ARRL

Login Help

Forum Home - Rules - Help - Login - Forgot Password
Members can access, post and reply to the forums below. Before you do, please first read the RULES.

Displacement Modulation

Feb 5th, 23:29

K0WUQ

Joined: Dec 3rd 2012, 11:13
Total Topics: 0
Total Posts: 0
In a former life on another forum, a great man once taught me, "Sometimes people are just ... wrong."

Looking at p. 58 in my 1966 Radio Amateur's Handbook, they introduce the idea of 'Modulation' with a Figure 2-67 that shows graphic representations (like oscilloscope traces) of audio, RF, a blend of the two and finally, Amplitude Modulation. Example C in the drawing takes the RF trace and bends its mathematical centerline to exactly match the audio. This means that the upper and lower envelopes of the RF also perfectly match, so that the RF amplitude is constant throughout. They do not give this pattern a name, so I call it Displacement Modulation, since the audio signal simply displaces the RF up or down at every instant in time.

In their discussion of this drawing (note in the following that KHz was then called Kc, Mhz was Mc), they point out the supposed physical impossibility of using the pattern in subfigure (C) for radio communication of the audio content. On Page 58, they say:

“Figs. 2-67 A and B show two such frequencies, and C shows the resultant. The amplitude of the 1-Mc. current is not affected by the presence of the 1-Kc. Current, but the axis is shifted back and forth [i.e. up and down on the drawing] at the 1-Kc. rate. An attempt to transmit such a combination as a radio wave would result in only the radiation of the 1-Mc. frequency, since the 1-Kc. frequency retains its identity as an audio frequency and will not radiate.”

This explanation is wonderfully clear, direct and authoritative and seems perfectly sensible. And it is just ... wrong.

I have experimented with this type of modulation as much as I could for the past 18 months or so, and have proven the following about it (my experimental methods are cheap and simple and can be duplicated by any amateur to verify or disprove what I claim here):
- This modulation method DOES propagate through space
- Adding this modulation method to a typical RF amplifier is inexpensive and simple
- Demodulation can be accomplished by at least two "primitive" types of receivers (solid-state kits readily available)
- Demodulated quality is at least as good as the best AM
- There is reason to believe that some modern receivers will not demodulate the signal properly - more experimentation req'd
- The modulation can be blocked with capacitance when desired - this is a strikingly strange property; it is NOT simple filtering

The Handbook DOES make some interesting claims for this method. These are important because they show great advantages in terms of bandwidth requirements:
- There are NO sidebands created (as in AM)
- There are NO harmonics created (as in FM)
- Bandwidth is identical to a CW wave, regardless of audio content

If there is any interest in experimenting with this, I will gladly go into more detail, maybe on the Homebrew forum or some such. If not, just forget I said anything. Obviously, this is NOT a method currently authorized for communication, and thus limited to "laboratory" type experimentation only.

Larry K0WUQ
Feb 8th, 12:12

K0WUQ

Joined: Dec 3rd 2012, 11:13
Total Topics: 0
Total Posts: 0
RECEIVERS

I think I should say a little more about the receivers used to demodulate this type of signal. Of course, it's hard for me to say "a little more"; as the late Lowell Thomas said, I am now at the age where everything I say reminds me of something.

Both receivers I used are Vectronics kit products - I first built their multi-band receiver re-branded as the MFJ-8100K; later I built their little 40 meter receiver, the VEC-1140K. In many ways, they are quite different, but both are compact devices and will run for hours and hours on a single 9-volt battery. Both kits are under $100 US each; the MFJ comes complete with a nice metal case, but you have to order the case as an extra for the VEC-1140, with the whole thing still coming in at $80 or so.

I believe (but have not proven) that these receivers properly demodulate the wave because of one common property: Although the circuits are quite different, BOTH receivers demodulate by immediately heterodyning the received signal with a tunable local oscillator. This is the feature that lets both units handle CW, AM and SSB without any 'mode switching' whatsoever. The MFJ-8100 is a modernized solid-state version of the venerable regenerative receiver developed in the 1920s. The VEC1140 is a more modern approach they call a Direct Conversion Receiver (all the details of this are hidden inside a single chip). They offer this receiver in distinct 20, 30, 40 and 80-meter versions, depending on the values of a small handful of passive components.

I'm quite sure there are many other receivers that would work. More experimentation required - ha.

Tuning is simplicity itself. For CW, you just tune onto the station and adjust the beat frequency whistle to a tone you like (or that makes it stand out the best from the background noise). For SSB, you tune onto the station and hear a voice like either Donald Duck or Darth Vader; you then adjust tuning until a quasi-human tone is obtained. For AM, the MFJ-8100 lets you lower the level of local oscillation so you can just tune through background noise until the audio comes through. On the VEC-1140, you will hear the high-pitched beat frequency which becomes lower as you tune in; just about the time you get to 'zero beat', the audio pops up and you can tune for optimum fidelity.

In both cases, tuning the Displacement Modulation signal is exactly the same as tuning in an AM signal.

The MFJ-8100K would make a great parent-child or grandparent-grandchild project, especially since it will accommodate two pairs of headphones for short wave listening! Modern stereo headphones are used with both receivers; the jacks on the MFJ are 1/8-inch phone jacks, while on the VEC-11 series, the older 1/4-inch jack is provided in the kit - modifying this with a 1/8-inch jack would be easy for any amateur.

Larry K0WUQ
Feb 14th, 11:57

K0WUQ

Joined: Dec 3rd 2012, 11:13
Total Topics: 0
Total Posts: 0
THE PROOF OF THE PUDDING IS IN THE EATING

A natural question that occurred to me is this:
"How do you know you're getting what you think you're getting? The Handbook says (or said 56 years ago) that this doesn't happen. You say you receive this signal as though it were AM - how do you know it isn't? I mean, yeah, a transmitter can be rigged to synthesize such a modulation in the output tank circuit. But that signal is an RF voltage - even in the antenna it's a voltage with a corresponding RF current. But what if it doesn't RADIATE that way! What if it somehow simply becomes an AM signal as it gets out the door?

I think this is a very good 'challenge question', and it's one that I HAD TO know the answer to. So, here's what I did:

My old Eico Model 460 DC-Wide Band Oscilloscope now has its 'permanent' location in a separate room from my 'shack' - a distance of maybe 15 to 20 feet from the transmitter. This isn't exactly DX, but it's far enough that we know what shows up there isn't caused by the electromagnetic field of the amplifier's tank coil or some such; IF the signal can be picked up at the scope's normal location and IF the modulation appears there, it HAD to get there by propagation through the intervening (admittedly, short) space.

In the scope's working location, I have the chassis grounded to a ground bus that is connected to an outside ground rod. This means that if I tie a few feet of wire to the Vertical Input terminal, I can see signals - lots and lots of them! For one thing, broadcast station WHO, a 50,000 watt blowtorch, is about seven miles due south. What I needed was selectivity. Here's how I got it:

My father's brother Wendell Cottrill had been a US Navy radio man in Dubbiya Dubbiya Two, The Big War. He later worked that into a design career for outfits like GE and Honeywell. Anyway, back in the 1970s, Uncle Wen gave me a bunch of war and industry surplus pieces, including some nice DC meters and a handful of largish transmitter-type tuning capacitors. All this was still in my junk box. And, I had a nice coil I bought new back in the 1960s, nursing my delusion of building a 1KW push-pull linear amp for 40 meters. This was an Air Dux (R) coil 3 inches in diameter x 10 inches long, wound exactly 10 turns per inch out of what appears to be about no. 14 tinned hard copper wire.

I had already used one of these capacitors to build my primitive antenna tuner, but there was still a nice one that had only about a dozen widely spaced plates - a real high voltage job. All I had to do was determine how much of this giant coil to use with it. I set up the capacitor across the Vertical terminals on the scope, set the capacitor to its mid position, and set up the coil across it in such manner that I could play with tapping any number of turns desired. I used a 6ft scrap of insulated hookup wire at the Vertical Input terminal as an antenna. With the transmitter running CW, I was amazed to find that just seven full turns was about right. So, I cut the coil off at seven turns (leaving plenty of lead length at both ends) and solidly re-built my 7 MHz heavy-duty tank circuit for serious use. I found that both my crystal frequencies were within the tunable range.

For an audio source, I used a little single-transistor howler (much like a typical code oscillator, but with no pitch adjustment). Strangely, the output from this tiny creature is not a sine wave - it is more of a 'shark fin' wave, somewhat rounded with a little 'plateau' near the bottom of the curved leading edge (a 'front porch' for all you old TV men out there - ha). This was applied to the Audio Input terminals on my amplifier using a couple of short lengths of hookup wire.

I started up my amplifier with the key locked down. I tuned in the unmodulated signal on the 'new' tank circuit and adjusted the scope so I had an RF signal a couple of inches high. Note that on this scope, an RF signal cannot be resolved into visible cycles - the best horizontal sweep is much too slow for this, so a steady RF signal is always just a uniform horizontal smear across the screen. Anyway, I used a low sweep speed, since the audio modulation is what I was looking for. I went into the shack and switched on the little audio oscillator, then went back to my workbench to have a look.

THERE IT WAS! Well, almost - I had to turn down the Vertical Gain to get it all in, and it wasn't the shark fin. It appeared to be a sort of 'tombstone wave'. Obviously, a little distortion had crept in somehow - ha. Studying this a while, I realized what the problem was.

What I was seeing was EXTREME overmodulation! The audio amplitude was actually MUCH larger than the RF amplitude. I was absolutely astounded. It was as if the roles were reversed - the audio seemed to be the carrier with the RF riding on it, like ocean waves coming in on the tide!

Yes, distorted - but absolutely proving that Displacement Modulation really does work, with propagation between sender and receiver! The first time you see this on the scope, it is ASTOUNDING! At least, it was to me.

Since then, I have replicated propagation numerous times at reasonable audio levels, using a small audio amplifier and recorded tracks as the sound source. It works as anticipated every time.

Larry K0WUQ
Feb 19th, 12:49

K0WUQ

Joined: Dec 3rd 2012, 11:13
Total Topics: 0
Total Posts: 0
MORE DISTANCE

A few days ago, I tried a test receiving the signal modulated by a recorded track over a distance of about 30 meters from the transmitter. Again, I am operating in the 40M band. I was elated to find that I could easily receive and demodulate the signal. I was NOT elated at the audio quality, which was heavily distorted, especially when the VEC-1140K was tried.

This at least proves that the modulated signal DOES propagate over a reasonable distance. I think that the distortion was a combination of my antenna methodology and the high sensitivity of these little receivers. Here's what I was doing:

I set up the receivers out in our old 'tin shed' barn, where the 'shocker' (high voltage pulse source) for our electric stock fence is housed to keep it out of the weather. The shocker was powered off and the high-voltage wire to the fence disconnected from it. Then, I used a 6- or 7-ft length of insulated wire to connect the fence to the Antenna terminal of the receiver to be tested. Another piece of wire was used to connect the receiver's Ground terminal to the earth ground that's driven in directly below the shocker. My reasoning was that with such a small driving signal, a couple of thousand feet of antenna (averaging maybe 3-4 ft above ground) would be good. But, this was a poor choice - I should have just taken 20 or 30 feet of wire out with me and strung it outside as a temporary antenna.

I should have also tried 'no antenna, no ground' and 'ground without antenna' modes, but I was so disappointed with the performance I was getting, I just didn't think of it. I might try these when we get another comfortably warm day. I have a hard time outside on cold days because of my long-term affliction with Raynaud's Syndrome, especially affecting my hands.

But anyway, Displacement Modulation does give us 'action at a distance', so getting it exactly right is just a matter of time and experimental patience. I'm sure I'm getting some distortion at high audio gain because of inadequate audio gear. I can observe this even when testing with antennaless receivers at the work bench right alongside my scope. Another problem is fairly bothersome AC 'background hum', which I'm fairly sure is coming across as AM (I can pick this up when monitoring CW sending, with the modulator totally dead). I haven't determined whether this problem is due to my amplifier's unregulated power supply, lack of shielding, or other factors. Some experimentation required, as usual.

Larry K0WUQ
Feb 21st, 14:08

K0WUQ

Joined: Dec 3rd 2012, 11:13
Total Topics: 0
Total Posts: 0
BETTER TEST, BETTER SETUP

Yesterday I went back out to the tin shed with just the MFJ regenerative receiver, this time using the receiver ungrounded with just the 6ft insulated wire as the antenna. I unplugged the fence shocker so I wouldn't be overwhelmed by 'lightning' pops once per second. Reception of the audio track was strong enough and pretty clean except for momentary distortion on the loudest sound peaks. A pretty impressive demonstration of the modulation method, anyway.

Note that with this method of modulation, most audio distortion transmitted will result in a small amount of amplitude modulation (because of the difference in shape between the upper and lower RF envelopes). This means that at those moments, sidebands will be created, though their amplitude should be pretty small. Something to be careful of, though, since one of our big 'selling points' is minimal bandwidth.

I had an old Knight Kit stereo amplifier sitting around, not used in decades, so I decided to try it out, in place of my old 10-watt Heathkit one-channel solid-state amp. The Knight amp is all tubes and also rated at 10 watts (per channel), but it has transformer output with 4-, 8- and 16-ohm taps. I set it up as the modulating amp using the 'A' channel's 16-ohm output (to get maximum audio signal voltage for the modulator input). I cabled my mic preamp into (believe it or not) the magnetic phono input on the amp. I did this to get maximum gain for the mic input. Fired her up, and Bob's your uncle - it works, with MUCH less 60Hz hum than what I was doing before!

Before that, though, I made a significant mod to my homebuilt RF amp. Because of the components used in the bias strip (cathode bias, grounded grid) I was limited to 100 mA of overall tube current (and about 1/7 of that is diverted through the Screen grid, so no output power from that part of it). I had noticed that biasing for higher current lessened audible distortion, so I put a big 560 ohm resistor (from the junk box) around (i.e. in parallel with) the existing bias strip. This lessens the range of adjustment, but increases the maximum current I can drive by 30-40 percent! Sure enough, this allowed higher power operation of the tube and peak distortion was less noticeable at fairly high modulation. The increased power did NOT make the plate of the tube run red hot, though I'm now running about 40 watts on a tube rated at 20.

Incidentally, my modulation transformer is a speaker output transformer cannibalized from a Heathkit transistor radio I built back in 1959. It is roughly the size of a golf ball. It is used 'backwards' with the (assumed 16 ohm) output coil used as the primary and the high-impedance side used as the secondary. Because it is connected to the control grid of the tube, it handles no current and therefore consumes zero power. The reason a power amp is used to drive it is the audio voltage required on the low-impedance input, not any need for audio power.

The scope tells me I'm still not using full modulation, even with the observed distortion. Now that (I think) I have a better audio driver, I'll go back to doing more experimenting with the recorded track.

However, using the Electrovoice DL-42 microphone, I don't perceive any audio distortion coming through the MFJ receiver. But of course, hearing your own voice in headphones while speaking into a mic is always self-deceptive to some degree - you always sound better than you really are.

Larry K0WUQ
Mar 9th, 14:07

K0WUQ

Joined: Dec 3rd 2012, 11:13
Total Topics: 0
Total Posts: 0
SUCCESSES and NOT SO MUCH

By now, I should be able to report full success at this every time I try it. But, I am forced to admit that I cannot. Much of the time I am getting unexpected results, as seen on my scope driven by my makeshift tunable tank circuit. In all cases, the 'failure' of my method is evidenced by getting low-level AM instead of the Displacement Modulation desired.

This doesn't mean I haven't proven anything; but to a scientifically skeptical reader, it throws doubt on either my claims, or at least, on my methods of achieving them. Obviously, it's a lot easier proving something to yourself than proving it to an 'unbiased' outside-world observer. But, here are the things I THINK my efforts have shown, so far:

1. It is easy to create a Displacement Modulated wave in an RF amplifier and it is provably possible to propagate such a wave across 'space' to a receiver that can detect and demodulate it.
2. However, our ability to propagate such a wave is not sure-fire - there are problems of unclear origin.
3. It has not been proven that the created wave is totally free of a low-level amplitude modulation component - this AM component has been observed (always at very low level) even when Displacement Modulation is STRONGLY propagated.
4. High modulation levels have never been established without noticeable audio distortion at the receiver; this distortion has not been visible at the scope trace, however.
5. Differences in setup, procedure, etc. that lead to success or failure of the modulation are not clear; a variance that seems to work one time fails when I try to duplicate it another day. This means there are unknown systematic errors.
6. The modulation, when done at 7 MHz RF, can be filtered out by a .1 mF 'blocking capacitor', although any AM present will make it through perfectly. This doesn't happen when a much smaller capacitor is tried (e.g. .001 mF). Thus, it is not simply low frequency blocking through a passive high-pass filter. The main reason this might be important is that any 'capacitance' effect in the atmosphere between transmitter and receiver could diminish propagation.

Sometime soon I will craft Yet Another Reply that will 'flesh out' these observations, just for those who enjoy slogging through the weeds.

Larry K0WUQ
Mar 9th, 17:11

K0WUQ

Joined: Dec 3rd 2012, 11:13
Total Topics: 0
Total Posts: 0
AN AHA! MOMENT . . .

A little while ago, while cleaning out one of the horse stalls, it occurred to me: What if I'm ALWAYS creating AM, and the large amount of distortion I get on demodulation IS the AM?

In Displacement Modulation, ANY error between the shape of the upper and lower RF envelopes will be, by definition, distortion. That is, to be UNdistorted DM, the moment-by-moment algebraic DIFFERENCE between the two envelopes must be CONSTANT. This will not be true if ANY AM is imposed on the RF signal along with the intended DM.

The same thing would be just as true the other way around - what if the demodulation I've been listening to all along has really been the AM part? The DM part would be heard as extreme distortion.

I will figure out a way to test this - probably involving a .1 mF capacitor ;-)

Larry K0WUQ
Mar 14th, 13:01

K0WUQ

Joined: Dec 3rd 2012, 11:13
Total Topics: 0
Total Posts: 0
CLARIFICATION ON CAPACITANCE EFFECT

I have now done a simple set of experiments clarifying the 'blocking capacitor' effect mentioned earlier. The results are interesting and useful.

My initial observation of the effect was that when Displacement Modulation was achieved, the oscilloscope showed the modulation clearly when the Vertical Attenuator was used in its DC mode, but the modulation disappeared in its AC mode (i.e. the signal was shown as nothing but the RF carrier, exactly like a CW transmission). Looking at the scope schematic, the ONLY difference is that in AC mode, the first thing the Vertical Input signal meets is a .1 mF blocking capacitor (basically, this allows you to measure a small AC ripple on a high-voltage DC power supply, for example).

Using the scope in its DC mode and scoping a good-quality low-level modulated RF signal, I tried a few values of capacitors between the top end of the tank circuit and the Vertical Input terminal on the scope. Results are as follows:

.001 mF old-fashioned tubular: Modulation blocked except for tiny residual ripple
260 pF disk: Modulation displayed at original amplitude, slightly distorted
520 pF (two disks in parallel): Reduced modulation amplitude, more distortion
780 pF (three disks in parallel): Reduced amplitude, severe distortion
1040 pF (four disks in parallel): Reduced amplitude, modulation clearly visible but no longer resembles original - note that this capacitance closely matches the small tubular tested first (.001 mF = 1000 pF)

So we see that small capacitors allow the modulation to get through in a distorted mode, UNLESS old-style tubular capacitors are used! This is useful because:

In the Displacement Modulation wave, ANY deviation from the upper RF envelope being identical to the lower envelope is in fact low-level AM. This is a basic type of distortion we could get by, for example, driving the modulation beyond what the amplifier will allow (basically, over-modulation). By using the blocking capacitor, the scope trace will show us ONLY the AM modulation component. So. IF AM exists, it will get past the blocking capacitor and will show up on the scope. This AM will be very low-level compared to the Displacement Modulation developed, and will produce correspondingly low-level sidebands. I assume the AM and sidebands would be distorted versions of the audio input.

Thus, we could use an ordinary oscilloscope with such a blocking capacitor as a simple 'over-modulation detector'.

Larry K0WUQ

Back to Top

EXPLORE ARRL

Instragram     Facebook     Twitter     YouTube     LinkedIn