AGC Slope

Tim K9WX

I’m two weeks and 500 QSOs into my Flex 6600 with a Maestro.  Most of those QSOs are CW with some FT8 and RTTY.

I have a number of questions based on my experiences to date, especially today when I operated SO2R CW for the first time.  I will raise those questions here as a series of postings, so that anyone with advice or insight can respond selectively.

The operating setup is identical for all questions.

·        SmartSDR 2.2.8

·        N1MM+ 1.0.7201.0

·        Windows 8.1

·        Headphones are plugged into the radio, not the Maestro, speakers are not used

·        Paddles are plugged into the key jack on the radio and the Iambic setting set to Enabled and to B

Q2: I’ve become familiar with the use of the AGC-Threshold control but find there is too much variability or range between the weakest and strongest signals, requiring me to turn the AF knob up to hear weaker signals and down for louder signals.  Some radio manufacturers incorporate a “slope” function in their AGC so that signal response beyond a certain level can be relatively flat and signals that exceed this level are close in AF output level.  Is there a similar adjustment available in the Flex?  

Tim K9WX

Mike-VA3MW

Hi Tim

I had one of the engineers explain to me the best way to set up AGC-T.  I need to do this in a video, but in the mean time.

• Select your antenna and pre-amp/attenuation 
• Select a frequency void of signals based on your RX bandwidth
• Start with the AGC-T turned up and slowly bring it down until you hear a reduction in noise.  This will only affect the noise, but not the received signal.

That is now the correct setting for AGC-T.  Correct setting of AGC-T should manage this.  I pulled the following from the manual just in case you hadn't had a chance to review it and for others who  might not have seen it.  (page 52)

AGC-Threshold

The AGC-T has 2 modes of attack.  1 mode attacks the short lived hi burst signals like lightning and the other is the more normal slow or fast attack rate.  

Automatic Gain Control (AGC) is a feature which automatically adjusts the Slice Receiver’s audio gain (volume) based on the strength of signals in the receiver’s passband filter. The goal of AGC is to amplify weak signals and attenuate strong signals so that they all lie within a comfortable listening range.

The receiver Automatic Gain Control Threshold (AGC-T) can be adjusted for optimum performance in noisy or quiet environments. The AGC-T sets the maximum gain applied under any circumstances. Since the noise floor is relatively constant on a given band at a given time, the AGC can be adjusted using the threshold control so that the AGC never applies gain to noise, but it will apply gain to signals just out of the noise. In doing so, the AGC can reduce the level of noise you hear, and help signals pop out of the noise.

The AGC system in SmartSDR is a dual track system, meaning that it can track both slow and fast increases in signal strength, making appropriate gain correction decisions in the presence of each. The speed of the AGC (FAST, MED, SLOW) determines how quickly or slowly the AGC recovers after attenuating a strong signal. You can easily hear this by tuning to a CW signal and going through the three settings. On FAST with a strong signal you can hear the gain pump up and down while on SLOW it recovers after a longer pause once the signal stops. When the signal stops, you will hear the noise floor increase as the gain returns
.
How to Set AGC for Different Operating Conditions

If you are listening to a loud voice signal, AGC SLOW will resist increasing the gain between syllables and therefore reject most of the noise which is at a level far below the signal. FAST and MED provide faster levels of recovery for situations when you want the system to more closely follow the dominant signal in the passband. Any time you have a very strong signal that causes the gain in the AGC to be reduced, you could experience a loss of gain to a weak signal you are listening to. The filter passband edges, which are continuously adjustable, and TNFs can be used to eliminate signals that might interfere with AGC operation.

The operator might prefer to use SLOW settings when rag chewing in a high signal to noise environment where there isn't much QRN and the noise floor is stable. This keeps the gain at more of a constant level that is less distracting. If the operator is trying to pull a weak CW signal out of the noise, they may prefer to use FAST mode to quickly ensure that the long-term average of the noise floor doesn't overcome the signal and prevent it from being heard. MEDium is a reasonable compromise.

When AGC is set to OFF, a fixed amount of gain determined by the AGC-T setting will be applied to both fast and slow signals regardless of their level. The more you increase the AGC-T, the more gain is applied and the louder the signal and noise will be. The benefits of increasing SNR with AGC are lost in this mode. Another disadvantage of turning AGC off is that the operator must adjust the 'RF Gain' manually to avoid distortion due to overload by strong signals.

Operating with AGC turned OFF may be desired by operators who want to avoid having a strong signal drive a weak nearby signal into the noise floor resulting from AGC audio attenuation, such as when operating digital modes.

To adjust the AGC Threshold, tune to a quiet spot between stations. Starting with the AGC-T at a high value adjust the AGC-T slider to the left (lower gain values) until the background noise just begins to decrease. This is the AGC-T "sweet spot" or the "knee" of the AGC algorithm. Depending on band conditions, if the AGC-T is set below 50, you may have to compensate for the loss in audio gain (volume) by increasing the Slice or master AF volume to a higher value. When you get the knee and AF volume adjustment correct for the band conditions, it will keep the volume of strong signals constant which will allow weaker signals to be heard even with AGC in FAST mode. Thus AGC-T is one of the most important adjustments, and often overlooked, to achieve the maximum weak signal receive performance out of the FLEX- 6000 series SDRs.

Varistor

It is important to point out that setting AGC-T, or the RF gain of the radio, is not a "set it and forget it" step. That is, it must be done as band conditions change and be actively managed.

Required RF gain is also different from band to band. For example, it is a lot less on the lower bands and higher on 15 and 10.

Tim - W4TME

True dat!

Matt NQ6N

I've been curious about the question of AGC Slope with the 6000 series for a while so I was happy to see this question asked.  (There is a definition of AGC Slope provided in the below article):

http://www.qsl.net/va3iul/Files/Automatic_Gain_Control.pdf

From my google search it appears that PowerSDR had a configurable "Gain Slope" for AGC. But in reading the description of this it's not clear whether it's different from the AGC-T setting in SmartSDR.

It seems that by definition the AGC-T is the "knee", and the AGC-Slope determines how much gain reduction is applied based on the strength of the input signals which exceed the threshold. 

I'd be very appreciative if someone could describe in detail the architecture of the 6000 series in the context of the diagrams on page 2 of the above paper.  Does the AGC-T implicitly adjust the slope? What was the thought process for not including a user-adjustable AGC-Slope?

73,
Matt NQ6N

Mike-VA3MW

Hi Matt

Did you get a chance to review the reply to the original question?  It did cover most of what you ask.

Mike

Matt NQ6N

Mike - I did, it talks about AGC-T but not about AGC-Slope at all, or about the design decisions in whether to make that user configurable.  That's why I typed up my question

73,
Matt NQ6N

Tim K9WX

Matt, thanks for posting this info. What you describe is how I have been setting the AGC-T and I am happy with the way that works. But I agree with Matt NQ6N that the info you provided does not address my question about AGC slope. He has expressed the question more eloquently than I did.

Mike-VA3MW

I poked one of our awesome engineering team.  They will dive under the covers for you on this. 

Eric-KE5DTO

So AGC is a complex animal.  Anyone who thinks otherwise hasn't ever implemented one.    The AGC system in SmartSDR is a dual track system.  Both tracks work in a similar fashion, but each is targeted at different response times.  This helps prevent fast impulse type spikes from dominating a slower changing signal (such as a sideband signal).

The AGC-T slider is actually setting the maximum gain that can be applied to any signal.  We always attempt to bring the incoming signal up to some target level, but at times, it doesn't make sense to apply that much gain.  For example, when amplifying the noise floor, you don't want to hear that at the full target level (if you want to try this, just slide the AGC-T all the way up -- it isn't pleasant).

As others have mentioned, band conditions can affect this since what is being adjusted is the max gain against the input signal.  So if the ambient level of noise changes, this will definitely change how things sound.

Today there is no slope applied.  We use a flat response.  There are several implications from this.  First, if 2 independent signals are evaluated one-at-a-time such that each of them are strong enough to hit the target level without hitting the max AGC gain, they will play at exactly the same audio level (i.e. the target).  This is the AGC doing it's job.  Even if one of these signals is significantly weaker than the other, they will sound the same (again -- this is when they are evaluated one-at-a-time).

Second, if you put both of those signals inside the same passband, the stronger signal will dominate the AGC and the weaker signal will sound quieter.  If one signal is 10dB stronger than the other, the weaker signal will sound 10dB weaker in our audio.  This is the direct no-slope mapping in play.

I agree with Mike that a combination of adjusting the filter (to eliminate an offending signal) and adjusting the AGC-T should give the desired result of a comfortable audio level for a wide dynamic range of signals.

tl;dr SmartSDR does not use a slope AGC.

Alan - KA4B

Eric, I am interested in knowing more about the 'dual track system' and how slow, medium and fast relate to the two tracks.  My question has to do with minimizing the impact of lightning crashes on ssb signals on 80 meters and increasing intelligibility without damaging the hearing.  Thanks!

Ken - NM9P

If you have lots of lightning on 80, I surely wouldn’t use Slow. But medium should be OK.

Mark K1LSB

Eric, I have good results with lightning crashes on 75 meters by setting the AGC Speed to Medium and sliding the AGC-T control toward the left -- keep in mind that the crashes will be reduced but so will your ability to pick up weak signals.

Bill -VA3WTB

Such is life in the world of communications...

Matt NQ6N

Thanks for the excellent explanation, Eric!   I very much like the AGC-T control and have had great success using it in combination with the narrowest filter bandwidths to dig out incredibly weak CW signals. 

Marty Ray

To mitigate the effects of static crashes when using CW, I set the sidetone to a very low frequency, (less than 400Hz), and use a very narrow filter. 73, Marty N9SE

Eric-KE5DTO

Sorry for the delayed response as I have been out of the office for a week.  So the dual track has 2 different time constants that it tracks and both of these are affected by the AGC Mode (Slow, Med, Fast).  It isn't really any more complicated than that the time constants are longer for Slow than Med, and for Med compared to Fast.  For any one of these settings, the dual track has a "fast" track with a much shorter time constant for handling impulses, static crashes and such without capturing the AGC making the radio appear deaf for many seconds while it recovers.