More Adventures with Icom TX Audio

There are a bunch of posts on this blog which detail my experiments with transmit audio on the IC-7300. Since then I have switched to an Icom IC-7610, which seems identical in the TX audio realm.

Over the past few years, I have experimented with several different audio processing equipment. I have a small recording studio setup, being a lifelong music lover and musician, so why not try some of that outboard equipment for speech processing?

Modern Icom radios have a narrow transmit bandwidth. 100 Hz to 2900 Hz is the widest setting on radios like the IC-7300 and IC-7610. That is plenty of room for basic voice communications, but there is an interest in seeing just how good you can make an Icom sound on the air.

My Ten-Tec Jupiter, in comparison, has a maximum transmit bandwidth of 3900 Hz. It also has a hi-boost function which boosts the modulation high frequencies, a TX low frequency roll-off which is selectable from 0 to 1270 Hz, and an 8000 Hz receiver bandwidth. If you are trying for what they call “eSSB”, the Ten-Tec Jupiter is a much better fit. In fact, most Yaesu and Kenwood transceivers have a much wider configurable transmit bandwidth than modern Icom transceivers.

So, Icom is a challenge. It’s difficult to include lower or higher frequencies into the transmit audio.

I had read something about voice processor technology using harmonics to “enhance” the sound. I had a few on hand to try out.

The dbx 286s mic preamp/processor, the Aphex Aural Exciter Type C² with Big Bottom, and the Behringer DEQ2496 Ultra-Curve Pro. The first two are used regularly in broadcast and/or PA systems as voice enhancers. The methods behind this kind of processing are proprietary in each case, but essentially they add EQ and harmonics to your voice to make it sound deeper and more radio DJ-like. The Behringer Ultra-Curve Pro is already a staple in ham radio shacks as a highly configurable (and therefore quite annoying to set up) processor-in-a-box.

This was an absolute waste of time. I am sure that there is some combination of settings which will work well, but my patience ran out. There are eSSB web pages which describe in great detail how to set up the Behringer DEQ2496, but I really just don’t have the desire to painstakingly mess on and on and on with my transmit audio. I respect the people who do, and they truly do achieve remarkable results, but that is just not me. Also, as mentioned above, there is a limit to what you can achieve insofar as eSSB is concerned on a modern Icom radio.

Transmit Bandwidth

Let’s take the “WIDE” transmit bandwidth setting on the IC-7300 and IC-7610. It is essentially an audio band pass filter which has a lower frequency corner of 100 Hz, and an upper frequency corner of 2900 Hz yielding a transmit audio bandwidth of 2800 Hz or 2.8 kHz. The image to the right illustrates this range drawn on a frequency response graph of the Behringer XM8500 microphone.

Filters have a shape. In the case of the transmit audio band pass filter, it attenuates everything below 100 Hz, above 2900 Hz, and allows everything else between to go through. This shape is not square although that would be a perfect filter. The filter corner frequencies slope off gradually like this: /▔\
The 100 Hz and 2900 Hz are usually considered to be points at which the amplitude of that frequency has been reduced by half, or 3 dB.

Let’s take the 100 Hz frequency. If the audio filter in the transceiver attenuates 100 Hz by 3 dB, then if you boost the 100 Hz frequency audio by 3 dB before it is delivered to the transceiver would yield a 100 Hz level which is at 0 dB, or not attenuated at all. Similarly, if you boost the 2900 Hz frequency by 6 dB before it is delivered to the transceiver, you will have an overall 3 dB boost at 2900 Hz after the audio filter. Don’t forget that there are transmit tone controls in the IC-7610 (and the IC-7300) which allow you to cut or boost “bass” and “treble” frequencies in addition to any outboard processing.

To some extent, you can overcome the filter attenuation by boosting the low and high ends of the transmit bandwidth limits and experimenting. Experimenting is key. Every voice is different, and so everyone will have a different cut/boost EQ configuration which suits their own voice. I recommend using another radio with a sound card interface (Yaesu FT-991A in my case) to receive your transmitted audio and record it on a PC. That way you’re hearing exactly what you sound like on the air. It is interesting to experiment varying the transmit power (into a dummy load, of course) on the transmit radio and the RF gain on the receiving radio to approximate poor signal conditions.

What I settled upon

After messing around with all of that outboard audio processing gear, I actually went back to my trusty UR6QW equalizer V7. It does exactly what I want in one small package which includes a microphone preamp, noise gate, compressor, adjustable drive level, 10 band graphic EQ with ±12 dB of cut/boost, and a delay/echo function which can be used to apply a small amount of plate reverb to the audio, enhancing legibility. It also contains galvanic isolation, eliminating any ground loop hum.

My IC-7610 Audio Setup

I used a MXL V67i condenser microphone which connects via XLR to the UR6QW equalizer V7, which is then connected to the microphone connector (or the modulation input on the rear ACC1 socket) on the IC-7610. The following settings were arrived upon during an evening of iterative testing with KG5PUF using the FT-991A to receive the transmitted audio which was then recorded via USB to a PC for playback and tweaking, and also various tests using software-defined radios around the world thanks to a great collection of these at WebSDR.org. This allowed for testing in real atmospheric conditions, with fading, noise, static crashes, and at various locations around the world.

Note/Update

Soon after I published this post, I switched from the MXL V67i microphone to a Shure SM-7B I had handy. I have to say that using powered microphones like the MXL invites a lot of background noise into your audio since the gain of the microphone is way more than a dynamic mic like the SM-7B, and it can cause issues with the noise gate setting, making it difficult to squelch the noise, and that’s even before the signal hits the microphone amplifier in your chosen audio processor or radio. The EQ7 settings for the SM-7B are more or less the same as the below settings for the MXL 67i, but with more gain (overcoming the lack of phantom-powered microphone gain) and with the 2.0kHz EQ slider set one click higher than the 1.3kHz slider.

Some of the more astute amongst you will have noticed that I’m using the compressor on the IC-7610 and also the compressor on the UR6QW EQ V7. That may seem a little odd at first, but this practice is used often in recording studios, usually running the signal through an aggressive fast-acting compressor then into a softer, more rounded sounding compressor. The B172A from Black Lion Audio does this incredibly well, and I might experiment with that soon.

Why this is better than eSSB

eSSB (Extended Single Side Band) is a popular way to get that “radio DJ voice” on ham radio. There is often deep bass and crisp treble associated with eSSB and typical setups consist of a Behringer Ultragain Pro, a Behringer Composer Pro XL, and some graphic or parametric EQ. That’s a lot of relatively expensive gear to use on a ham radio just for a bit of EQ and compression.

One of the main reasons eSSB is frowned upon by some is the large, occupied bandwidth of the transmitted signal. Higher frequencies require more bandwidth. Some radios allow up to 12 kHz wide transmit bandwidth. It depends on the regulations in your part of the world, but in the USA, the regulations specify:

No amateur station transmission shall occupy more bandwidth than necessary for the information rate and emission type being transmitted, in accordance with good amateur practice.

Code of Federal Regulations, Title 47, Chapter I, Subchapter D, Part 97, Subpart D, § 97.307 Emission standards.

While it can be argued that 2.4 kHz is more than enough bandwidth to transmit intelligible speech, it is the nature of amateur radio operators to explore options – whether that’s using add-on external audio processing gear or modifying your radio to slow down the ALC and artificially (and some say needlessly) increase the average output power level, or whatever. Hams are curious, and I believe it is very important to foster that.

However, in a crowded band situation such as a contest or in times of higher sunspot activity a transmission occupying over 4 kHz can get a little overwhelming since it occupies more bandwidth than is strictly necessary which means less room on the band for many more signals which occupy less bandwidth. It sounds amazing, but it takes up 2x 2.4 kHz “channels” worth of bandwidth.

Further Experimentation

As I wrote above, I am sure that when using a bunch of outboard gear there is some combination of settings which will work well, but my patience ran out and I settled back with the UR6QW EQ V7. I do have some interesting equipment to experiment with, and I will.

I also investigated using the rear accessory jack for a modulation input. Looking at the IC-7610 schematic, the AMOD line is a more direct line to the A/D conversion than the MICI line. I was curious as to how this affected the transmitted audio compared to using the microphone input.

I built a cable to interface the Icom 8-pin microphone jack to the appropriate pins on the 8-pin DIN ACC1 connector on the rear of the IC-7610. Unfortunately (or fortunately) using the rear ACC1 jack as a modulation input sounded exactly the same as using the front microphone connector.

At least it gets the microphone plug and cable out of the way, making room for a 90° USB-A adapter which then gets the USB plug out of the way…

More experiments to come…