Talk:Automatic gain control

Untitled
The discussion in the article, currently in the second paragraph, about the operation and effects of channel fading in AM radio, and how it can be repaired with an automatic gain control, is suspect.


 * I don't clearly understand the AM radio example in this article. Is there maybe Automatic gain control (AGC), because the power the AM radio receives gets less and less the further we move away from the sending mast? Thus, the signal nets to be amplified (i.e. "normalized"), and that just is what AGC does?


 * Thanks, --Abdull 18:14, 4 Apr 2005 (UTC)


 * Yes, you are correct - as you move away from the transmitter, the signal strength you receive on your radio gets weaker, and the AGC automatically increases the amplification of the signal so that you don't notice any difference when listening to the radio. The signal can also be affected by other factors, such as another transmitter operation nearby with a similar frequency, which interferes with the transmission you are trying to pick up. AGC will automatically adjust the gain to take care of this too. Obviously it also works in reverse - if you get closer to the transmitter, the signal will become stronger, and the AGC will reduce the amplification so that your radio doesn't get louder.


 * Hope this helps -- Kim 8:00, 18 Jul 2005 (UTC)

It is usually [not] advantageous to reduce the gain of the front end of the receiver, which has benefits for both the amount of noise in the signal, and performance in the presence of strong signals, such as blocking rejection.

I removed the 'not' in the above. I believe it's incorrect to say that running the front end at full gain improves performance on strong signals - it does the opposite! The reason receivers block is that the front end gets totally saturated, pushed well into its non-linear region and effectively 'jammed'. If the gain is reduced, it can remain linear and handle the signal. Maintaining linear operation is also essential for reducing unwanted intermodulation distortion, and lower gain means lower noise too. This applies equally to AM and FM receivers - good FM designs will have AGC on the front end for exactly these reasons - it is a frequently mentioned fallacy that 'FM receivers don't need AGC'. Yes they do if they are to be any good! Cheap designs might omit it, but professional systems never do, especially for narrow band FM receivers. FM and AM receiver AGC design is different however - in FM, only the front end is usually adjusted, the IF stages are allowed to saturate to take maximum advantage of the capture effect, and these stages operate at full gain the whole time. This is OK since the front end and IF conversion keep out-of-band signals out of the IF chain (and the front-end AGC keeps the input signal level at a comfortable level that the IF chain can handle without blocking). AM receivers on the other hand must always be linear to preserve the modulation. Thus the two designs will differ considerably, with no AGC in the IF chain on FM, but AGC in the IF chain on AM. Both designs will benefit where AGC is applied to the front end.Graham 01:10, 24 December 2005 (UTC)


 * THis sounds like you are talking about gain compression--Light current 07:45, 4 February 2006 (UTC)

Adaptive and automatic
Is adaptive gain control the same as automatic gain control? —ShaneCavanaugh 07:03, 8 February 2006 (UTC)


 * Depends on how it adapts I suppose. 'Adaptive' sounds more intelligent to me than automatic. AGC is just designed to keep the output level constant with varying input. What does adaptive do?--Light current 20:24, 8 February 2006 (UTC)


 * Not all AGC systems keep the output level constant. Some compress the dynamic range, i.e. a compressor, where, say a 40 dB range at the input becomes a 20dB range at the output (2:1). Others might do this over some range then limit to a constant level above a certain threshold, like a VOGAD. Perhaps this is the difference.... but I've not heard the term adaptive gain control as a very widely used term. Graham 23:41, 8 February 2006 (UTC)

Telephone example
I don't see what the current telephone example has to do with automatic gain control. --Abdull 18:41, 2 March 2007 (UTC)

Biological? Really?
We have to insert natural selection into this article? Seems as though this could also fall under intelligent design- why even have that discussion on this page? —Preceding unsigned comment added by 141.131.3.22 (talk) 19:31, 13 October 2010 (UTC)


 * I reworded it; I agree there's no reason to invoke evolution here. Dicklyon (talk) 22:03, 13 October 2010 (UTC)

Is there a point to mentioning a specific brand name?
"Vactrol" redirects to "Resistive opto-isolator." The caption text for the schematic of an AGC circuit featuring a resitive opto-isolator said "...a Vactrol resistive opto-isolator" with "Vactrol" wikilinked. The only edit I made in the article was to remove the "Vactrol" and wikilink "resistive opto-isolator" instead. My edit was reverted with a comment of "More informative the other way." Someone please explain to me how using redundant terms and wikilinking to a redirect page is "more informative." -- Joe (talk) 01:26, 9 May 2013 (UTC)


 * The point of leaving Vactrol is that it's the part used in the image, according to the source; it's also a widely used and widely known device name. The point of linking Vactrol rather than resistive opto-isolator is that it would be harder to know what Vactrol refers to the other way.  As the linked article  points out, it's a genericized trademark; so probably we should downcase it.  Dicklyon (talk) 02:05, 9 May 2013 (UTC)

AGC in the 30s
Article said: 'By the early 1930s essentially all broadcast receivers included automatic volume control'

I've fixed this as its so plainly wrong
 * 1) First a sizeable percentage of the market in the early 30s was homemade receivers, often of the 2 or 3 valve reaction type, few of which featured agc.
 * 2) Second, a large %age of radios in the early 30s were 1920s sets, which did not normally feature agc
 * 3) Crystal radios, regenerative sets and other such lower cost technology radios formed a large percentage of the radio market in the early 30s, and didn't generally feature agc.

The reference is someone's book in which he says "AM radio receivers incorporating this AVC circuit came into use about 1930, and it has been included in every set since then" which, quite simply, is not true. A look at any collection of early 30s sets soon confirms this. My 1934 set has no agc.

In his book he also says "This was an invention made by Wheeler in 1925, which maintained a constant sound level from a radio receiver while tuning to various broadcasting signals of differing strengths" which again is not true. It is inherent in the design of simple/early AGC that there is output variation with input level variation. (Think about it for a moment, the ouput level controls the gain, yet he proposes constant output level, which implies constant gain, with widely varying input level. Its impossible.) Its only the practice of agcing the last amplifier stage while deriving the agc signal from its input, not output, that brought agc close to the goal of constant output with variable input.

82.31.66.207 (talk) 00:46, 10 October 2013 (UTC)

Confusion
> AGC is a departure from linearity in AM radio receivers.[3]

no, change in gain does not equal nonlinearity. A key point about agc is that gain is changed without losing linearity. That someone writes something in a book does not make it so :)

> Without AGC, an AM radio would have a linear relationship between the signal amplitude and the sound waveform – the sound amplitude, which correlates with loudness, is proportional to the radio signal amplitude,

no, its proportional to the change in rf amplitude 82.31.66.207 (talk) 21:57, 10 October 2013 (UTC)

More issues
> To produce the least noisy recording, the recording level should be set as high as possible without being so high as to clip or seriously distort the signal.

Not really, its not that simple. Making full use of tape's dynamic range causes hf compression, hence its desirable to avoid doing so. 82.31.66.207 (talk) 07:43, 11 October 2013 (UTC)