The AM Forum

Would I be louder if I spent my 1500 W PEP on a carrier + one sideband instead of on a carrier and two sidebands?

Jon

Limit negative peaks to 90% and drive positive peaks beyond 100% !!
Keep it real DSB AM!!

It's a wash.  Unless the receiver is filtering out one sideband only, the receiver will see a 3dB decrease in recovered audio.

Moreover, by running only 2/3 of a complete AM signal you deprive us of the option of tuning to the side that has the least amount of interference.

So it would make sense when one side is being obliterated and you know it -- deliberate SSB QRM right on the net frequency, for example. Then I might as well send a carrier + the other sideband, full bore, right? 

My homebrew rig can't do this but my old CE 100V + linear can.

Jon

This happens quite frequently. Idiot slopbucket will try to qwerm Ashtabula Bill or somebody with a zero-beat signal, say, by uttering their catcalls on LSB. Just nailing the receiver down to 4Kc and tuning to the high side on the R390A brings it back crystal clear no pain.

Please keep your AM signal intact. Do not neuter a sideband.

I find this very handy when I'm working 20 meter AM or an AM crowded 40 meter band.

SDR software removes this kind of crud like cutting a worm out of an apple.
Last night a moron on 160 was 3 KHz away. Stock radio could hear him but sdr removed it like a zit.

Does 1500 W PEP SSB-AM equal 1500 W PEP DSB-AM?

Yes and no.  1500 Watts PEP DSB AM = 1500 Watts PEP SSB AM, but the SSB AM signal will have much audio distortion.  They are not truly equivalent.  But the peak-to-peak recovered audio is the same. 

There was another posting a few weeks ago on an acronym for sideband-
equivalent AM.  I don’t recall the exact term.  But it reminded me that in only one engineering textbook (which I believe I have in my library) have I ever seen a dissertation on the distortion in SSB-AM.  So far I can’t locate the paragraphs.

I did some empirical tests to show the distortion of ssb-am.

We usually do signal testing with single-tone sine wave modulation with a 1000 Hz. tone of some specified modulation level.  For most work, the level is usually restricted to less than 80 % modulation, because most r.f. signal generators can’t be trusted to do a good job of generating 100 % modulation.

If we are talking about generating a 1000 Hz. sine wave modulated AM signal, for example, via a “SSB AM” signal, we are saying we will have a carrier signal and a second signal of lesser r.f. level.  The frequency difference of these 2 r.f. signals will be 1 kHz., to generate the AM signal.  (In a ssb exciter, your 1000 Hz. audio is converted to a single r.f. signal along with the re-injected or unbalanced carrier signal 1000 Hz. away.)

To generate a 50 % AM signal with ssb-AM, the second signal is 6 dB lower in level than the “carrier” signal.  To generate a 100 % modulated AM signal, both signals must be equal.  Here the distortion is most apparent.  This signal is identical to the 2-tone signal generated by an ideal ssb exciter fed with 2 audio tones of matched level!  Some of you, at this point, may be able to visualize the classic 2-tone signal’s envelope used for testing linear amplifiers with a SSB exciter.  It is not identical to a 100% modulated DSB-AM signal!  There is a well-defined V notch in the envelope at the “zero envelope point” or “negative 100 % modulation” point called “baselining” by many AM’ers today.  A true sine wave AM signal envelope will have a rounded shape at this “negative 100 % modulation” point, identical to the positive peak envelope shape.

I used a HP606B and HP8640B r.f. generators with a coupler/splitter transformer to combine generator outputs.  The receiver is a Kenwood TS-430 in LSB mode for distortion reference only and AM mode with 6 kHz. bandwidth filter.  Coupler output to TS-430. 

I did some recovered speaker audio THD distortion measurements on the signals generated:
  a.  TS-430 in LSB mode – one carrier signal only, tuned for 1000 Hz. tone received, a healthy speaker level is needed for the HP 331A distortion meter.  I measured 0.7 % THD.  (A well-designed receiver could produce a much lower distortion level than this.   This measurement shows the basic receive audio distortion level the -430 can give.)
  b.  TS-430 in AM mode – one signal from HP8640B generator only , 50 % modulated DSB-AM signal; 0.7 % THD.
  c.  TS-430 in AM mode – two r.f. generator’s carriers adjusted for 50 % modulated ssb-am envelope; recovered audio is 10 % THD.
  d.  TS-430 in AM mode – two r.f. generator’s carriers adjusted for 100 % modulated ssb-am envelope; recovered audio is 14 % THD.

A E.E. math geek could calculate the exact theoretical audio THD with the SSB-AM signals at 50 % and 100 % modulation.  In general, as the modulation level is reduced from 100%, the distortion will drop. As the modulation level approaches 0 %, the THD will approach 0 % !

In practice I surmise that, transmitting ssb-am, without heavy speech clipping the average peak modulation is below 50 % and the resulting distortion level is not noticed by most people to be much worse than many other AM signals.

Below are 3 photos of the ssb-am signal and DSB –AM signal envelopes:
1. 75 % DSB-AM
2. 75 % SSB-AM
3.  100 % SSB-AM

Tom, was that the "exhalted carrier" AM discussion?

I'm trying to pull up the thread too. At first I thought it was the type the Drakes produced, but that was controlled carrier, where the audio level swung the carrier up and down.

Was it this one:

http://amfone.net/Amforum/index.php?topic=6919.0

There was one other relevent Technical Forum topic, Feb. 1st.  This is one I was trying to think of:

"What is AME???

http://amfone.net/Amforum/index.php?topic=6907.0

- - - - - -

The maximum % of modulation achievable does depend on the circuit implementation.  Theoretical limit is 100 % as I said, but if the carrier comes from the unbalanced balanced modulator, the maximum mod. level will probably be 71 % as John says.  If the carrier is re-injected by a separate path, as in the FT-101EE (IIRC), 100 % can be obtained, like in my photo.

It can be demonstrated grapically with vectors.  SSB with carrier produces "quatrature distortion", a combination of amplitude and phase modulation.  With DSB AM, the quadrature distortion of USB cancels the quadrature distortion of LSB, to produce pure distortion-free AM, with no phase modulation component.  The only way to make a SSB with carrier signal reasonably distortion free is to reduce the percentage of modulation to a low value, at which point the quadrature distortion becomes negligible. 

The same thing happens when you receive SSB with a diode detector.  You have to turn the rf gain down to clear up the distortion.  Think of the BFO as the "carrier" and the SSB signal as the sideband.  If the rf gain is turned up, the sideband signal is strong, in other words, the percentage of modulation is high, and you hear lots of distortion.  Turn down the rf gain, reducing the "percentage of modulation" of the BFO, and the distortion goes away.

The only thing SSB with carrier is good for is for the carrier to produce a reference signal for the BFO  to lock onto, or for emergency reception using a receiver without a bfo.  It is nothing more than SSB with poor carrier suppression.  It is not AM. 

Buggar! I guess I am going to have to find a way to keep my Collins hF-80 system from knocking off one sideband. I got the exciter going today bypassed the Collins 2.75 KHz filter with a .1uF cap. Sure sounded better. My Racal though, was picking up both USB and LSB. What's up with that? Maybe it doesn't knock out one SB. Manual says it does.  Mmmmm.

Paul
VE7KHz

Paul, I think you bypassed the sideband filter that passes the one sideband and rejects the other.  If so, you solved the problem!  The DSB with adjustable carrier that you get from the balanced modulator is perfectly good, and it can produce perfect AM.  Just tune the amplifiers right, and don't clip the peaks.

Um, but you should use compression or peak limiting to keep from overmodulating.  Overmodulation from a balanced modulator is worse sounding than overmodulation from a regular plate modulated rig, and it is very easy to do.  In fact, it is difficult not to overmodulate any transmitter that has extra modulation capability.  However, overmodulation will not cause splatter in that setup, as long as the peaks are not flattening.

Hi Bacon,

Thanks! Excellent info. i am going to review the manual today to find out what I have done. I have strapped the jumpers like the manual says, which should give 16KHz wide, no SSB filter signal, but it's not, which is why I bypassed the filter with the cap.

Will definately be using compression ahead of the exciter. Will be firing the PA today so may just be on the air tonight with a new rig.

Cheers

Paul
VE7KHz

One other thing to be careful about is to make sure the reinserted carrier is exactly in phase with the original one.  Any phase error will distort the AM signal.  In fact, if the carrier is shifted exactly 90 degrees, the result will be pure phase modulation, with zero amplitude variation!

As I understand it, Maj.Armstrong's prototype FM transmitter consisted of a balanced modulator with reinserted carrier in quadrature (shifted 90º) with the original.  The resulting phase modulated signal was then converted to true FM by applying the appropriate pre-emphasis curve to the audio input to the transmitter.

I once modified a Drake TR-7 to bypass the sideband filter so that the AM output was true DSB.  That was a problem with that rig; the phase of the reinserted carrier bore no relationship to that of the original suppressed carrier.  I found a  small interstage rf transformer in the rf carrier amplifier chain.  I observed that adjusting the trimmer that resonated the transformer shifted the phase as you tuned through the resonant peak.  So, using a scope, I carefully adjusted the transformer for  the  proper AM waveform.  Amazingly, just as the reinserted carrier was adjusted close to the proper phase, it sort of locked in, and did not show any tendency to drift, kind of the same way a regenerative receiver locks onto an AM carrier with just the right amount of feedback when it is tuned zero-beat.  The AM output from the rig suddenly sounded 1000% better and the result was much more punch than the stock bogus SSB-with-carrier pseudo-AM.

Yes, that is the Armstrong method of Frequency Modulation.  The problem was that you could not get more than about 0.5 modulation index from this quadrature technique, which meant 50 Hz audio would produce 25 Hz deviation from the modulator, and you would have to multiply by 3,000 to get 75 KHz deviation.  This forced Armstrong to start at very low frequencies, and then the selectivity of the low frequency stages fouled up the FM sidecurrents at high audio frequencies.  This is why he developed block conversion and amplification to repeat an FM signal - the modulator distortion was too high to run through it multiple times, plus the complexity of such huge multiplier stages was greater than that of a block conversion.

Good point on inserted carrier phase.  If it gets shifted compared to the DSB, you get some mix of PM and AM, and envelope modulation will be reduced and distorted, depending on the degree of phase shift, and the negative peaks will not be able to reach 100% negative.  Higher audio levels will cause a doubling back of the negative peaks in the positive direction in the signal envelope, which will sound terrible on an envelope detector, but still 100% negative will not be reached.  Not a good way to control overmodulation!

Didn't he use something in the vicinity of 40 mHz (40,000 kHz) for his first experimental broadcasts?  I think that's why the old FM band was something like 42-50 mHz. Dividing 40 mHz by 3000,  means he had to start out in the vicinity of 13 kHz.  That wouldn't accomodate 2nd order sidebands above about 6500~.

The original DSB signal could start out higher, say a more manageable frequency like 100 kHz.  Multiply that by 3000,  up to 300 mHz, then heterodyne it with a 260 mHz  local oscillator to produce the 40 mHz PM signal.  Then you would have to apply de-emphasis to the audio for uniform FM deviation.

Yes, that would work.  I'm not sure why Armstrong didn't do that.  It may have been a patent issue.

I believe that the 13 KHz issue was exactly the problem.  Actually I think Armstrong was using a little less than a 3000 multiplier, and a little more than 0.5 modulation index (at the lowest frequencies), and his modulator was working around 16 or 17 KHz. Second order sidecurrents at the modulated generator frequency were pretty much irrelevant above a few hundred Hz, because the necessary modulation index dropped off with increasing audio frequency.  At 5000 Hz, 25 Hz deviation would have been a modulation index of only 0.005, and the Armstrong method did not produce any second order sidecurrents at the generator frequency anyway, because it was just DSB with 90 degree shifted inserted carrier.  This limitation is part of the reason you don't want to exceed about 0.5 modulation index at the generator frequency with this technique.  It caused distortion.

I'm guessing that he just barely squeaked by with 50-15,000 response.  Also, maybe he specified distortion at 80% mod or something.

Engineers wanted direct, high-stability crystal frequency control, which made PM to FM conversion the thing to do, one way or the other.  With crystals controlling the conversions, the 100K - 300 MHz - 40 MHz idea should have been OK.

Oops, darn, I hit quote instead of modify again.  Disregard.

Interesting topic  here. Poor mans AM one sideband with carrier. Yep thats how we do it on the other 3 WBCQ frequencies 5110,9330, and 18.910.  Slop bucket broadcasting is permitted under FeeCCee law . but must follow ITU standard which is the carrier must be at -6DB from the peak modulation value. The law is 50 Kw minimum power out be it carrier power for full DSB am or 50Kw PEP SSB. The 50Kw PEP SSB is much more cost effective than a full power AM transmitter to operate. There is a degree of distortion when recieved on a diode detector. The transmitting equipment is using old TMC exciters that have 8Khz filters. the overall audio reponse is much better than typical slopbucket.but the low end suffers a bit. The audio processing equipment is feeding the balanced modulators directly . For a while a phasing type generator using DSP audio phase shifting tecniques designed by Pete McNulty WA1SOV was in use on 9330. The audio quality was stunning .We hope to place it back into service soon. The only way to recieve poor mans AM is use of a sync detector . This is the other end of the equasion. A diode detector falls apart with any great degree of modulation. Most listeners are probably using inexpensive portable radios without sync detectors .Most of the programming on the frequencies mentioned are bible beaters. They do not seem to care about audio quality. Just as long as their "word" gets out. Such is life in another radio world in the HF spectrum .De Tim WA1HnyLR

sdr decodes it ok fine also

Carrier 6dB below peak means 100% positive modulation maximum, so 50KW PEP would be 12.5KW carrier out.  I see how that's cheaper to operate than AM with 50KW carrier out, even with the efficiency of a linear amp.  That's an interesting power regulation.

cool

Exactly!  I never listen to both sidebands with my Flex.  But if someone checks in and I'm listening to the other sidebend I won't hear him.  Additionally, if there is a carrier on one side, I will not be able to do exactly what VJB says in this quote

Use both sidebands, PLEASE!

Nice 5 year old thread.

Well then I shall renew the request -- Please keep your AM signal intact. Do not neuter a sideband.

5 year old thread - nontheless it's a topic banging around from time to time.  My EICO 753 does that with the AM mode.  There is no reason why a SSB with carrier left intact shouldn't sound OK especially if a 6 KC filter method is used. Just might be a nice thing if the carrier osc. could be moved from one side of the filter to the other - to accomodate the above mentioned commentary about slopbuckets and carriers.

The phasing method wouldn't work due to phase shift problems over the modulating AF frequency spectrum.   


Al

You'd be better served by starting a new topic rather than necroing this old one.  You'll probably generate more interest.

I don't know the answer to your question, but I suspect not.  You can always bypass the sideband filter.

Ive received excellent reports from a TS-950SD with the audio switches set for 3300Hz and have switched sidebands when needed to either get away from or annoy a slopbucket. With help from the LK-500 with the external PAC-5 transformer and the fan on high it seems to get thru just fine with just a little help from the rigs compressor.

There is also a mod that allows it to transmit thru the 6KHz AM filter but I havent done that as its a bit complicated to still use the 2.7 filter on SSB.

Hopefully the overhauled 100V will be on line soon and I can test AM on LSB, USB, and DSB.

Carl

The WBCQ setup is interesting.   The solid state 50 KW rigs like the DX50 are supposed to be around 90% efficient. 

I think at one time Ten Tec made a fake AM rig, the Paragon maybe, back in the 1980s.  I occasionally hear carrier+one sideband on the air, most recently from a KWS-1.  Sounded like crap.

                                 WHY ONE SIDEBAND PLUS CARRIER IS NOT THE SAME THING AS AM

One sideband + carrier is not AM, in the sense of the mode that is the reason for the existence of this bulletin board.  One sideband + carrier is just what it says.  Slopbucket with poor carrier suppression. It is AM ony in the strictest sense that SSB is merely a special form of AM. The specific form of AM portended by this BB is full carrier double sideband amplitude modulation. We just call it "AM" for short.

Due to inherent fundamental mechanics of the modulation and demodulation process, one sideband + carrier cannot be received on an envelope type detector such as the diode detector without severe distortion, unless the modulation percentage is kept very low - something on the order of 30% maximum.

The only reason SSB + carrier even vaguely resembles AM when copied with an envelope detector, is that the peak-to-average ratio of the human voice waveform is extremely high. While voice peaks drive the detector into unintelligible gibberish, the average modulation level remains at a low percentage of modulation, so that the signal may come through reasonably intelligibly, but it still sounds like crap and the voice peaks are distorted. That is the same reason why an analogue VU meter dwells in the 20-30% region most of the time while it spikes up into the red zone on louder voice peaks.

We can make another analogy by considering the detection process in a receiver.  In the early days of SSB, most amateur receivers didn't have a product detector, so the majority of hams received SSB with a diode detector and BFO, in exactly the same manner as they had copied CW ever since CW had replaced spark. But to clarify the SSB signal, not only did the receiver have to be tuned exactly on frequency, the rf gain had to be turned way back with the audio gain set at or close to maximum. The BFO provided the reinserted carrier, and the output from the i.f. amplifier provided the modulation.  If the rf gain was set to maximum as with normal AM reception, the i.f. output to the detector was too great, the diode detector was overloaded, and all that came out of the speaker was unintelligible distortion. In other words, the percentage of modulation was too high.  By cutting back the rf gain control, the carrier (BFO) remained  the same, but the percentage of modulation (from the receiver's final i.f. stage) was reduced to a point where the signal could be detected with little or no audible distortion. The diode detector still demodulated the signal in the exact same way that it demodulates a full carrier AM signal. In this case, the "carrier" is the BFO and the modulation is provided by the received SSB signal.  But since there is only one sideband, the modulation percentage must be kept low by retarding the rf gain setting on the receiver to eliminate the severe distortion.

This distortion that occurs when SSB + carrier is received with an envelope detector is called "quadrature distortion" and can be fully demonstrated vectorially. This is fairly well explained in Single Sideband for the Radio Amateur, the popular ARRL publication that first appeared in the 1950s.



                         THE PROPER TRANSMISSION AND RECEPTION OF ONE SIDEBAND PLUS CARRIER

As explained above, one sideband plus carrier cannot be properly received with a normal envelope type AM detector.  But it can be satisfactorily received by treating it as a regular SSB signal, or distortion free, by using a properly locked synchronous detector, which is nothing more than a product detector with an additional phase-locked loop circuit to lock the BFO exactly in synch with the received AM carrier. Unlike the envelope detector, the product detector can handle the full output from the receiver's i.f. amplifier, so the rf gain does not have to be reduced for distortion-free reception.  With a synch detector, SSB + carrier can be made to sound exactly like regular DSB AM.  In fact, using my Sherwood synch detector, I have had another station switch between AM and SSB + carrier, and could not tell any difference in the sound. When the synch detector is used for reception, the transmitted carrier serves only as a pilot carrier to give the BFO something to lock onto, and does not actually demodulate the sideband directly.

But, if SSB + carrier is transmitted with the intention of receiving the signal with a synchronous detector, it is very wasteful of power to transmit a full carrier.  Only enough carrier is needed to provide a reliable lock with the PLL (phase-locked loop). With a good synchronous detector, the carrier can be reduced to at least the vicinity 20 dB below the p.e.p. of the sideband, and still get a reliable lock. By using a pilot carrier reduced to this level in combination with the synch detector, the annoying frequency error that usually occurs with amateur SSB and gives it the "Donald Duck" sound, is eliminated, but the reduced level of transmitted carrier would not waste an excessive amount of power nor cause the linear amplifier to overheat or run at low average efficiency.  Years ago, before the advent of satellites, AT&T's transatlantic and ship-to-shore radiotelephone circuits could be monitored on an ordinary short wave receiver, as they transmitted SSB with a pilot carrier.  

This is where the ESSB guys are missing the boat; if they transmitted a low-level pilot carrier and used synch detectors for reception, their signals could be made to sound very close if not identical to regular full carrier double sideband AM, while maintaining nearly the same power efficiency as ordinary fully-suppressed carrier SSB.

yes and that works well provided the intended listeners can move their bandwidth to cut out the QRM or narrow it to receive only your transmitted carrier+SB.  And thank you Don, you are 100% right and that was a very interesting illustration of why it sounds distorted.

SSB+carrier can't be received with a Costas loop sync detector either, which depends upon the phase relationships of both sidebands to maintain phase lock.   The Costas loop doesn't require a carrier at all, just both sidebands.  However I'm not aware of any Costas sync detectors in any HF receivers with the possible exception of the Sony ICF-2010 which as I recall uses a AM stereo demodulator chip for its sync detector.  A Costas detector could be implemented in a SDR also but again I'm not sure if this has been done in the popular SDR hammy software.

 ;D X2 !

Then why do so many report many riceboxes on AM as having an excellent sounding signal on a BA receiver? Ive no complaints either listening to them on a BA or a ricebox in AM mode.


OK, all you guys with controlled carrier modulation get the hell off the forum, His Lordship has spoken ::)

All the modern riceboxes that I know of to-day run both sidebands in AM mode.  Many of those rigs can be made to sound just as good as plate modulated AM, and one could be fooled into believing they were running a converted broadcast transmitter. Some of the early US-made rigs like the KWS-1 and the Drake TR-7, and some first primitive transceivers had an "AM" mode that ran SSB+carrier.



 "Controlled carrier modulation" is still full carrier double sideband AM, whose carrier varies in amplitude at a syllabic rate in step with the voice. Controlled carrier was first introduced in the 1930s, using normal high level plate modulation. One way this was accomplished was using a special saturable reactor to control the primary voltage to the plate transformer in the final amplifier power supply to follow the variations of the modulator plate current.  Another method was to wire the class B modulator, class C final and power supply in series, so that with no modulation the modulator just drew static plate current, and thus limited the DC input to the final.  On modulation peaks, the modulator drew full current, which also passed through the rf final in the series circuit, bringing up the DC input to the final up to full value.

Rigs  like the Drake TR- and T4- series, Heath DX-40 and DX-60 and Knightkit T-60 and T-150 used controlled carrier screen modulation, all full carrier double sideband AM just the same.

I always try to control my carrier.  Whether it's changing with the amount of modulation or not, I sure like to set it to where I like it.

Wow. Having trouble believing my eyes. Have tolerance your hinus.The AM'ers as a group are better combined as apposed to be broken into fractured and bickering groups.

Jim
WD5JKO

Relax Jim, that was just some Yankee sarcasm ::)  Ive no idea where that long post of Dons originated as it looks as if it was copied from somewhere.

However I still maintain that the various phasing rigs of old that allowed switching to DSB, LSB, and USB AM sounded good. I cant remember trying AM on a NCX-3 or NCX-5 when I had them as loaners. At home from the late 50's I went from various CE 10A to 20A, Phasemaster, HT-37, and then a long term 100V to C Line, to TS-930/940/950SD.

There are a couple of CC rigs here that get used with the Alpha 76PA and I flog a T-150A at the summer cottage.

Carl

You could build a detector which can utilize both sidebands, for instance synchronous double-sideband product detection either with hardware or in software definition. Since the upper and lower sidebands are mirror images above and below zero ( the carrier frequency), they can be combined in phase (not in quadrature as with SSB demod in the Phasing or Weaver method) to reinforce the signal.


http://www.arrl.org/files/file/Technology/tis/info/pdf/9709qex003.pdf

Mike WU2D

Some interesting reading on this thread. In 1964 I put together a 6146 class C transmitter with a homebrew modulator using a pair of 807s. Our simple Lafayette HE-30s were AM mode radios with CW/SSB capability done by dumping a BFO into the diode detector. You had to turn off the AVC because the BFO drove the AVC so that the S meter read around 30 over nine. Of course, with the AVC off, the RF input would over drive the later IF stages into full limiting/distortion. You had to turn down the RF Gain so that the RF/Mixer/IF stages had a chance of keeping the input linear. The Af gain was turned way up to hear the signal out of the detector.

Perhaps SSB plus Carrier should be thought of as an alternative to DSB AM when conditions warrant it.  Many people have made mention of tuning in one of the two sidebands when interference on the other calls for it. The bandwidth is often either shifted or cut in half under these conditions. So, under certain conditions, SSB AM might make sense to fight through the crud. IF you really need to get through, just switch over to SSB and run full bore anyway. So, what's the point. IF you are looking for maximum detected audio and want to rely on the carrier for detection, The carrier at 375W and peak SSB signal at 375 W will get you at 1500W PEP. You will get the same recovered audio with DSB AM if you detect both sidebands since they are each at half the voltage level of the SSB AM peak sideband voltage. The difference is that you use half the bandwidth. So, if you throw away one of the DSB AM sidebands, you are losing 6 dB of recovered audio.

The deal is the detection process. The diode detector has a logarithmic transfer function. Many mathematical functions can be represented by a series summation of other simpler functions, much like a square wave can be represented by a weighted sum of odd harmonics. For the log function, the principle term in the sum is a square term. hence the term "Square law detector."  A problem with this detector for DSB AM is that the square term not only multiplies the carrier times each sideband to down convert to baseband, it also multiplies the sidebands by  each other to produce second harmonic distortion. Other, higher order terms produce some additional distortion, but most products are at RF frequencies, and the coefficients of these terms drop off in value quickly. 

IF distortion reduction is your thing, and you want self detecting of DSB AM, then something like the precision detector cleans it up by not being a square law detector. It operates more like a self driven synchro detector. The diode type detector works fairly well with SSB AM. You probably have done it many times without realizing it. Every time you tune off to the side of a DSB AM signal, to get away from an interfering SSB station, you are eliminating one of the sidebands and detecting SSB AM with your own simple diode detector. Any distortion you hear from the desired signal detection process would be the higher order terms coming into play and/or level issues between the carrier down on the slope of the passband filter or other issues. IF you have an SDR setup, try it with the SDR receiver set to AM, tune in a DSB AM signal normally, and then pull in one side of the filter to the carrier. Either way, you have made the DSB AM signal into an SSB AM signal.

My disappointment in this whole thing is that the precision detector, perfect of DSB AM, is not good for SSB AM. An SSB AM signal, say a 1 KHz tone signal plus a carrier, would look just like a two tone SSB signal. and a precision detector (rectifier) would have a harmonically distorted waveform. It almost works since most of the harmonics of voice signals of 1.5KHz and up fall above the receiver audio stage upper cut off frequency.

Bottom line, consider SSB AM as just another tool to fight interference, especially if you push it to full bore on both the carrier and sideband. When the conditions are clear, run full DSB AM and enjoy the copy!

But if you are going to run SSB, what's the point of running all that carrier?  Just have the other op turn on his BFO, or better still, if he has the capability, switch to synchronous detection. Use just enough carrier to give the synch detector something to lock onto, or to use as a reference for zero-beating the frequency.  With a diode detector and BFO turned off, any SSB + carrier modulation to a depth greater than about 30% is wasted anyway, since the quadrature distortion renders it unusable.

A diode detector is not a square law detector, because the logarithmic characteristic occurs only at very low levels, right at the threshold of response.  Usually, the i.f. amplifier delivers more than enough signal to the diode detector to make it function approximately as a linear detector.

The square law detector is interesting.  The early AM broadcast sets from the 20s and early 30s used triodes wired as grid leak detectors, because even though this detector had a square-law characteristic, it offered much greater sensitivity than would a simple diode detector.  Unfortunately, this square-law characteristic resulted in considerable audio distortion. Most of those old radios had a crackly sound even from strong local stations.  Since most people had never even heard a radio play before, this distortion was accepted by the public as something normal for the medium.  But it so happens that the square-law characteristic gives the audio output a distorted waveform that is the exact opposite of the distorted waveform that results from envelope detection of SSB + carrier.  Theoretically, a true square-law detector at the receiver would give perfect reproduction from a SSB + carrier signal modulated up to 100%, but the DSB AM signal would appear distorted!

I'm not sure if anyone has ever run any experiments with this, but it would be an interesting investigation. Maybe dig out some of the square-law detector circuits from the early days of radio, or use modern solid state technology, and build a true square-law detector.  I suspect the threshold of detection would have to be adjusted to closely match the level of the received signal, in order to get the two distortions to exactly cancel. It would be interesting to see how a SSB+carrier signal received with a square-law detector would perform in the presence of noise and interference, compared to DSB AM received with a linear envelope detector, considering of course, the total sideband power transmitted in each case.

PS: The reason you can tune to one side of a DSB AM signal, using envelope detection and no BFO but narrowing up the receiver to receive only one sideband, and not have severe distortion, is that removing the other sideband lowers the modulation percentage of the signal received through the narrow filter enough to stay below the threshold of severe distortion.

Yes, run both sidebands so those who want to only hear half the signal can.

Sounds kinda funny, dumping on people who run "SSB + carrier", because you want the ability to choose one sideband + carrier to listen to...

Actually, I have to admit there is some merit to that.  With DSB (with or without carrier) the listener can choose which sideband to receive, depending on the QRM that pops up while the station is making a transmission.  Somewhat a form of diversity reception, particularly useful during long "old buzzard" transmissions. SSB + carrier would be useless in that case, since there is no way the transmitting station could be aware of which sideband is suffering from the most interference, not being able to monitor the frequency while transmitting. If that one sideband gets obliterated by interference, you are up the creek without a paddle.

The Sherwood synchronous detector has on "offset tuning" mode that allows the listener to shift to one sideband, by moving the receiver to one side of the passband without losing BFO lock.  This allows it to work just like it would with envelope detection, but the product detector receives the signal more effectively without the distortion.

I don't know if this was answered on AME. It was a mode selectable on a Motorola ALE type radio. Very user un-friendly. It was not designed to be a HAM radio. It's specific purpose was for the HF emergency systems in use throughout the USA.
The AME was Amplitude Modulation Equivalent. A compatible sideband transmission.

http://en.wikipedia.org/wiki/Compatible_sideband_transmission

The transmitters at WBCQ that use the reduced carrier and are modulating  USB. The audio is exceptional for SSB, but is more susceptible to selective fading distortion than a full carrier. The selective fading effect is eliminated with the use of a sync detector. On my R390A the audio is lower.


This is what Allan / Timtron use at WBCQ on  5.110, 9.330, etc.................7.415 is full carrier 50kw....I can't believe they get that power from single phase 220 and three other transmitters running 12kw ea with the reduced carrier. And I think he has a commercial AM (5KW) older plate modulated TX on the same pole.