turn.
This gets compounded when you consider that "dumb" digipeaters reach beyond the metro. They are also bringing in packets from other areas outside of the metro, notifying everyone in the metro that someone in Rochester just made a left turn.
Some "old school" hams see this as a feature, rather than bug. They see gateways as irrelevant. It doesn't matter if the packet has been gated, because the goal is to connect all APRS operators over RF. It used to be that northern MN, as well as southern MN, subscribed to this model. Both were "digipeater-heavy", with few gateways. Wisconsin still largely adheres to this model.
While I certainly see value in connecting relatively distant areas over RF, I wonder how most APRS operators are using APRS. Are they sitting at a terminal, watching RF traffic in real-time? Or are they using internet-based sites such as aprs.fi to show traffic that may have happened hours ago, and which includes -IS (Internet Service) traffic?
There is also a distinction to be made between rural areas and congested metro areas. A single "dumb" digipeater, with no other link to a gateway or other digipeaters, may provide local operators with an RF link to each other beyond their simplex range, much like a standard FM repeater does. And, if that "dumb" digipeater is within range of another "dumb" digipeater, then the range of the individual RF operator is expanded considerably. In theory, they could use messaging to communicate back and forth just like on a voice FM repeater. But is this how most operators use APRS?
In contrast, "smart" digipeaters usually also function as gateways. Local examples would be W0PZT-1 in Medina or W0YC-5 at the U of M. Packets that they can't immediately gate (due to high traffic levels) are digipeated to the wider world, hopefully to be gated elsewhere. But they won't digipeat traffic that's already been gated -- as far as they can ascertain.
The problems surface when gateways such as N9MEC in Apple Valley delay gating packets by a minute or more. These "smart" digipeaters re-transmit duplicate packets sent by "dumb" digipeaters, believing that they were never gated. In other cases, packets re-transmitted by "dumb" digipeaters can fool the "smart" digipeaters into believing that the packet they received is an entirely different packet than one that has already been gated, because the path is different.
About the only guarantee a "smart" digipeater offers is that it generally won't re-transmit packets that it has already gated.
Yet another consideration is propagation. When tropo rolls in and we directly receive packets from Wisconsin, Duluth, Willmar, and Iowa, the last thing we need is digipeaters magnifying that traffic.
The proper role of digipeaters
For RF-only APRS operators, digipeaters seem critical. Yet, it should be obvious that you can get too much of a good thing. There is no governing or co-ordinating body to regulate APRS, beyond the basic FCC regulations. Someone could set up three 50W "dumb" digipeaters in the metro next to each other, and they would bounce all traffic back and forth until the N-path expires. This wouldn't expand the coverage at all, but it would generate a lot of 'noise'. Next, consider that some digipeaters reaching the metro ID once every minute. These IDs would also be re-transmitted back and forth across the metro, compounding the impact of an inconsiderate operator. It's doubtful that these scenarios are what RF-only APRS operators have in mind when promoting digipeaters.
To my mind, the proper role of a "dumb" digipeater would be to bring in traffic from outside the metro, presumably to be gated. There is very little need for metro traffic to be digipeated within the metro area. What need there is for that can be handled by "smart" digipeaters, and really provides more of a case for additional RX-only gateways than it does additional "dumb" digipeaters.
Once upon a time, this is the model that the metro used. Digipeaters on the fringes of the metro would repeat traffic from outside the metro, to be gated by our efficient network. The goal was always to get packets gated, not repeat them endlessly. This worked well for RF-only operators and others who used a hybrid approach.
There may very well be a role for a few low-power (5W) digipeaters in low-lying or otherwise obstructed areas in the metro that our network of gateways can't reliably reach. Examples would be Downtown, Uptown, the airport, or even low spots in Eagan. Realistically, though, TX or RX-only gateways would be a better solution for these areas.
Rural areas are probably the best role for digipeaters, provided that they can reliably reach a gateway. This allows APRS traffic to be used to determine real-time propagation, and connect APRS operators to the wider world by way of the -IS. ideally, the two work hand-in-hand.
And if the internet should fail in a particular area, the digipeater would allow local RF traffic to reach unaffected areas. Even "smart" digipeaters would continue to function as "dumb" digipeaters if the internet fails, so the typical EmComm argument tends to fall flat. And a TX gateway will continue to transmit RF messaging to/from stations in its local area without the internet.
"Dumb" digipeaters will remain the favored option at sites without internet access, but there's little reason not to add an RX-only gateway nearby. Look at the example of the RX-only gateway AE0RF-10 next to the "dumb" KF0ZH digipeater. Or the RX-only KC0CAP gateway near the powerful "dumb" KD0JOU-2 digipeater. Any RF-based messages that these digipeaters re-transmit will be immediately gated and passed on to a TX gateway. This makes RF-only operation more powerful, not less.
The Big Picture
Digipeaters aren't inherently "bad" or "good". In the anarchic model that APRS exists in, a properly-functioning network relies on individual operators to make sensible decisions on the need for a digipeater in their particular local situation. What value would a digipeater add that a gateway couldn't provide? Is there another digipeater within range of your proposed digipeater that could leverage the coverage over a wider area? Would your proposed digipeater reach a gateway, allowing information to be shared? Is there potentially a base of RF-only operators locally that would benefit more from a digipeater than a gateway? Are you willing to devote the time it requires to set up audio filtering and levels on a digipeater to make it beneficial to the community? Is there already a gateway providing service to the local area?
Generally speaking, digipeaters are about as difficult to set up as a gateway would be. Virtually all TNCs that can be set up as a digipeater can be configured to be an I-Gate or "smart" digi with the addition of an internet connection. Often, using software like UI-View or APRSIS32 is much more difficult to configure as a digipeater than as a gateway.
Just because you can set up a digi, that doesn't mean that you should.
In most of outstate MN, the more critical need is for gating capacity. You will probably get more impact solving that problem than you would get by adding a "dumb" digipeater. Consider more modest solutions, like a gateway with a 1/4 wave exterior antenna on top of the roof of a rambler, than a big digi on a tower.
Another possibility might be a satellite gateway set up on 145.825 to make satcom more valuable.
Consider your options. A "dumb" digipeater may not be the best solution for your local area.
