Welcome to W2BRI's Digital Voice Site

AOR's ARD9800 Digital Voice Modem
A Review

Review, Audio Samples, Sidebar, and Note on Quality


In these digital times, where amateurs can bounce weak signals thousands of miles and establish reliable communication with ease, the efforts to improve our digital communication arsenal are gaining great momentum. With each new digital mode applied to moon bounce, meteor scatter and DX, the old assumptions and records fall away bringing new expectations and desires. At the forefront of this new, inventive effort is the drive to deliver a digital voice product for amateur radio that will bring with it advantages over current modes and techniques. 

It isn't enough to take current phone communications like two meter FM and turn them digital, rather, it is the passion of modern day inventors to give current amateurs a mode that's superior to what we have, better audio quality, higher signal to noise, and simultaneous data with voice. With this in mind, the people from AOR have released the ARD9800 Fast Radio Modem, a digital voice modem for the amateur operator and ham market. 
The ARD9800 is a small modem device 3x1x6 inches able to connect to almost any amateur operator's rig. This device is unique - it can deliver near FM-like sound quality on HF and still maintain strict FCC bandwidth limitations. 

The real trick to decent quality digital audio is the compression engine used to fit the necessary data stream within regulated bandwidths. You see, in the past, decent audio quality required so much data, the 3.0 khz HF phone requirement was just too narrow. However with the advent of faster chips, and better compression schemas, the dream of better HF audio communication is finally here. AOR is the first commercial player out of the gate with a product that starts to bring us obvious benefits over current HF SSB communications. 

Just as SSB came into its own back in the 1950's and replaced the AM mode for the most part, so too digital voice will shape the future of the HF bands. As my wife wandered by my shack one evening and heard my friend Juan coming out of my HF rig, I remarked that we were using forty meters. Her face changed into surprise, and she said, "That's HF?" Yes, I replied, and explained that we were using digital voice over radio. "I thought you were on FM or something like that," she said. To say the least, she's not a fan of HF noise and was completely impressed by the audio quality and lack of hiss, hash and crackle. Imagine working your schedule on HF, but no longer contending with the natural noise that comes with HF usage. Digital voice technology brings the dream of high quality audio framed against a noiseless background closer to practical reality. 

Connecting the ARD9800 to your radio is a fairly easy process if you have the necessary cables. A special microphone cable connects your radio to the modem via your rig's microphone port, and your radio's headphone jack gets plugged into the modem's audio in connector. I highly recommend buying these cables from AOR for your specific radio (the modem works on all modern HF rigs according to AOR), because figuring out the pin assignments on the microphone cable can get frustrating and the manual is not as clear as I would like it to be. My understanding is they are constantly upgrading and working on the manual. As of the time of this writing I was supplied with a second manual with more improvements included. I myself got a hold of the AOR made cables, and it made things much more pleasant. 

Once the connection between the modem and the radio are complete, all one needs to do, is get on frequency with another modem owner and try your hand at digital communication. The modems can be used on any HF band except 60 meters, and will work in any mode: AM, FM, and SSB. When a user presses the PTT on the their modem's microphone, a short ring is heard. This ring is a digital header being sent by your modem to alert other modem users on frequency that digital communication is about to be sent. Once the short header ring sounds, those listening with their digital modems will hear their speaker go quiet, the familiar HF din will cease, and clear audio will start to be heard. It's almost magical to see these modems work the first time and see the amazing transformation that takes place when digital communication takes over. 

However, this new product is not without its faults, and readers should not believe at this early stage of development digital voice is a perfected mode of operation without its own shortcomings. It too has its own quirks and operational issues that must be explained. For instance, the modems sometime do not recognize the initial data header and will not decode the digital voice stream. As a result the receiving station can end up listening to data noise instead of digital voice. Also, it is sometimes hard to get the audio levels just right to make communication possible. There are instances where the modems work only one way, from station A to station B, but not in the reverse. Station A never receives the digital audio, only digital noise. In the many experiments made with this product, it was never fully clear why the modems did not always work perfectly, although, it was probably do to operator error on one side of the transmission. [Editors Update from Paul, KQ6EH - Paul Metzger: Many of us have had this issue, we have found it to be directly related to ground loops, no true low impedance earth ground attached to both the ARD9800 and transceiver, local RFI, or power supply issues. With a little experimenting, all problems are cured. Although these issues can be quite tricky if they are compiled at the same time. When I had originally attempted a QSO with you during the time you had the two test units loaned out to you by AOR, I could only hear you. While you were very patient during that time, I had adjusted all of the levels on my ARD9800s with no avail. It wasn't until later that I had found the cure. A true low impedance earth ground. Since I had posted this fix to the www.rfelectronics.com web site, others have used it with equal success.]

Also, AOR has instructed owners to make sure certain conditions are met when using the modem. For instance, a switching power supply cannot be used with the product or on the radio utilized with the modem. It seems that these power supplies introduce unwanted distortions in the transmissions that impede successful contact. I cannot explain this completely, however I have had to switch to a different power source to make the modem work right. So a linear power supply or battery must be used instead. In addition, all receive and transmit DSP must be turned off in order for the modems to work properly. Users also must make sure all narrow filters are turned off so the full data stream can get through the IF or audio bandwidth of the radio on transmit and receive. 

Despite all these issues, potential buyers should realize that this product represents the cutting edge of amateur radio technology and that all their work and experimentation is truly groundbreaking and unique. Buyers will be working with an entirely new mode of communication for HF, and will discover the characteristics and limits of the technology. It is a truly fantastic opportunity to be at the forefront of a new amateur mode, helping shape and drive the way this new technology is applied and used. 

These modems are also capable of other functionality, like slow-scan TV with an optional memory module, and as a data modem, but for the purpose of this article, the topic of digital voice communication was the most compelling feature and therefore the complete focus. Those amateurs stuck in the old ways beware, for a new mode is upon our service that will alter the radioscape of our hobby. As more amateurs experiment with digital voice, the nature of our bands will change bringing the new sounds of data intermixed with the cadence of SSB communications. As digital techniques improve, and new modes are introduced, ham radio will further evolve into an exciting and rich future. For those that think everything has been done before them, pick up this device and see that we are simply scratching the surface of what is to come. 

By Brian Levy, W2BRI 
 
 

Audio Samples

1. K0PFX in Bridgeton, MO, talks to W2BRI on 20 meters, digital voice (660K)
    File: mcontact1.mp3

2. W6DMV makes a short contact with W2BRI on 40 meters, digital voice (206K)
    File: w6dmvshort.mp3

3. W6DMV changes his power setting while transmitting and shows how the modem can recover if the power drops too low. (1.45MB)
    File: w6dmvpower.mp3

4. An example of the data header and what the modem sounds like if the analog data is heard (68.9K)
    File: ring.mp3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 

Email sidebar from Charles Brain,
G4GUO, codec inventor on sync issue

From my work, I found that the reason it does not always sync correctly is due to the code that estimates the frequency offset error. It looks as if they (AOR) have just copied the DSP code I gave them. My prototype now has a switch so once I have managed to correct the TX offset I can disable the AFC and that works much better. The units have to estimate the error to about 1 Hz so it has a lot of work to do. The problem with switching supplies is due to the high peak to mean ratio of the waveform, the electronics in the PSU can't keep up with it. Personally I like to run my prototype units off a separate supply to the main transceivers supply as well.  As you have found it is very important to make sure the waveform is transmitted without distortion. You really have to have you equipment set-up correctly to use this mode.

 - Charles

On the Subject of Audio Quality
with the AOR Modems

Throughout the testing of these modems it must be said that the digital audio quality will vary depending on the strength and nature of the contact. There were contacts where the audio far exceeded the quality of the samples listed on this page. These samples were simply the result of recordings that were taken at certain opportunities. I was amazed how well these modems could reproduce the analog audio with certain contacts, and the modems would scale the audio down when needed because of poor conditions. I would rank these recordings on the MOS system at 3.5-4.0, there were contact where the MOS was closer to 4.0-4.5. 

Also, some have questioned the advantage of digital voice over SSB, since from these recordings they sounds similar in performance. The distinct advantage is the freedom from background noise, and the ability to easily hear the voice communication. Notice that on these recordings you hear nothing but the person you are in QSO with, all other noise factors are gone. Also, from a signal to noise level, throughout the tests we were able to establish QSOs at very low power levels, levels that when listened to at analog mode, were not easily copyable. As long as the digital stream is in place, even at power levels of 1-5 watts, the digital reproduction will be nominal if band conditions and the technology allow for it. 

- Brian

 

 Back to the Digital Voice Home Page

Check out my other websites: W2BRI's Magnetic Loop Website |
allaboutprojects.com: My Project Management Website for management

Copyright 2002-2003 Brian Levy