Running AM with the IC-756 Pro III
W1AEX
There are quite a few people running the 756 Pro series of rigs, who have discovered that it has a very nice AM receiver. The available bandwidths include 3 kHz, 6 kHz, and 9 kHz positions, with the two wider ones being capable of providing some very pleasant listening experiences, especially when a good quality speaker or audio system is connected. Unfortunately, the AM transmit side of the stock IC-756 Pro series is a somewhat different story. Even with a high quality, flat response microphone, the stock internal audio produces a very restricted and gritty sounding signal, with no processing or equalization. There is a bit of deception going on here as well, with the way ICOM chose to implement the built-in monitor of the 756 Pro series. When you listen to yourself in the monitor with a pair of headphones, your transmitted signal seems to sound pretty decent on AM. Don't be fooled by this, they are picking off the monitor signal before the DSP and ALC circuits become involved in the process. The two mp3 clips below illustrate the difference between what is heard on the monitor, and what is actually heard by stations listening to you:
Pro III with stock audio on AM as it sounds through the Pro III monitor size is 150 kb - mp3 What people think they sound like with the stock Pro III audio on AM. This audio sample was made by patching the headphone jack monitor output of the Pro III directly into the line input of one of my soundcards.
Pro III with stock audio on AM as it sounds on a Yaesu FRG-7700 size is 310 kb - mp3 What your AM signal actually sounds like to someone else over the air. Fairly restricted, with a midrange peak, and the characteristic sound of a carbon microphone from an old telephone. The gritty distortion in the audio is a by-product of the agressive ALC action as positive peaks are limited to no more than the carrier power level.
To improve the performance of the Pro III on AM, a good place to begin is to stop the aggressive action of the internal ALC. Once that's accomplished, the second step is to address the lack of audio processing in the AM mode. With both of those items addressed, the 756 series of rigs can sound very good on AM. The mp3 audio clip below is an over the air recording of my IC-756 Pro III running with external ALC and processed audio:
Pro III with external ALC and audio processing on AM size is 410 kb - mp3 What the Pro III sounds like on AM without the internal ALC and internal audio. The receiver is a Yaesu FRG-7700 in the 12 kc bandwidth position. The line-out of the FRG-7700 was directly connected to the line-in of a station computer and the recovered audio was recorded with Adobe Audition.
Defeating the Internal ALC
The aggressive clamping action of the ALC is the major obstacle to overcome when trying to attain a healthy AM signal. The ALC effect is very evident when seen on an oscilloscope that is set up to monitor the RF envelope. The dirty truth is that the Pro III is simply unable to reach anywhere close to 100% modulation with the internal ALC present. When you reach somewhere around 50 - 60 percent modulation, you will begin to see the ALC restrain the positive peaks, preventing much more than 70% modulation at best. As the rig attempts to develop peak power, the internal ALC will actually cause it to heavily downward modulate. This produces a very gritty signal with strange sounding artifacts. If you don't have a scope, you can easily view what is happening by watching the AM power output of the Pro III with an average reading wattmeter. In the video below, the Pro III is running 15 watts of AM carrier. Watch the downward deflection on the meter as the rig is modulated.
If schematics scare you, then the picture below will reassure you. This takes less than 5 minutes to build, and it has a dramatic effect on how the Pro III behaves while running in the AM mode. The positive terminal of the 3 cell battery pack is connected to one end of the potentiometer, along with the shield side of the connecting cable, so that it's grounded when connected to the rig. The negative terminal of the battery pack is connected to the other side of the potentiometer. The potentiometer wiper is connected to the center conductor of the RCA jack, allowing an adjustable ALC range from 0 volts to negative 4.5 volts. The 756 Pro series of rigs only require an ALC control voltage range from 0 volts to negative 4 volts so after initial testing with this circuit I inserted a series diode into the lead going to the rig. This drops the voltage to -3.9 volts and also prevents a positive voltage from inadvertenly being fed into the transceiver's ALC jack.
The effects of this circuit are quite dramatic when operating the Pro III in the AM mode. It's very important to keep the internal ALC out of the picture, so to accomplish this, set the front panel RF power level of the Pro III to maximum output. Then plug the external ALC circuit into the jack on the back panel. Key the Pro III while in the AM mode and then use the potentiometer to set the output of the Pro III to approximately 20 watts of output. Modulate the rig, and observe that it now is showing a healthy amount of upward deflection on an average reading power meter. The video below gives a good indication of the improvement when compared to the first video at the top of the page.
It would be a very simple matter to build up a negative power supply with an LM-7905 regulator to eliminate the battery in the circuit above, if you wished to make something a bit more permanent. The Pro III is designed to accept negative voltages from 0 to minus 4 volts through the back panel ALC jack. Just remember, don't ever apply a positive voltage to the ALC jack on the back panel. This can cause serious issues to several devices that are in the internal ALC circuit path. When running SSB, CW, FM, or any of the digital modes, it would be best to disable this circuit. It shouldn't cause any harm, but it could cause errors while using the digital modes, and would be a hinderance to proper operation with modes other than AM. Also, keep in mind that as long as the battery is plugged in it will slowly discharge through the 1 megohm pot.
Simple Processed Audio for
Running the Pro III on AM
When I contemplated using external audio, it dawned on me that I already had an audio chain in place, through the digital mode interface between my station computer's sound card and the Pro III. The interface that I use is a SignaLink SL-1+ but anything that offers a decent audio transformer to isolate the sound card and the Pro III would work just fine. The transformer in the SignaLink SL-1+ is a 600 ohm to 600 ohm transformer that looks very much like what was used on old telephone modem cards. The setup is very simple as illustrated in the picture below.
I grabbed one of my 5 dollar, home-made, flat response condenser microphones and made a shielded cable with an XLR connector on one end and a 1/8 inch stereo plug on the other to allow connection to the microphone input of a sound card. It sounded very nice with headphones plugged directly into the computer, however, when I tested the audio through the Pro III it sounded very boomy and muffled. Obviously, some audio processing was going to be necessary. A search of readily available programs brought me to an amazing piece of sofware called Voice Shaper.
Voice Shaper Audio Processing Software
This amazing little program, written by Alex, VE3NEA, is ME/2000/XP compatible and allows 7 points of EQ adjustment between 70 Hz and 6000 Hz. It also offers a compressor, limiter, noise gate, and a brick wall audio bandpass filter as well. Another nice feature is that you can save as many different presets as you like. The nicest surprise is that this amazing little program is free! The learning curve with this software is very reasonable, so with a little experimentation you can produce many different audio profiles. I won't get into the finer points of using this program since the author covers this in great detail with an online tutorial. The best thing to do is to play around with all the features and see for yourself how it works. The screenshot below shows one preset that I use on AM when conditions are good. The compression is set at 1.2:1 and the Output Gain is running at -12 dB.
The real truth is in what
is actually heard with the Pro III and external sound card
audio processing. Using the EQ curve in the screenshot
above, and a compression setting of 1.2:1 I recorded a few
mp3 files of the Pro III in action. The sound clips
derived directly from the Pro III's headphone jack tell
the story about how deceiving the built-in monitor of the
Pro III can be. What you hear through the Pro III monitor
is NOT what people hear on their receivers during an
actual QSO!
The two audio clips below are samples of the Pro III with Voice Shaper processed audio injected into the back port.
Pro III with Voice Shaper audio on AM as it sounds through the Pro III monitor size is 340 kb - mp3 Wouldn't it be nice if it actually could sound like this? This audio sample was made by patching the headphone jack monitor output of the Pro III directly into the line input of one of my soundcards. Obviously, the monitor tap in the Pro III picks off the sample before the DSP and ALC do their work.
Pro III with Voice Shaper audio on AM as heard by a Yaesu FRG7700 with its 12 kc AM filter size is 410 kb - mp3 In this audio clip, the ALC has been defeated, and processed audio is being fed into the back panel of the Pro III. This is a best case receive scenario, and represents what is heard with a receiver designed for AM reception. It sounds much better than what is heard through most sideband receivers where they tend to use a narrow filter in the AM mode. You can still hear things happening on some sibilant peaks, but the audio is generally smooth and not too hard on the ears. This audio clip was made by patching the line output of the FRG-7700 into the line input of one of my soundcards.
I can still hear some crunching evident on sibilant
sounds, but it probably would not be too apparent on the
other end of a QSO. The brick wall audio filter in Voice
Shaper does a great job of stopping speech frequencies
that are way beyond what the rig's DSP will tolerate. This
is a big problem that people run into when they use one of
the cheap 3 band type mic preamp/equalizers such as the
little Behringer 802. That third EQ band is totally
outside the range of the Icom's DSP, and if you attempt to
crank it up, not much high frequency energy will get
through.
I believe the trick is to insert some presence rise near 3000 Hz to avoid sounding muffled. Under rougher conditions, simply make a profile that rolls off the low end and then increase the Voice Shaper compression level and boost the slider near 3000 Hz to brighten the audio response. Again, don't rely on the Pro III monitor when making audio level adjustments, it simply does not accurately tell you what is happening. Depending on the capabilities of your computer, you will experience some degree of latency that will be noticable if you monitor yourself. The slower the computer is, the more latency you will experience. It was quite noticeable (more than half a second) when I used an older system with a 2.4 GHz processor and improved drastically when I installed my recently retired gaming system that uses a 3.0 GHz hyperthreading processor. Annoying audio dropouts may also occur with a slower computer, so keep this in mind if you intend to use an older system.
When setting up your transmit audio it is best to use an easily built RF sampling monitor such as the one shown in the diagram below. This circuit is based upon a design from the East Coast Sound section of the AMfone archives. There is a huge repository of excellent AM information at the amfone.net site! The original design works fine for power levels up to about 30 watts, but it tends to emit smoke at higher power levels! This one has handled power levels up to 175 watts without any difficulty. I run the audio output of the detector below into a control box with a relay that switches my headphones between the receiver output and the AM Diode Detector output whenever I key the rig. I set the "Audio Level Adjust" variable resistor to match the volume level of the receiver. When used with good quality headphones, this monitor circuit will produce a very accurate representation of your transmitter's AM audio. It's easy and inexpensive to build and it works fine!
As an alternative, you can use a receiver and plug some headphones in to allow you to make level adjustments through your soundcard software and the Voice Shaper software. Keep in mind that there may be a bit of overload induced blocking going on when using a second receiver, but it will give a pretty good indication of what your audio sounds like. I found that it was critical to watch the sound card input and output levels to prevent getting into a situation where distortion was introduced before audio was injected into the Pro III. A very helpful utility called QuickMix can be used to save your Windows mixer settings to a file when you find settings that are acceptable. When other programs alter the settings you can load your audio profile back into the Windows mixer with the QuickMix utility. I have three sound cards in my station computer, so it was essential to use QuickMix to keep track of the settings for all three of them! The sound card I use for producing audio for the Pro III is a cheap Sound Blaster Live PCI card. QuickMix is a free program that can be downloaded here:
QuickMix Soundcard Settings Utility
I found input and output sound card level settings that produced clean audio which then allowed me to use the "Output Gain" slider in Voice Shaper to control the audio level on the fly. This setting is calibrated in dB and makes it simple to fine tune your settings. This setting is also "remembered" when you save each Voice Shaper profile onto your hard drive. After a lot of help from others over the air, I found a setting that produced a clean, well modulated signal without sounding overdriven. Your own settings will vary from mine, but you will find a setup that will work with your equipment. I pretty much leave all the levels alone now, and have made up 5 or 6 different Voice Shaper profiles that can be switched on-the-fly to accomodate conditions and the receiver being used on the other end.
One other important factor to consider with the Pro III is the power level it is used at in the AM mode. I have found that if you climb much above 20 - 25 watts output things begin to fall apart rather quickly. Typically, I run between 10 - 15 watts into my AL-80B 3-500 amp to produce between 100 - 150 watts of carrier output. If my audio and RF levels are set carefully, and external ALC is applied, with an average reading wattmeter I can see the meter deflect upward very slightly on audio peaks. As expected, a peak reading wattmeter will show a greater increase in power as you modulate. Unfortunately, even with the external ALC applied, the 756 Pro series of rigs cannot reach 100% modulation. With a 1000 cycle tone, the rig reaches close to 90% modulation and then distortion rears its ugly head, as the video at the link below shows:
ICOM Pro III Scoped On AM (Video hosted at YouTube)
The 90% number is no accident, as this is the target value the DSP is aligned to allow from the factory. The service manual details the adjustment of the AM modulation setting on the bottom of page 4-6. Using the adjustment setup conditions they described (1000 cps tone at 10 mVrms applied to the front panel MIC connector with MIC level at center and RF power at maximum) I confirmed that my unit was just about reaching the 90% mark with no distortion evident. Being curious, I peeled back the tape that covers the adjustment ports on the DSP and made the attempt to adjust it to allow a higher percentage of modulation.
I quickly found that for every 1% gain in the positive direction, the DSP moved about 5% in the negative direction. Needless to say, with this behavior, the rig hit the baseline in very short order, and the best I could do in the positive direction was a few percent above 90%. It appears this adjustment is a dead end and the investment in time is not worth the effort for such a small gain. Most users have found that with more complex voice frequencies, typically 80% - 85% modulation is about all that is seen during a QSO. This is a limitation that appears to be inherent with the way ICOM has implemented the AM mode in the 756 Pro series of transcievers. With that being said, I don't recall any complaints about my audio sounding low when I use my Pro III on AM.
One last thought. Keep in mind that you can use the same external audio chain to experiment with other voice modes as well. You can easily tailor the audio beautifully on SSB and FM. If you have an interest in using your IC-756 Pro series transceiver on AM, you might want to give these simple methods a try. If you already have a computer and a digital interface, or even a simple 1:1 transformer, you won't need to invest in a pile of hardware to get decent results.
The two audio clips below are samples of the Pro III with Voice Shaper processed audio injected into the back port.
Pro III with Voice Shaper audio on AM as it sounds through the Pro III monitor size is 340 kb - mp3 Wouldn't it be nice if it actually could sound like this? This audio sample was made by patching the headphone jack monitor output of the Pro III directly into the line input of one of my soundcards. Obviously, the monitor tap in the Pro III picks off the sample before the DSP and ALC do their work.
Pro III with Voice Shaper audio on AM as heard by a Yaesu FRG7700 with its 12 kc AM filter size is 410 kb - mp3 In this audio clip, the ALC has been defeated, and processed audio is being fed into the back panel of the Pro III. This is a best case receive scenario, and represents what is heard with a receiver designed for AM reception. It sounds much better than what is heard through most sideband receivers where they tend to use a narrow filter in the AM mode. You can still hear things happening on some sibilant peaks, but the audio is generally smooth and not too hard on the ears. This audio clip was made by patching the line output of the FRG-7700 into the line input of one of my soundcards.
I believe the trick is to insert some presence rise near 3000 Hz to avoid sounding muffled. Under rougher conditions, simply make a profile that rolls off the low end and then increase the Voice Shaper compression level and boost the slider near 3000 Hz to brighten the audio response. Again, don't rely on the Pro III monitor when making audio level adjustments, it simply does not accurately tell you what is happening. Depending on the capabilities of your computer, you will experience some degree of latency that will be noticable if you monitor yourself. The slower the computer is, the more latency you will experience. It was quite noticeable (more than half a second) when I used an older system with a 2.4 GHz processor and improved drastically when I installed my recently retired gaming system that uses a 3.0 GHz hyperthreading processor. Annoying audio dropouts may also occur with a slower computer, so keep this in mind if you intend to use an older system.
When setting up your transmit audio it is best to use an easily built RF sampling monitor such as the one shown in the diagram below. This circuit is based upon a design from the East Coast Sound section of the AMfone archives. There is a huge repository of excellent AM information at the amfone.net site! The original design works fine for power levels up to about 30 watts, but it tends to emit smoke at higher power levels! This one has handled power levels up to 175 watts without any difficulty. I run the audio output of the detector below into a control box with a relay that switches my headphones between the receiver output and the AM Diode Detector output whenever I key the rig. I set the "Audio Level Adjust" variable resistor to match the volume level of the receiver. When used with good quality headphones, this monitor circuit will produce a very accurate representation of your transmitter's AM audio. It's easy and inexpensive to build and it works fine!
As an alternative, you can use a receiver and plug some headphones in to allow you to make level adjustments through your soundcard software and the Voice Shaper software. Keep in mind that there may be a bit of overload induced blocking going on when using a second receiver, but it will give a pretty good indication of what your audio sounds like. I found that it was critical to watch the sound card input and output levels to prevent getting into a situation where distortion was introduced before audio was injected into the Pro III. A very helpful utility called QuickMix can be used to save your Windows mixer settings to a file when you find settings that are acceptable. When other programs alter the settings you can load your audio profile back into the Windows mixer with the QuickMix utility. I have three sound cards in my station computer, so it was essential to use QuickMix to keep track of the settings for all three of them! The sound card I use for producing audio for the Pro III is a cheap Sound Blaster Live PCI card. QuickMix is a free program that can be downloaded here:
QuickMix Soundcard Settings Utility
I found input and output sound card level settings that produced clean audio which then allowed me to use the "Output Gain" slider in Voice Shaper to control the audio level on the fly. This setting is calibrated in dB and makes it simple to fine tune your settings. This setting is also "remembered" when you save each Voice Shaper profile onto your hard drive. After a lot of help from others over the air, I found a setting that produced a clean, well modulated signal without sounding overdriven. Your own settings will vary from mine, but you will find a setup that will work with your equipment. I pretty much leave all the levels alone now, and have made up 5 or 6 different Voice Shaper profiles that can be switched on-the-fly to accomodate conditions and the receiver being used on the other end.
One other important factor to consider with the Pro III is the power level it is used at in the AM mode. I have found that if you climb much above 20 - 25 watts output things begin to fall apart rather quickly. Typically, I run between 10 - 15 watts into my AL-80B 3-500 amp to produce between 100 - 150 watts of carrier output. If my audio and RF levels are set carefully, and external ALC is applied, with an average reading wattmeter I can see the meter deflect upward very slightly on audio peaks. As expected, a peak reading wattmeter will show a greater increase in power as you modulate. Unfortunately, even with the external ALC applied, the 756 Pro series of rigs cannot reach 100% modulation. With a 1000 cycle tone, the rig reaches close to 90% modulation and then distortion rears its ugly head, as the video at the link below shows:
ICOM Pro III Scoped On AM (Video hosted at YouTube)
The 90% number is no accident, as this is the target value the DSP is aligned to allow from the factory. The service manual details the adjustment of the AM modulation setting on the bottom of page 4-6. Using the adjustment setup conditions they described (1000 cps tone at 10 mVrms applied to the front panel MIC connector with MIC level at center and RF power at maximum) I confirmed that my unit was just about reaching the 90% mark with no distortion evident. Being curious, I peeled back the tape that covers the adjustment ports on the DSP and made the attempt to adjust it to allow a higher percentage of modulation.
I quickly found that for every 1% gain in the positive direction, the DSP moved about 5% in the negative direction. Needless to say, with this behavior, the rig hit the baseline in very short order, and the best I could do in the positive direction was a few percent above 90%. It appears this adjustment is a dead end and the investment in time is not worth the effort for such a small gain. Most users have found that with more complex voice frequencies, typically 80% - 85% modulation is about all that is seen during a QSO. This is a limitation that appears to be inherent with the way ICOM has implemented the AM mode in the 756 Pro series of transcievers. With that being said, I don't recall any complaints about my audio sounding low when I use my Pro III on AM.
One last thought. Keep in mind that you can use the same external audio chain to experiment with other voice modes as well. You can easily tailor the audio beautifully on SSB and FM. If you have an interest in using your IC-756 Pro series transceiver on AM, you might want to give these simple methods a try. If you already have a computer and a digital interface, or even a simple 1:1 transformer, you won't need to invest in a pile of hardware to get decent results.