ARRL

Register Account

Login Help

K0WUQ

Joined: Mon, Dec 3rd 2012, 11:13 Roles: N/A Moderates: N/A

Latest Topics

Topic Created Posts Views Last Activity
Displacement Modulation - I Was Just . . . Wrong Jul 31st 2023, 11:52 1 1,424 on 31/7/23
Displacement Modulation - Using an SSB Receiver Jul 23rd 2023, 15:19 1 1,388 on 23/7/23
Displacement Modulation - Next-To-Last Mystery Solved Jul 3rd 2023, 15:09 1 1,462 on 3/7/23
Why Displacement Modulation Is NOT the Easiest Thing In the World Jun 23rd 2023, 16:12 1 1,459 on 23/6/23
Why Displacement Modulation Is the Simplest Thing In the World Jun 23rd 2023, 14:51 1 1,354 on 23/6/23
EICO 723s Jun 23rd 2023, 02:23 3 1,687 on 11/7/23
Run Plate Transformer at Half Voltage? Feb 28th 2022, 12:21 1 4,296 on 28/2/22
How To Do AM? Feb 25th 2022, 21:40 1 5,326 on 25/2/22
Flip Antenna Phase? Feb 14th 2022, 23:55 6 4,405 on 16/2/22
Heterodyne Rockets Feb 13th 2022, 12:52 3 4,557 on 15/2/22

Latest Posts

Topic Author Posted On
Displacement Modulation - I Was Just . . . Wrong K0WUQ on 31/7/23
Well, after four years, I have The Final Answer. The Radio Amateur's Handbook, 43rd Edition 1966 was NOT wrong . . . I was. And, I hereby apologize to the editors of the Handbook (ancient and modern), and to everyone who has spent some of their valuable time following along through these topics.

Displacement Modulation, as I have described it, CAN be created in the ways I have discussed AS a current wave and voltage wave in the plate circuit of a simple transmitter - BUT it cannot propagate through the ether, just as the Handbook article warned.

Until yesterday, there has always been a Great Mistake in my test methods: I assumed that if I was intercepting the 'ether wave' from a distance of 15 or 20 feet from a fairly low-power transmitter (30 to 60 watts plate power) that I was seeing the pure wave accurately. But until yesterday, 30 July 2023, the only 'distance testing' (about 100 ft out) was with receivers that appeared to handle the modulation perfectly.

Yesterday, for the first time, I lugged my 'wave sensing' oscilloscope (EICO 460 equipped with a 40M tuned tank at the Vertical Input terminals and with antenna and earth ground) outside over the 100 ft or so to our old corrugated tin barn. Once set up, I was able to see the signal thrown over a reasonable distance, for the first time. What was detected was nothing but a weakly modulated AM signal. There was no evidence whatsoever of my carefully nurtured modulation.

This is exactly the result predicted by some engineer friends of mine at the QRZ.com forums. They explained the theoretical basis for the impossibility of propagating both RF and audio frequency concurrently (which is exactly what I was trying to do, when you think clearly about it) and so I finally realized I had to do such a test. I really expected to be vindicated by the observed physical reality - but, it was not to be. As an old farmer finally admitted as he was struggling to give travel directions to a passing stranger, "Well ... ya just can't get there from here!"

So, I apologize to everyone who wasted any time over the last couple of years in following along with this. Perhaps a few useful things were learned along the way - there were for me, for sure. The only way what I did could be used for anything would be a case were we needed hard-wired RF with AF modulation; but, in such a situation, AM would work just as well and is just as easy to produce.

One thing my engineer friends tried to point out (though their equations were as incomprehensible to me as Swahili) was that my method would ALWAYS produce some AM (and thus, some sidebands, of course), even if perfectly executed. This is why my 'distance' experiments with three different receivers appeared to validate the method - what I was tuning was really just AM after all.

Using the equipment I designed and built, as well as purchased and modified, I will now turn to learning all I can and perfecting (as much as possible) AM done at the control grid of the RF amplifier. This method, though somewhat inefficient, allows modulation with a minimum of expense at the audio end, because the AF power requirement is incredibly low even when excellent results are achieved.

I am, as always:
UNDER CONSTRUCTION
THANK YOU FOR YOUR PATIENCE

73 to all,
Larry Cottrill K0WUQ
Displacement Modulation - Using an SSB Receiver K0WUQ on 23/7/23
Well, last week I got myself a "new" receiver via eBay - a quite nice looking Heathkit HR-1680. The Copyright date on the book I downloaded says 1976, so this design is almost 50 years old, practically state-of-the-art by my standards. It's very clean except for just a thin film of oxide on the RCA connectors in the rear, which was easy to fix up.

After connecting this to my outside antenna and ground to check out its basic operation and get used to the controls, I moved it into the workroom and set it up in place of my MFJ Regenerative Receiver (my transmission monitor). This means it has a good ground to a driven rod outside, but only has about a 3ft dangling wire for an antenna, and is physically a stone's throw from my EICO 723 transmitter (maybe 14 or 15 feet). OK for preliminary testing.

Today I took this receiver out to 'the old barn', a ritual I refer to as 'portaging the new receiver'. This is an old corrugated steel shed that can't keep out rain, wind or snow but DOES have these critical features: 120VAC power, an 8ft driven ground rod AND connectivity to The Giant Antenna. This antenna is usually called the 'electric stock fence' and comprises maybe a quarter mile of bare wire located 3ft above ground level on average and is well insulated from ground except for contact with about a thousand weeds of variable moisture content. I just disconnect it from the fence charger and use a couple of AWG 14 insulated hookup wires to connect it and the ground rod to whatever receiver I'm working with. The closest point of the antenna to the station is approximately 1 wavelength away (40 meters in my case, or just over 130 feet) - not exactly DXing, but better than just testing from the next room in my basement.

This receiver is quite sensitive and is very simple to set up and use. Both testing locales produced almost identical results, even in terms of signal strength and background noise (which was pretty horrendous). I first ran the transmitter without voice modulation - finding it on the dial was easy due to the incessant 60 Hz hum that I have never been able to keep out of my transmitters. (Someday when I'm gone all anyone will remember about my voice online will be that '60 cycle hum'.)

The most surprising (to me) and disappointing finding was that I am able to tune the signal as both upper and lower sidebands, at good signal strength. Sidebands just aren't supposed to be happening with this modulation! But, there they are. When I found this on the short haul testing, I really hoped they would disappear after portaging the receiver out a good distance (i.e. hoping this was a 'local' effect due to proximity), but not so. The recorded voice sounded just like 'sideband voice' as you tune into it. Nuts.

On CW, I got exactly what I hoped I would get - as you tune onto the station, the CW whistle (from the carrier) drops down and as you get close to 'zero beat', the voice quickly rolls into place. When you get it at just the right spot, the voice quality is really good with no noticeable distortion. Although the sideband voice was intelligible, this CW version was almost full fidelity, the two flaws being that (a) the eternal background hum is noticeable, and (b) strangely, you don't seem to be able to quite get to 'zero beat' - just down to a very low growl and then the tone heads upward again on the other side. I suppose this second effect could be a receiver alignment problem.

Two interesting controls on this receiver have noticeable effects on the whole process. The Preselector control is probably the most 'front end' control on the unit, and is supposed to be readjusted any time you change bands, to get optimum reception. At 40 meters, the 'sweet spot' is just past 50 percent rotation - roll it to either side of that and the audio quality drops rapidly (as does the background noise).

The RF Gain control is more interesting in terms of my modulation. It's a very sensitive adjustment - turn it counter-clockwise just beyond a certain point and ALL audio just 'rolls off the table'. However, just to the right of that spot there's a little magic zone where the background noise drops off way more than the demodulated audio - so, though the audio is pretty low level, it stands out clearly and you can really hear the quality of the spoken voice. Fascinating. (Remember, this is in the CW Mode setting.)

CONCLUSIONS:

Ordinary amateur voice communication could be carried out with this modulation today. All that's required on the transmit side is a slightly modified CW transmitter fitted with a small audio amplifier, mic, mic preamp and an old-fashioned audio output transformer. For receiving, apparently all you need is a classic SSB receiver with a good CW mode setting.

Of course, nothing is proven about the 'punch' of such a signal in competing with noise, nor of its ability to interfere with other traffic. Having it show up as tunable sidebands is certainly the greatest disappointment. As usual, more experimentation required ...

EDIT:

Of course, as often happens, my observations might be a lot better than my conclusions. I am very ignorant of how SSB actually works on the receiving end, and I ASSUMED that a receiver in its LSB or USB mode can ONLY tune that kind of signal. Again, silly me - that's probably not how it works.

In another thread a long time ago, someone said that an SSB receiver had no problem tuning an AM signal. I assumed that was because it would just tune onto one of the sidebands (which would always be part of such a signal) - but maybe he was really saying that even in a 'sideband' mode the carrier would be detected and demodulated. If that's the case, then the above-stated 'evidence' of sidebands on my signal is simply bogus.

One thing I recently equipped myself with is a little device called TinySA - a really neat piece of gear. However, looking at discussions about it, it doesn't seem to be able to show the presence of sidebands on a modulated signal. It just can't slice the universe fine enough to separate them out. What kind of device would be needed to do this? Or, is there another testing technique that would actually prove the presence or absence of generated sidebands?

This is a crucial question, because the narrow bandwidth of the Displacement Modulation signal is the only 'selling point' that makes it worth the bother of doing this.

Larry K0WUQ

EICO 723s K0WUQ on 11/7/23
Well, the filter capacitors in my newer 723 bit the dust a couple of days ago. I had just powered up, and was tuning up when I began hearing some snapping and popping in my headphones (i.e. coming through my monitor receiver). I immediately thought of those original style tubular electrolytics breaking down (they are red cardboard tubes resembling roadstand fireworks), and shut her down quickly. Fortunately, I already had some replacements on order - I am as smart as the next guy and ordered four of them, enough to re-fit both my 723s. Mouser didn't have 40uF 450V in stock, so I ordered 47uF 500V ones, certainly close enough to work.

Today this order arrived a couple of hours after the regular mail, and I wasted no time getting them installed in both transmitters. They are half the physical size of the original ugly ones. I was pretty much convinced that the transformer in the older 723 was shot, and when I powered her up, I was not disappointed - all I got was smoke from the transformer.

When I powered up the newer one, things were much better, which means I responded properly in immediately shutting down as soon as I sensed there was trouble. While I was in this unit, I also fixed up the 120VAC output that is used to pull the antenna Send/Receive relay. That had mysteriously stopped working the other day - this turned out to be a cold solder joint from the factory! So, it was easily fixed.

Altogether, a pretty blessed day at the workbench. All I need for the older transmitter is an impossible-to-get power transformer that will fit in the limited space designed for it. Meanwhile, the second unit is working fine and I'm back to "cookin' with gas" running with that one.

Larry K0WUQ
Displacement Modulation - Next-To-Last Mystery Solved K0WUQ on 3/7/23
In my earlier discussions, I mentioned that a mysterious problem was that the modulation could not be obtained when the plate tank circuit was fully tuned to resonance - instead, I had to significantly de-tune from the Peak Efficiency Power tuning point, in the direction of reduced capacitance, for the modulation to get through to the antenna. This has been true in my little homebuilt rig, and also true to almost exactly the same extent on my simple modification of the venerable EICO 723 60-watt transmitter (ca 1960).

So last night, I suddenly wondered, What would happen if we forced the plate tank to resonate at a higher L/C ratio? In other words, push the capacitance radically smaller by radically increasing the coil inductance? This morning, I decided to try this.

In the EICO 723, the plate tuning capacitor is directly tied to the 'bottom end' of the mult-tap coil (a coil about 3.5 inches long and almost 1.5 inches diameter, maybe 25 or so unevenly-spaced turns) - the whole coil is used for 80 meter tuning, which I don't care about. I first removed the wire link from the capacitor to this point. Then, I unsoldered the wire at the 40-meter tap and moved it over to this point at the end of the coil. This means I'm using the former 80-meter coil to tune to 40 meters with the selector switch at the '40' position.

Next, reducing the tuning capacitance: My first stab was a porcelain-bodied trimmer of unknown value between the capacitor and this coil end (meaning that this cap and the tuning variable were in series). When I fired her up, I could never get the plate current to dip - obviously, this series cap must be too small (according to the 723 parts list and schematic, the tuning variable is 140pF max.). I still have a couple dozen 280pF 3KV disks lying around, so I put one of those in the series position. Voila! I was now able to tune the plate tank to resonance.

I was immediately able to observe something I was sure would happen, although the effect was more pronounced than I expected: The change in the tank circuit caused the tuning 'peak' (the sudden dip in plate current at resonance) to become a 'hill' - the current dip was no longer a sharp 'hiccup' but rather more of a 'shallow breath'. What this means is that the Q of the plate tank had been greatly reduced, meaning: (a) much 'softer' tuning to resonance, and (b) much less peaking of efficiency and RF power (similar to the de-tuning I was doing before)!

I now started up my monitor receiver and monitor 'scope (the one with the tunable tank circuit and short antenna on the Vertical Input terminals). When I keyed the transmitter, I was met with the first really interesting discovery of the day, to wit:

During all my experiments with this modulation, I have been plagued by a sizeable AC hum which shows up as a displacement modulation of its own. This varies in degree from time to time, but is always there. It even showed up as hum in my CW signal when I was tuning for that. Always very frustrating, and nothing I've tried has ever removed (or even reduced) this. Now, however, with a much lower Q tank circuit, this is almost entirely gone - the 'scope trace looks almost like a textbook picture CW signal. Fascinating!

So then, I powered up my audio amp and started my 'test tape' of a male speaking voice. To my amazement, the modulated wave pattern looks and sounds flawless, WITH NO DE-TUNING whatsoever! My little Heathkit 10-watt amplifier has its own Volume control, and using this easily controls the precise amount of modulation. I observed no obvious audible distortion. This was exactly like the best results I had ever obtained previously - and WITHOUT the annoying background hum!

IMPORTANT Observation: What this means is, we can now tune the rig in a 'normal' manner, without monitoring a transmitted signal during the process! I regarded this as one of the most egregious flaws that needed resolving to make the method truly practical.

CONCLUSION: For the modulation to make it to the antenna jack, DON'T use a tightly-tuned high-Q plate tank (the kind you would design for highest plate efficiency). Instead, provide a lower Q tank of higher L/C ratio, and accept the lower efficiency and softer tuning to peak power as simply a 'cost of doing business' with this form of modulation.

THE REMAINING MYSTERY: Are there modern, state-of-the-art receivers that can detect the audio out of this signal? My little Regenerative and Direct Conversion receivers handle it easily, but most serious hams would see these as practically toys compared to the full-functioned, multi-hundred-dollar units everybody wants today. So far, I have never had a response to a CQ I've thrown out there. So, are modern receivers just 'blind' to these signals, or am I just waiting for the golden moment when I'll get lucky?

EDIT: Well, I was too enthusiastic about the 60 Hz hum - it's still there, just less obnoxious than it was. Pretty much inaudible when the voice modulation is going, but still noticeable when the voice is quiet. As I said, this effect varies over time - when I saw a really smooth-looking 'scope trace it must have been at a 'quiet moment'. I was a little while ago speculating that the running of the heat pump might have something to do with it, but just now spent a few minutes with the heat pump definitely not running and there was basically no relief from the effect. So, who knows? Sorry I jumped to conclusions on this one.

Yet Another EDIT: Further experimentation resulted in further simplification. With the crystal I got with my first 723, marked '7183 KC' it turned out that the fixed cap in series with the tuning cap is unnecessary - the plate tuning variable has enough range that it reaches 'dull resonance' with the 80-meter coil, simply at a very different spot in its rotation. It was silly that I failed to find this on my first try, since I think we all know that tuning variables cover a huge range of capacitance values.

Also, somewhere along the line I should have mentioned that in getting this to work on the EICO 723, the grid tuning has to be set to a pretty low value, under 1.0 mA, for the modulation to work. I am convinced that most of the problems encountered in my experimentation were not true 'design' problems, but rather, failures to tune things up to agree with some Laws Of Physics that were poorly understood by me.

Larry K0WUQ

EICO 723s K0WUQ on 27/6/23
Today I made yet another simple modification. I put almost 200 ohms of resistance (takes about 6 watts) in the cathode circuit of the amplifier tube. The reason for this is that plate tuning for Displacement Modulation is pretty far from the Peak Efficiency tuning point, so the plate current runs above the 120 mA that EICO considers to be the running limit for the tube. Resistance in the cathode circuit creates 'cathode biasing', causing the control grid to be relatively more negative and thus, choking down on the plate current. With this in place, I was able to stay down around the 120 mA point with good-sounding modulation achieved.

Tuning for good modulation sort of requires three hands, because the grid, the plate, the loading capacitor and my 'random wire' type antenna tuner all affect the outcome. This is the one serious flaw in using this form of modulation - the Goldilocks Syndrome, where it's only satisfactory when everything's 'just right'. But, "It works if you work at it." Once you hit the sweet spot, my little 10 watt Heathkit audio amp gives full modulation with the volume knob turned only halfway up, showing the minimal power required to modulate in this way.

To be perfectly honest, the modulation is not perfectly distortion free - but I think that comes from the non-linearity of the tube used as the final amp. There might be another beam power pentode of the same power rating that could be substituted (maybe with a little re-wiring of the tube socket), which would have essentially linear characteristic curves. If that could be done, almost perfect modulation with zero sidebands would be achieved.

A also bought (eBay) a 3-inch D'Arsonval 5mA meter to replace the original jittery one, but haven't seen it on the doorstep yet. That will spoil the authentic look of the front panel, but the convenience of a decent meter will be more than welcome.

Larry K0WUQ

Back to Top

EXPLORE ARRL

Instragram     Facebook     Twitter     YouTube     LinkedIn