Repeater Logging
Over the summer of 2024, I migrated a high profile repeater back to it's original site. After an initial burst of enthusasism, and several months, a number of complaints regarding it's behavior and coverage were lodged with the management.
An effort was launched to understand what was happening and why. You can learn a lot about a repeater system by listening to the downstream audio and signal. However to do that diagnostic, you do have to understand how a repeater is operating, and where the audio / modulated signal is coming from at any given moment.
Two or more stations received on the input at the same time with similar signal strenghths has a very distinctive buzzing sound, which the repeater will attempt to reproduce on the output. The buzzing that goes away and one station continues reasonably clearly is an indication that it was a simple 'double'. This happens all the time and it's my opinion that being able to recognize this sound is a skill that FM amateur radio operators should aquire in the first year of FM radio use.
The RSSI indicator, the s-meter being a very coarse one, on a user radio tells you how strong the repeater is to you, but tells you nothing of how strong the transmitting station is to the repeater. When a station sounds scratchy, noisy, or otherwise difficult to read, is that the link from them to the repeater, or the repeater to you?
Well, the first line of assessment is listening to the signal when the originating station drops. At that moment, all of the audio you are hearing is being produced by the repeater controller, and the downstream link conditions to you. For that 1 second of silence, followed by a courtsey tone, you are hearing what the repeater's signal to you is, not influenced by the quality of the input signal. That's the window to self evaluate your radio's operation.
In my case, I'm the repeater operator. I have more options. I can go listen to the repeater's receiver. I can hear what the signal sounds like to the repeater, but can I see the RSSI?
The repeater in question is built around the Icom FR5000 series devices, attached to a CAT 300 repeater controller. The CAT 300 is connected to the Icom Repeater using the Icom's Accessory Connector in what is the obvious manner. Pin 10, however, is 'assignable', and one of the options (on page 26) is that it can be an RSSI signal. The operator that came before me helpfully configured pin 10 for that purpose, and hung it out on a RCA connector to be read.
A volt meter, and turning the volume up on the local speaker, and away we go.
I can now read suprisingly fine grained RSSI information, along with listening
to what the repeater is demodulating on it's input. Note, the speaker on the
repeater is presenting received audio, it does not play what the repeater is
transmitting, which, would include things like ID, announcements, and courtsey
tones.
Wonderful, I can now stand in a repeater dog house, at the base of a tower, and listen to the repeater audio and see the RSSI on my volt meter. A 4 digit voltmeter shows a total working range of between about 1v and 2v, with about 3 digits of resolution in the RSSI data. Individual stations from fixed locations generally see less than 0.002v fluxiation in the signal strength over the course of several minutes of transmission. An interesting note is that the RSSI voltage reading never actually gets to zero. This is an indication that the noise floor is detectable on the RSSI reading, so a change in this base reading over time would indicate the noise floor moving.
I need to be able to record this somehow, because seeing it is good, but being able to review it is much, much better. The first attempt, point a camera at the volt meter, pipe the audio from the repeater's audio out to the microphone input of the camera and record the result.
Having another helpful party do the same thing, but to a radio several miles away listening to the output of the repeater results in three and a half pieces of data that can be assembled into a useful diagnostic log:
- repeater input audio according to the repeater
- repeater RSSI
- repeater output audio
- repeater output s-meter
This worked, but had the failure that I actually had to be out at and visiting the repeater site during the event, because I don't have cameras that are smart enough to automatically record only the interesting information, nor do any of them have the ability to run unattended for weeks at a time. We need full fidelity audio, and detailed RSSI readings, but only when something interesting is happening. Additionally, the second recording from off site suffers the same problems, with mostly attended operation. Finally, to make this diagnostically useful, I have to merge two cameras worth of data into a single file, representing the audio in and out as left and right channels, with a picture of the volt meter and the remote downstream radio's s-meter, another manual process.
Most of the value I was getting out of this system diagnostically was in the RSSI and upstream audio, so for my next phase, I focused on making that system provide long term unattended recording.
Version 2 specification:
- Record audio
- Record RSSI
- Run for weeks unattended
- Record when something 'interesting' was happening.
Consulting the availble set of tools, the obvious choice was to get a dedicated audio recording platform, of which many are available. The one selected out of the pile of parts was a Zoom 4 track handy recorder. Feeding audio into one of it's inputs got goal 1 out of the way. Setting the recorder to be 'sound level activated' isn't quite the 'interesting' mark, but it serves as a partial solution. The Zoom audio recorder is quite happy to sit for weeks, unattended, recording data. So that's need 3.
How to get RSSI recorded? I have a 4 track audio recorder, but audio recorders don't record DC voltage. Can I convert the DC voltage readings to audio in a way the audio recorder can save them?
Again, consulting the pile of parts lying around, an Arduino Uno R4 Minima was located and pressed into service. A sketch was rapidly developed that reads the voltage on an analog pin, does some math on it to map it to a frequency, and then generates that tone on the single DAC pin the R4 has. That is then piped into the recorder as right channel audio.
The result of this is a stereo file that contains repeater input audio on the left channel, and an RSSI tone on the right channel. The RSSI tone, for completely arbitrary reasons, is updated 30 times a second, which means that if you listen to it, it might sound like a data channel, and I suppose in some way, it is. A 30 baud, 400 ish symbol AFSK data channel.
At a human factors level, this is a pretty easy format to share with people to help them understand what's happening. A low tone tells us that you aren't actually getting into the repeater that well, and is coupled with the audio heard. We have found examples where people were making into the repeater just fine, but wildly overdriving or underdriving the microphone, resulting in problems.
The system is in place, recording traffic and providing us diagnostic logs. It is frustrating that so far, this data has led to locating problems not in the repeater, but in various user's radio setups. But without this information, that would have been very difficult to detect without visiting every user's radio and observing how they operate it.
Plans exist to expand this platform with more data logging, but that's for an updated post in the future.