AM radio receiver - design

s t s

Some loops are built over a ground plane, other use a transformer and float above the ground.

I'm pretty sure I'd hear a 3db reduction in noise with synchro. I'm not sure why you think the snr would be 3db better. If it is because you are listening to just one side of the AM signal, you need to remember that you are getting half the energy as well. Thus the narrow bandwidth doesn't buy you much unless you managed to avoid the QRM/N. Synchro helps a bit with fading since that phenomena is due to the mixing of signals (direct and reflected). [Narrow the BW, less room for constructive and destructive interference.]

It's really hard to design a decent synchronous AM receiver. Most of the built-in synch demods are real crap. Even Drake has to rev their synch a few times. Sherwood sells an external synchro box because everyone else does a piss poor job. Probably the R8b and AR7030 are the only receivers out there with passable synchronous demod.

oise

for

I

st

I've seen Lankford's work. A lot of it is dubious at best. I recall he had a "paper" where adding a lowpass filter in the audio path would reduce fading.

The phasing schemes work if the QRM/N is local. I have a ANC-4 for that purpose. MFJ makes one as well. Still, nothing is a good as using a Wellbrook. There is something about the "large aperture" of the Wellbrook that reduces the local noise.

Having hacked with the older ALA100 (there are a few versions now), I find a 4x4 loop of copper pipe (easy to build) does the job for AM BCB. If you want a deep null, you need to give the antenna space. I've done handheld direction finding with the ALA100 and a 2x2 loop. For most urban areas, about 4x6 is the practical limit. I find the attenuator turning on for some local stations with that much aperture. In the boonies, anything goes I suppose. I have use the ALA100 with an

8ft on a side loop. I could pull in NDBs from 1000 miles away in the daylight, but only in remote areas.

h.

is

,

ent

n

ess

p

You'd have to do an A/B test quickly. 3dB isn't huge.

Go back and re-read what I said. Look carefully at the noise it removes. It takes out 1/2 the noise. Since we are talking about random stuff the difference is 3dB.

No, it is because the sync demod rejects half the noise in the combined sidebands. The is double sideband with a carrier. It is standard AM radio not SSB. The 3dB improvement happens only if you have both side bands and the carrier. Only if you have both sidebands can you have a signal like:

Y =3D f(t) * cos(wt) + sin(wt)

That equation can't exist in a SSB world.

Please look through the papers and on the URL I postedm and post any dubious material you find. Inquiring minds want to know!

Phasing schemes are also helpful to null out unwanted stations on the same frequency Mike

What sort of interference? You might be better off hunting for the source and replacing the offending components (if they belong to you) or having the source fixed (if the equipment belongs to someone else).

You might be expending a lot of time and money better spent replacing a bad fluorescent ballast or something.

--
Paul Hovnanian     mailto:Paul@Hovnanian.com
------------------------------------------------------------------
"Si hoc legere scis nimium eruditionis habes."
(If you can read this, you\'re overeducated.)

ise

or

th

ch

same

=A0 =A0 =A0 =A0Mike

That silly paper about the elliptic filter reducing fading comes to mind. As far as I know, Lankford writes those papers in the format of a technical journal, but they are neither published or peer reviewed.

dth.

at is

ks,

nd

onent

in

nd

on

r

less

e

oop

First, I need to dig up a communications book since I don't know if you noise model is legit. But look at this this way. You are receiving one side band. Half the signal, half the noise. Where did you gain 3db?

y(t) =3D (1+f(t))*cos(wt)) + n(t) where n(t) is the noise seems like a more reasonably model. .

First, I need to dig up a communications book since I don't know if you noise model is legit. But look at this this way. You are receiving one side band. Half the signal, half the noise. Where did you gain 3db?

y(t) = (1+f(t))*cos(wt)) + n(t) where n(t) is the noise seems like a more reasonably model.

There's what looks to be a reasonable account of synchronous detection of AM at

Using a quadrature regenerated carrier would resolve one sideband as noted above, so up to half the signal power would be unused and such an improvement in SNR seems unlikely. Another use of quadrature in this way is to detect NBFM.

Unfortunately, there appears to be material on numerous other web sites that claims SNR improvement, for an in-phase regenerated carrier, because it's claimed that the detector is insensitive to the phase component of the noise, which is probably not the full story.

Chris

e

width.

that is

orks,

and

mponent

lt in

and

tion

for

nd less

a

the

th

es

loop

s

Trust me it is. With noise, you can say that any function is part of it and then say that the rest of it is everything but that function. This is a nice thing about talking about random stuff.

I am talking about a normal AM signal with two side bands and a carrier. You are changing to talking about something else.

[....]

e

s,

in

n

for

nd

a

the

th

es

AM

ated

al

hat

The OP is talking about AM band radio. This is not single side band stuff. There is a huge difference between the two in terms of sync demodulation's effect.

In single sideband, the phase of the carrier is modulated.

The OP is talking about AM band radio. This is not single side band stuff. There is a huge difference between the two in terms of sync demodulation's effect.

In single sideband, the phase of the carrier is modulated.

- - -

The most common meaning of 'single sideband' is SSBSC ... suppressed carrier, so there is no 'carrier' the phase of which to be modulated. SSBSC is simply the audio band, or a spectrally-inverted representation of the audio band, translated up to a band of radio frequencies.

One of the well-established benefits of synchronous detection of AM (i.e. DSB with carrier) is that it can be used to resolve the signal content of each sideband independently, whereas an envelope detector relies on coherence between the sidebands and the presence of a carrier of adequate strength - if either of those conditions is not met, the result of envelope detection will contain harmonic distortion. Consequently, synchronous detection can help when the desired AM signal is subject to frequency-selective fading ... as explained in the document to which I provided a link. However, when both those conditions _are_ met a simple (no quadrature) synchronous detector provides the same SNR as an envelope detector because it does the same thing: it translates the pair of sidebands down to baseband, symmetrically about 0 Hz (all the mod signal and all the noise). The quadrature version translates one of the sidebands down to baseband (half the mod signal and half the noise) giving the same SNR.

If what you're talking about is different from either of these cases then that isn't clear from what I've read so far.

Chris

.

ome

e

er

t

LL

ion

s

th

e

he

nd

of

ted

's

Yes but we keep referring to the carrier so the meaning here would be for the single side band with carrier case. The OP is in fact talking about standard AM which is double side band with carrier.

[...]

Yes but that has nothing to do with my point. I explained how a PLL based detection reduces the noise in a normal AM radio signal. The OP wants to make a very good radio for the normal AM band.

pe

It also assumes that the SNR of the whole situation is large (unless this is what you mean by "adequate strength"). When the SNR is poor, the assumption that the carrier multiplying by the side bands will be almost all of what ends up as base band signal quits working.

e (no

nds

e

I think I was completely clear in the first posting, but that you jumped to assuming something because you were not thinking about the OP's case. The OP is making a radio for the standard AM band. Keep that point foremost in your mind as you reread what I wrote back there and read this explanation below. They are two different cuts at trying to explain the same ideas.

Disregarding signal strength, the signal in a normal AM radio is ideally:

Y =3D f(t) * sin(wt) + sin(wt) (1)

Where:

f(t) is the program material that is centered around DC. It contains frequencies from 20Hz to roughly 7KHz. The peak amplitude is always less than one.

sin(wt) is the carrier. It appears twice because this could also have been written as:

Y =3D (1 + f(t)) * sin(wt)

The version (1) of the equation is the better because it breaks the carrier and the sidebands apart and makes it far more clear what the signal is:

What you actually receive will be:

Y =3D fn1(t)*cos(wt) + fn2(t)*sin(wt) + f(t)*sin(wt) + sin(wt)

The fn1 and fn2 are two parts of the noise. Any function (noise or signal) in a band that does not include DC can be broken into two parts in this way. Since you know single side band methods, I won't bother to explain how. Just remember to remind yourself that this is not SSB radio before reading the next bit.

Now that we have what the receiver actually has as a signal. The next step in the argument is to assume that we have a PLL that is locked onto the carrier and that its bandwidth is much less than 20Hz so that as far as we are concerned, it is a constant frequency sine wave. We can multiply each term in the equation by sin(wt)

fn1(t)*cos(wt) * sin(wt) no base band result

fn2(t)*sin(wt) * sin(wt) base band is fn2(t)

f(t)*sin(wt) * sin(wt) base band is f(t)

sin(wt) * sin(wt) DC result

Notice that only the fn2(t) appears in the result.

Now for the case where we don't use a PLL. To keep it simple I will show only the second power term of the diodes function. The diode curve is actually a long series but I am too lazy to write an infinite number of terms. I guess it would be valid to say that I am using an RMS based detector. I will also not break down the fn1() and fn2() into series. Just remember they are. I also won't show the signs.

fn1(t)*cos(wt) * fn1(t)*cos(wt) fn1()^2 noise fn2(t)*sin(wt) * fn1(t)*cos(wt) no base band result f(t)*sin(wt) * fn1(t)*cos(wt) no base band result sin(wt) * fn1(t)*cos(wt) fn1() noise

fn1(t)*cos(wt) * fn2(t)*sin(wt) no base band result fn2(t)*sin(wt) * fn2(t)*sin(wt) fn2()^2 noise f(t)*sin(wt) * fn2(t)*sin(wt) fn2()*f() noise sin(wt) * fn2(t)*sin(wt) fn2() noise

fn1(t)*cos(wt) * f(t)*sin(wt) no base band result fn2(t)*sin(wt) * f(t)*sin(wt) fn2()*f(t) noise f(t)*sin(wt) * f(t)*sin(wt) Program material squared sin(wt) * f(t)sin(wt) Program material

fn1(t)*cos(wt) * sin(wt) no base band result fn2(t)*sin(wt) * sin(wt) fn2() noise f(t)*sin(wt) * sin(wt) Program material sin(wt) * sin(wt) DC

If you sum it all up, you will find that both fn1() and fn2() appear in the resulting function that goes into the squareroot if this is an RMS.

I think you are agreeing with me...I said the synch detector improves SNR only near threshold. When you are above threhold say 15 dB SNR, there is no SNR improvment. The SNR improvment occurs only near threshold, i.e. the diode envelope detector SUFFERs fom the threshold effect when the carirer componet becomes noisy, but the synch detector does not suffer from a threshold effect.

SNR RF SNR Audio SNR Audio Envelope Det Snync Det

15 dB 15 dB 15 dB 10 dB 9 dB 10 dB 3 dB 0 dB 3 dB 0 dB -6 dB 0 dB

In some cases this can be helpful..

Mark

If the below is what you mean by threshold then yes, but remember that it is a smooth curve of decreasing advantage.

That depends on how many digits you use. If you use enough digits, there will always be some advantage. The OP said that he had lots of noise.

This sync-demod problem sounds counterintuitive. Please explain why various manufacturers have such problems. Links to schematics would really help. This smacks something going on the i want to understand.

mind. As far as I know, Lankford writes those papers in the format of

You seem to have misread what he said, He didn't say "adding a lowpass filter in the audio path would reduce fading."

To quote a couple of lines from his article: "The distortion which one hears from strongly fading MW and SW signals manifests itself as high frequency sound akin to noise. So it would seem that an appropriate low pass audio filter should substantially improve the audio quality of strongly fading MW and SW signals."

"They (the eliptical filters) provide about as much reduction of fading distortion and reduction of flutter (including sub-audible hetrodyne) distortion as the very best AM synchronous detectors, such as the Drake R8B. At the same time (they) also reduce high frequency splatter."

Seems reasonable to me. Mike

Sync sounds good on paper. In reality it is hard to do. You radios that "growl" at you. If the lock isn't perfect, you get low frequency grunge. You also can get a slight frequency offset (translation).

Sherwood makes the best synchro:

Most high end radios have a 455KHz output for a pan adapter and/or external PLL demod.

Since I own a radio with synchronous detection, I can tell you in real life, you don't better SNR with sync. It helps with adjacent channel interference. It also helps with fading since the bandwidth of the signal is reduce. [Audio bandwidth is not reduced, just signal bandwidth). The best way to get rid of noise is the magnetic loop antenna.

You might want to go to rec.radio.shortwave. Lots of medium wave listeners there too. Pretty much you will get the same answer. There is some guy there who has multiple Wellbrook loops. He uses a phaser (MFJ I think) to use the 2nd loop to null transmitters on the same frequency. However, that doesn't sound like your issue.

In real life, you are probably better with a SqueezeBox or Roku and just streaming the AM station, assuming you have broadband internet and the desired station provides a stream.

To use a wellbrook, you need at least a moderately high end radio that has a 50 ohm input. Basically, we are talking about a new hobby. ;-) Most people who listen to medium wave or shortwave already own their gear prior to streaming, or they listen to signals that are not streamed (military, boats at sea, aircraft, numbers stations, etc.) In

2009, the Squeezebox or Roku is the way to go. The Squeezebox is very high quality. The Roku is a bit cheaper, but also not as good. I have the original Squeezebox. They added a boombox version, a high end version, and a cheaper version if you don't want a display. The Squeezebox lets you stream your own recordings (ripped CDs or purchased MP3) from a computer to the box. It has many radio stations preprogrammed, so you don't even need a computer unless you want to play your own tunes. Often you can get small discounts ($50 or so) if you order directly and can find the secret promotional code.

The Roku software isn't quite as good as what you get from slimdevices. There is a software version of the Squeezebox on sourceforge called softsqueeze. It emulated the function of the hardware using your PC. For me, it built confidence in how the final system would operate. That is, I set up my music on a linux PC, then use my windows PC to act like the squeezebox. When I decided to buy the hardware, getting it going was a snap. Just a matter of setting up the wifi.

Way short of any explanation.

Looked and listened, seems to audiofoolery, with price tags to match.

You have just assassinated your position.

Now you have drifted completely off topic.

I can think of a few likely problem areas. They can be fixed but they may give insight. There are two basic designs that I have come up with. In one the VCO is the local oscillator and the 455KHz is a reference making the whole thing a PLL. In the other there is a normal AM radio up to the detector and the PLL is added at that point.

(1) In the circuits I've mess about with, the reference oscillator or a VCO runs at twice the 455KHz number putting it at 910AM where it would perhaps prevent the good reception of one AM station.

(2) To make the needed stability of the 455KHz, you need a trimmed LC oscillator or a crystal. This costs a lot more than a 1N914.

(3) The servo gain crossover needs to be well below the audio but the locking range needs to be on the order of 1KHz.

On this last point, there are a few easy fixes:

If you make the loop filter of a PLL such that it oscillates when there is no feedback via the phase detector, it will sweep back and forth until it locks.

When not locked, you can widen the bandwidth.

When the filtered sync demod output goes negative, you can decrease the gain on the VCO filter.