IF AMP Queston

AM receivers have AGC to adjust the gain up and down. FM radios have limiting.

To expand on this a bit:

Typically you want just enough RF gain prior to your mixing stage to maintain a good S/N ratio. Typically this is going to be about 20 dB of gain. But you do not want too much RF gain (for a robust receiver

- always exceptions to the rule) because of front end compression issues.

As a result, most of your gain will be in your IF section, but becuase the incoming signal can vary in strength (assuming you want a high dynamic range receiver) then you can vary the gain (or limit it in FM) as stated above. (Or, you can use a log amp for pulse detection ).

On a sunny day (Mon, 22 Dec 2008 06:53:21 -0800 (PST)) it happened MooseFET wrote in :

You can then perhaps specify a minimum gain and a maximum gain.

I would think that the maximum gain for the IF amp would be: the minimum voltage required at the detector divided by the minimum output from the mixer.

And the minimum gain would be: the maximum voltage allowed at the detector divided by the maximum output at the mixer.

The ratio of maximum gain to minimum gain is the required AGC range.

With 'minimum output at the mixer' I mean for a sensible signal to noise ratio.

The maximum voltage at the detector is likely limited by distortion criteria, clipping perhaps.

The minimum voltage at the detector is limited by the sort of AM diode used, and, in case it is a product detector, maybe also by signal to noise.

The maximum voltage at the mixer output is limited by the mixer type and circuit.

Have I covered everything now?

blah blah blah. OP ... strange question, why, what, how. All philosophy.

Say you want to pick up aerial signals down to ~1uV. You'll be wanting to listen to these on a speaker or headphones, so say ~1V of audio. That's fixed the maximum overall voltage gain for the complete radio to 1 million. The audio amp' might have a gain of 10, hence the total RF gain needs to be ~100,000. Sounds a lot but gain is cheap. The clever bit is using it with loving care.

F***,Sh**. "RF" wrong. Read "IF". (Rare to find RF gain).

a-

s

A put down is not very effective if nobody knows what you are talking about.

6da-

eds

IF amplifiers used in HF receivers often have a gain of about 60 dB.

Leon

a-

s

RF gain is very common on high-spec. receivers, like the Redifon R551N I have.

Leon

I'd say it's even pretty common on mid-grade receivers, although in the form of a broadband amp rather than a tuned amplifier like high-end units have.

ote:

6da-

=BF=BDneeds

Twas same person berating himself :)

Rule of thumb: the IF amp has to have enough gain to compensate for the losses of the mixer(s) and filter(s) in the IF stage, but not much more.

For best overall noise figure, the amplifier gain closest to the receive antenna is the most important. It should be high enough to make noise contributions from the IF stages insignificant, but not so high that the signal level will be too high for the first mixer.

IF amp gain around 20 dB or so is common.

. . Really? Decades of engineering work out the window. Most of us have shot for somewhere between 50 and 80 dB and found that span decent to work with. Sorry, the RF amplifier stage sets the noise figure to a first approximation. All the stuff behind is is second order at best.

Yeah, I know. Deep space picovolt work requires something on the order of

100 dB (what a crystal bitch to tame) and crackerbox AM radio can get by with 40 dB, but 20 dB is laughable.

Who is this idiot? Wasn't he the one that was trying to adapt AM sideband principles to FM modulation? Somehow I don't think either Bessel OR Armstrong is his friend.

Jim

--
"It is the mark of an educated mind to be able to entertain a thought 
without accepting it."
        --Aristotle


...
> Rule of thumb: the IF amp has to have enough gain to compensate for the 
> losses of the mixer(s) and filter(s) in the IF stage, but not much more.
>
> For best overall noise figure, the amplifier gain closest to the receive 
> antenna is the most important.  It should be high enough to make noise 
> contributions from the IF stages insignificant, but not so high that the 
> signal level will be too high for the first mixer.
>
> IF amp gain around 20 dB or so is common.

If you only have 20 dB IF gain, where is the bulk of the gain going to be achieved? Something like 100 dB total is needed to amplify a 1 uV signal to something that is audible. And what about AGC? It won't be very effective with only 20 dB to play with.

Leon

shot

h.

I would have put it as the RF amplifier and mixer but once past there it is more like a second order approximation that the noise numbers are set.

sqrt(1^2 + 0.5^2) =3D 1.12

If your RF stage only gives you 6dB of gain it dominates the noise issue.

of

The earth bound end has the advantage of being able to spread things out. The amazing stuff is the stuff that has to be kept light enough to launch.

An old tube AM radio I fixed many years back had less than 40dB in the IF. The gain was something like 45 or 50. It could receive about the same stations as I could get on my crystal radio. Its only advantage was that you could hear them one at a time instead of all mixed together. I think that even 40dB may be a little low for a good AM radio.

and

Since the whole world is going digital, I'll give this one away:

If you single side band modulate the upper side band with a signal and modulate the lower side band with minus the signal and radiate the two sidebands as the carrier, you have a signal that can still be understood if it get clipped along the way. It still sounds like hell but you can make out what is being said.

. e e e

Expound, please? What do you mean "minus the signal" on the LSB. Let's assume a 1 kHz sine wave tone. I understand how to generate the USB, but what does "minus the signal" mean when you go to generate the LSB?

And if it doesn't get clipped, what does it sound like then? What you have described is a form of double-sideband-suppressed-carrier, but you don't mention reinjecting the carrier at ther receive end.

Jim

t's

ut

Imagine a single side band with suppressed carrier. Take the upper side band case. All of the frequencies in the audio input are shifted to be above the carrier.

Now do the same for the lower side band but flip the polarity of the audio input. You now have the mirror image to the upper side band but the phases of all of the components are inverted.

Now add these two sidebands together and put in a carrier. You now have all of the same frequency components as a normal AM signal but if you look at it on a scope, it doesn't look like a normal AM signal.

This gets sent through the channel ( or radio waves)

Consider the 1KHz modulation case. You can quickly cobble the signal together in LTSpice or a spreadsheet to see what it looks like. The envelope will have a 2KHz modulation on it and if you look at the zero crossings you will see that they wobble back and forth in time at a

1KHz rate. This is because both sidebands hit their positive peaks together at the zero crossings of the carrier instead of at the peaks of the carrier.

The modulation of the zero crossings means that even if this signal gets clipped, the information still will mostly get through. Taking the one KHz modulation case, the easiest way to think about what happens when it is clipped is to think of the clipping process as an AM modulation of the signal at 2KHz. If you put a complex waveform and a (2KHz + DC) into a multiplier, you get out each of the frequencies in the complex waveform and each modulation sideband:

Input =3D 99KHz, 100KHz, 101KHz * (DC + 2KHz)

Output =3D 97Khz, 99KHz, 101KHz 98KHz, 100KHz, 102KHz 99KHz, 101KHz, 103KHz

So we now have 97KHz, 98KHz, 99KHz, 100KHz, 101KHz, 102KHz, and

103KHz. When we decode the signal we have our 1KHz and its second and third harmonics.

Did you know that LTSpice can read and write *.WAV files? It seems to do one or the other just fine. When I try to make it do both, it seems to hang. This fact combine with my extreme laziness prevents me from producing a good spice schematic to demonstrate the process.

have

I'm sorry I didn't make clear that the carrier goes along on the trip. I hope the above makes it obvious.