The AM Forum

A friend has an Icom IC7300.  He says he likes it, but it has excessive carrier shift.
Why would a modern radio have carrier shift ?

Don W4DNR

What exactly does he mean by "carrier shift" — the frequency changes under modulation? If so, I'd say there's something wrong with his particular radio or power supply.


Don

Usually carrier shift refers to the modulation going more in one direction than the other, like 100% negative but much less in the positive direction.

 
    I have listened and talked to several 7300s on AM driving amps, like I do with my TS990.
 There is no frequency shift, six kc wide, wonderful audio . Maybe that person has an  AC volt
drop problem with the icom  12 vdc power supply when in AM mode.

 Bob  W8LXJ

My friend and I , both being in Broadcast AM for years, I took it to mean just that.   

Negative 100% and Positive 80% modulation wouldn't be that bad  ( in my opinion ) for a radio designed more for SSB.

Negative 100% and Positive 125%  ( if done without splatter ) would be better I would think.

Maybe he's just a 100/100 purist ?       

DOn W4DNR

According to 47 § 73.14 AM broadcast definitions:


Another term for carrier Amplitude regulation is carrier shift.

This seems to indicate as to how much carrier stability is maintained under modulation.

In many of the modern transceivers (and I have an Icom-7200) uses Digital Low power modulation with the modulator section feeding the linear amplifier section.

The internal ALC system largely regulates the output power under various conditions and when modulating in the AM mode, the output under modulation is similar to the older screen grid modulation in say the DX-60 where carrier power seems to drop under modulation.

Some control of power output can be done by biasing the ALC externally in many of these transceivers.


Phil - AC0OB

Phil,

This may be splitting hairs, but doesn't carrier always drop when modulation is at 100% ?

Modulate a carrier at 2000 cycles ( hertz )  and at 100% modulation ,
there is no carrier and all of the power is in the sidebands out at + and -  carrier .

Or is this saying that between voice peaks, the carrier is not only reduced by the modulation, but also that the whole waveform is being reduced by the slow ALC . . . ?     Reducing the possibility of 100% Positive modulation, or just reducing the whole AM envelope by a power percentage ?

Thanks

Don W4DNR

Perhaps the PDF below will explain the power relationships for Amplitude Modulation.

I am not clear about the last two sentences; were they directed toward high-level modulation systems or DSP controlled transmitters?  

A good explanation of the shortcomings of SS transceiver ALC, especially in the Icom 756 is given here:


http://www.w1aex.com/756AM/756AM.html


Phil

I wonder how the carrier shift is being measured. As an input, is a tone generator being used or your friends voice? How is the output being measured?

There's no accurate answer here without accurate measurements in a controlled test environment.


Don

A friend of mine has a IC-7300 and also complained about the "AM Carrier Shift".  When asked what he was referring to, he replied that when you transmit in CW mode the max power out is 100 watts, but when in AM mode, the unmodulated carrier dropped to less than 25 watts keeping the modulated peaks from achieving 100 watts on peaks.  To me that is not "AM Carrier Shift", but I believe some folks use that terminology to describe the above.  He called Icom about it and was told that in AM mode, the IC-7300 will not allow for 25 watts AM carrier in the software to help protect the finals.  I guess the designers believed that AM mode with 100% modulation was too much if peaks reached 100 watts.  Thus it will not modulate to 100 watts peak, but will go up to about 90 watts max.

73....Myron....K4YA

  On the Kenwood , when using it as an AM exciter, AM mode, power output wide open, carrier level on zero and then insert carrier for desired output of RF amplifier. Often run a passive grid driven 4-1000a in ab2, 4000vdc plate , 500vdc on the screen, 300 watts out carrier, 1,200 watts out on audio peaks.. If I want more power out, turn up the mic gain or increase the screen voltage to 600vdc and will go the 1500 out with no problems. THE KENWOOD shows about 10 watts output for carrier and 15 watts for audio peaks

Ok, that is not carrier shift but is a result of transceiver design and math.

25 Watts AM carrier results in 100 Watts Peak-Envelope-Power at 100% Modulation. The maximum PEP for SSB is 100 Watts as well, the limit of the internal Linear Amp stage.

See the listing in the PDF file for a transmitter capable of 25 Watts AM carrier and compare the real AM output power with the so-called PEP values.

The ICOM-7300 Specifications show this:

Output Power:
100W (25W AM)
RX Frequencies:
    0.030-74.800
Receiver Type:
    Direct sampling

Transmitter
Output power (HF/50MHz)   SSB/CW/FM/RTTY: 2–100W, AM: 1–25W
Modulation system SSB AM FM   
Digital P.S.N. modulation
Digital Low power modulation
Digital Reactance modulation
Spurious emission   Less than –50dB (HF bands), Less than –63dB (50MHz band)
Carrier suppression   More than 50dB
Unwanted sideband   More than 50dB
Microphone impedance   600Ω

https://www.icomamerica.com/en/products/amateur/hf/7300/specifications.aspx

Phil

Ah Ha !! ;D      ALC..... I hadn't factored that into the equation ! 
A perfectly created 25 watt carrier and 100 watts at peak modulation
is squished in the next stage by the ALC.

I'll have to do another scope measurement on my TS-2000.
Last time I measured, 25 watts carrier was pretty close to 100 peak
with no ALC.   Close enough that it isn't worth an effort to improve.

With a 200 watt radio like the TS480 ,  25 watts carrier should be well below
any ALC action.     

I have an old FT-757GT ( no DSP )  that always sounded good on AM, but I don't
remember looking at it on a scope.

Then there is the Valiant, the DX-60, and the DX-40 that need  some "tweaks".
At least when I retire, I won't run out of projects.

Thanks Everyone ! !   Great Information.


Don W4DNR

If that was true, it would produce maximum "carrier shift" when using A3a emission because the total power in the two AM sidebands at ±100% sinewave modulation is only 50% that of the unmodulated carrier.

Actually, the average amplitude of an AM carrier is constant during all sinewave modulation levels to ±100%, at which time the total r-f output power is 1.5X the unmodulated value, and the r-f current at the output of the transmitter rises to 122.5% of its unmodulated value.

Ladies and gentlemen, it is now post-time.

The electrons are lining up at the gate...

And they're off!

Average Power takes an early lead with Peak Envelope Power close behind and Power Factor lagging as they head into the first cycle.

Around the Smith chart they come. Now Peak Envelope Power is closing in on Average Power, with Root Mean Square challenging on the outside.

And down the transmission line they come. Average Power and Peak Envelope Power are now neck-and-neck with Root Mean Square trailing as Power Factor, Spark Gap, and Modulation Index go into a deep fade.

It's Average Power. It's Peak Power.

Peak... Average.

Average... Peak.

And at the wire... it's a photon finish!!!

term dropping at its finest !    seems the muse has visited  73

Ok, let's get this cleared up.

Carrier shift is caused by AN AVERAGE SHIFT in the carrier power (also known as a DC shift).  This is NOT related to asymmetrical modulation.  Not, not not.

Carrier shift will only occur if the integral (average) power of the modulating waveform is greater in one direction as compared to the other.

I regularly modulate 150% in the positive direction, and 90% in the negative direction (at the same time) and there is no carrier shift.  That is because the areas under the curves of the positive going waveform and the negative going waveform are the same.  The area == the average power.  It's the average power changing that causes carrier shift.

In high power AM broadcasting (>  25kW) there is a system that purposely creates carrier shift to save AC input power and that is:

AM Modulation-Dependent Carrier Level (MDCL)

and if you have further interest this can be found in

https://www.nrscstandards.org/standards-and-guidelines/documents/archive/nrsc-g101.pdf


Phil - AC0OB

Richard  Fry said :

Actually, the average amplitude of an AM carrier is constant during all sinewave modulation levels to ±100%, at which time the total r-f output power is 1.5X the unmodulated value, and the r-f current at the output of the transmitter rises to 122.5% of its unmodulated value.


Richard,

I was referring to "instantaneous" carrier loss at 100% modulation....
As the negative modulating voltage approaches the same but opposite polarity on the final tube plate, the carrier goes to zero.   Too much of this creates clicks or hash that is wide as a barn door.  I can see how ALC could squash the positive going peaks on a modern transceiver, but how can a "classic" transmitter with an ample HV supply be made to get into the region of carrier shift ?

As Steve noted …. "I regularly modulate 150% in the positive direction, and 90% in the negative direction (at the same time) and there is no carrier shift.  That is because the areas under the curves of the positive going waveform and the negative going waveform are the same.  The area == the average power.  It's the average power changing that causes carrier shift.

Maybe the question should be :  If someone wanted to illustrate carrier shift , how would that be accomplished  ?     
   

Don W4DNR

The default meaning of "carrier shift" to most readers/writers is rooted in the definition given in U.S. 47 CFR § 2.1 - Terms and definitions, which states:

      Carrier Power (of a radio transmitter). The average power supplied to the antenna
      transmission line by a transmitter during one radio frequency cycle taken under
      the condition of no modulation. (RR)

So if one asks about the instantaneous value of carrier power during amplitude modulation, that would need to be made clear in the question.

Below is a clip from a trade magazine that might be of interest here.

Note that the 930 kHz carrier in this SDR display of the r-f spectrum of this AM broadcast station essentially has constant amplitude during modulation.

(https://i.postimg.cc/QttrpgZ7/AM-Bandwidth-Article-RW-EE.jpg)

Average. I though I knew what average was, until I heard about average power.

What is average? Mathematically, the definition of average is clear - the sum of all measurement points divided by the number of points measured. Hence, the average amplitude of a sine wave is 2/pi, or 0.637 x peak. This works for voltage or current, but somehow, not for power! Nope, average power is different. Average power isn't average voltage times average current. That would be too easy. This is where √2 comes in and we start talking about r.m.s., the DC equivalent power.  So, it seems that the word "average" has different meanings in different contexts. When it comes to complex wave forms, like a modulated AM carrier, which average is average? Holy crap. This is why I get nervous anytime someone starts talking about average power as an absolute. Absolutely what?

Maybe I've got this all wrong. Or maybe I should just give up electronics and study something more clearly understood and unambiguous - like human psychology. ::)


Don

I just know that I can see "peak" modulation on a scope, or on a properly calibrated AM Modulation Monitor. 

Average power could be useful as an indicator of audio processing effectiveness.

Into a fixed resistance load, if the antenna amps go up, then antenna voltage should also go up.

This still doesn't explain how to achieve what is called  "Carrier Shift".

My point is …. if you know how to achieve "Carrier Shift" , then conversely, you should be able to prevent " Carrier Shift".

Don W4DNR

In the practical world I have always looked at carrier shift like this:

Put a steady AM carrier into a dummy load with an average reading (not peak) BIRD wattmeter inline. Then sock some audio voice yallos as you vary the audio gain up and down and observe the Bird reading. If the meter bounces upwards with heavy modulation, this is positive carrier shift.  If it bounces downwards (like with excessive alc limiting, or improper / light linear or light screen modulation loading) then it is negative carrier shift.

This effect really shows up with super heavily modulated AM balanced modulators, whereas with conventional plate modulation (with proper negative peak limiting and the rig running clean/ correctly) the shift can be slight or not there at all.

I've tailored my 4X1 plate modulated rig to run with little to no upward bounce and this is its best overall optimization for fully modulated cleanliness.

T

But, but, but...

The Bird model 43 measures voltage (it uses a detector diode), in a given direction, with a meter calibrated in watts. Its stated accuracy is +/- 5% FS when operated into a 50 ohm resistive load. I've never seen a Bird 43 in any RF lab I've worked in. An HP432A (+/- 1% FS) power meter, yes. Never a Bird wattmeter.

The HP 432A uses a thermistor, which measures heat. Heat, like r.m.s. heat. Power heat. The HP 432A is a power meter. A Bird 43 is a voltmeter calibrated in watts. So, when a Bird 43 meter needle is flicking around with modulation it's still measuring voltage (damped by the meter movement). I'm skeptical that a Bird 43 can measure a complex waveform power with any kind of accuracy. I doubt that even the HP 432A can keep up with a complex waveform, like voice.

Now, to top it all off, the word "average" does not appear in the manual for either the Bird 43 or the HP 432A. But I have a feeling the the Bird measures 0.637 "average volt-watts", and the HP measures 0.707 "average r.m.s.-watts". But I still don't know what watt is what, and which average is average. :'(

As for carrier shift, I don't have a clue where to even begin. The theory is one thing. Making meaningful measurements, can be quite another.

Alright, I got that off my chest. Whew! (But I still don't feel any better.) :(


Don

Hi Don,

I understand what you're saying. Making meaningful readings that we can actually use in the real world is what I am talking about.

I don't know of another simple method to get a handle on when we are getting carrier shift other than one of these meters, be it voltage or power derived.  I think looking at a voice waveform on a scope is not very effective to see carrier shift.

The Bird without the PEP kit, (a common ham meter) or any other voltage derived non-peak-reading meter may not show accurate power readings under modulation, but they can give us a relative idea of when we are putting power in one direction or the other. In my case I like to adjust things for zero carrier bounce at maximum modulation. Sure, I can get the meter to bounce upwards with my old class E rig or FT-102 running 200% modulation, but I prefer zero bounce at 130% if I can.  There are still a lot of guys using diode detectors rather than sync detectors.

Out of curiousity, I remember the popular Gates mod monitor used at most BC stations in the past having a "carrier" meter on the left side.  I never owned one. Did that indicate carrier shift or was it just a calibration for the modulation %?

Here's the manual if anyone wants to dig into it:

https://www.americanradiohistory.com/Archive-Catalogs/Gates-Harris/Gates-M-5693-Modulation-Monitor-1959.pdf

Interesting discussion.

T

Here is the FCC definition for the carrier shift of an AM broadcast station (from 47 CFR §73.14):

        Carrier-amplitude regulation (Carrier shift). The change in amplitude of the carrier wave
        in an amplitude-modulated transmitter when modulation is applied under conditions of
        symmetrical modulation.

In the past the FCC had a specified limit for AM broadcast station carrier shift, and all broadcast-approved AM modulation monitors metered/displayed it.  Now, due to "de-regulation" at the FCC, AM broadcast stations are not even required to have a modulation monitor.  However they still need to meet the technical standards required of AM stations by the FCC.

Yes, that's right - the FCC doesn't require monitoring, but they have to meet the regs.

About half of the (now around 600) modulation monitors that Radio Engineering Associates has sold have gone to broadcast stations.

AM radio is getting harder and harder to listen to due to all of the RFI that the power grid and communications infrastructure (DSL and the like) is spitting out these days, but that's another discussion.  Usually I just stream the same stations, even in the car, rather than contend with all of the noise and static.

Given that there are systems which intentionally produce carrier shift to save energy (perhaps by 3dB or more without severe effects) I don't know why there's a concern about, what I would assume to be, a relatively small amount that we may encounter in our transmitters (which aren't using extreme positive peak modulation). Sure, the old controlled carrier ham rigs sounded awful, but probably for more than that one reason.

So, aside from any difficulty in measuring carrier shift, why is it a problem? Will the guy on the receiving end notice that it's there, or not there? If this is an exercise in splitting hairs, I'm okay with that. But should I really be worried about how much carrier shift my transmitter has? I've never had a "you've got too much carrier shift" report, or heard anyone else ever get one, the exception being with a controlled carrier rig like the DX-60.

All that aside, I can see how an SDR transmitter could be a different animal. Having fooled around with a Flex-1500, there seem to more ways to make it play poorly than there are ways to make it work FB (which it can). Overdrive a single stage and it goes south quickly. It took some help from Rob, W1AEX for me to get my Flex settled in. Maybe SDR's are more susceptible to carrier shift if not correctly set-up? I dunno.


Don

An interesting sidebar to this discussion is that, by itself, amplitude modulation of an r-f carrier to ±99.99... % does not change the carrier amplitude at all, even for asymmetric modulating waveforms.

All of the results of that modulation appear in the upper and lower sidebands, which occupy r-f spectra separated from the carrier by the sum and difference of the modulating frequency/frequencies.

A time-domain view of that modulated waveform has an r-f envelope amplitude that varies along the time base — which is produced by the vector sum of the (constant amplitude) carrier and the two, varying sidebands, at each instant of time.

The average value of the carrier may appear to vary in a time-domain display if the modulating waveform is not symmetric, but this effect is still produced by the vector sum at each instant of time of the (constant amplitude) carrier and the two sidebands.

Don,

I don't think it is as much of a problem as it is a curiosity that very few people understand what it is or how it could be created or how it could be eliminated.  I am one that "should" know, but just never stopped to fully examine the cause and effect.

I've worked with 50KW AMs ( and lower ) since 1976 and I must have been lucky to have had transmitters that had ample power supplies and lots of RF headroom.     

I think this is a good discussion.

I would still like to know  ( besides improper ALC on transceivers that weren't really designed for AM ) how to create and / or  eliminate *Carrier Shift*.          Even the DX-60 that you mention operating *controlled carrier* ...... maybe more properly called screen-grid modulation ?   and sounding bad ..... Is *carrier shift* associated with screen grid modulation, or just IMPROPERLY designed  screen grid modulation ?      If properly designed, why wouldn't a DX-60 sound just as good as any other modulation scheme ?   Splitting hairs ? No, I hope I'm going to learn something.

Don W4DNR

Don,

The DX-60 was screen modulated, but intentionally designed to produce a controlled carrier response. This was an attempt to thwart the low plate efficiency of the screen modulated amplifier stage at low modulation levels.

Screen modulation, by itself, does not produce a controlled carrier. Heath, and others, created methods to accomplish it. It was a rudimentary implementation that many DX-60 owners have had to undo. Yes, it worked, but it was torture to listen to. Anyone running a stock DX-60 will usually be politely informed that eliminating the controlled carrier would be a welcome improvement.

Don

isn’t this definition

“Carrier-amplitude regulation (Carrier shift). The change in amplitude of the carrier wave
 in an amplitude-modulated transmitter when modulation is applied under conditions of
 symmetrical modulation.”

just another way to define modulation percentage ?

Hey folks, interesting topic!

"Carrier Shift" isn't an abstract, nor is it always properly demonstrated by scopes or "average" wattmeters.

The "shift" nomenclature was initially created by the lack of power supply regulation in legacy transmitters (tube-type), which would include overall capacitance and the ability to respond to transients (mostly high-energy, low-end audio). The plate current would dip (or increase) based on that regulation. "Positive Swing" isn't always a good thing!. Steve, QIX has properly defined it as an  "average" power deviation which is absolutely correct. At any given time (suspended in a time-based scope or panadapter) both sidebands should be mirror images, even if asymmetrical. That would include "carrier controlled" schemes if properly implemented, at a given time.

Since the majority of my operation is with plate-modulated old-school stuff, a starting point is the plate current indication under modulation.. Any meter deviation with modulation is a flag that "carrier shift" could be occurring.

Methinks this is where using separate power supplies for the modulator and RFPA come in.

There are a number of enhancements one can make to the DX-35, DX-40, DX-60, HT-40, Knight T-60 and T-150 and similar transmitters to improve the power supply and audio.

Once the power supply is beefed up you will automatically have less carrier droop during modulation.

By modifying the speech amplifier for more linear audio and wider audio frequency response, the audio can be much improved.

And finally, by removing the controlled carrier time constant circuit and going to direct screen modulation, higher modulation percentages can be had.

Down here,

http://amfone.net/Amforum/index.php?board=46.0 ,

and here,

http://amfone.net/Amforum/index.php?board=52.0

one can find improvements for most of the SGM transmitters mentioned above.


Phil - AC0OB

There were a few questions.... "How do I get rid of carrier shift"  and  "What causes carrier shift?"

These questions are like asking "How do I get rid of audio distortion?" and "What causes audio distortion?"

There are many causes of carrier shift, distortion and non-linearity, thus many solutions. Jeff/ NBC came up with the elegant suggestion of just looking at the plate meter on a plate modulated rig.

This is all about the stability of all supplies feeding the rig, using good tubes, good iron, regulated screens and grids when needed, operating parameters adjusted right, etc. And be sure the loading cap is adjusted on the heavier side (less C) when using a linear amp or when screen modulating. I've heard more than a few BIG screen modulated rigs that sounded horrible with negative carrier shift and crunchiness under modulation until the operator simply decreased the C2 loading cap value and it cleaned up.  The Slab Bacon with the big meat a-shaking was a perfect example with his 4X1 screen modulated rig. He found that rig sounded very FB once the loading was increased as the plate efficiency was deliberately reduced. (more heat = cleaner operation)

Carrier shift is just another symptom / problem of a rig that needs work - that can be cured with a shotgun approach - by running tones thru the rig, triangle, sine - watching the spectrum analyzer and oscilloscope, sweeping, etc., to trick out the rig.   Once a rig is optimized in all areas, fine tuned and is transparent with exquisite cleanliness, the carrier shift will be gone - just like the distortion, splatter and tuning instabilities.  IE, carrier shift is a matter of degree. It can be slight and hardly noticeable, just like a bit of audio non-linearity. Or it can be really bad as the carrier varies wildly under modulation like a controlled carrier system. If there is a bad carrier shift, then there is likely other audio problems related to non-linearity caused by the same source(s). I like to look at a rig as a complete system where everything depends on everything else, like a chain.  One bad link and the problems start.

T

Please avoid having any sort of poor power supply voltage regulation 'modulate' the transmitter.

By using CCS-rated, high quality parts (Dahl, UTC, Thordarson, Amertran, et al.) having current ratings of 2-3x the maximum expected amount, voltage regulation of 2.5 to 3.1% can be got over a 5:1 to 6:1 current range. This would be a FWCT or bridge with a choke input type of filter.
Even the older, 'high-DCR' parts work fairly well in a choke input circuit when run at 1/3 to 1/2 their continuous rating.
We (hopefully) don't skimp on RF circuit parts, pay the same attention to the power!

Such a supply won't fit inside the transmitters mentioned, but could be put in a separate enclosure. Hey lots of gear has a separate power supply.

Power supplies with capacitor input filters have poor regulation and are avoided. leave them for desktop SSB amplifiers.

A while back, I put together a small amplifier that allows me to use a QRP transmitter at 100W PEP or to drive a larger amplifier. It was an RF deck from a Transworld TW-100 and the filter ass'y from a TS-440.

Running it from battery power, as I typically do in the shack, I added a large hold-up cap inside the amp to counter the DC voltage drop of the cable from the battery. I used what I had in the junk-box, but one of those huge caps that are used on mobile stereo installations would have been a nice choice too.

I believe any 12V radio installation could benefit from doing similar, whether it be mobile, in-shack or battery, power supply configured. Having a stiff supply voltage should help with carrier shift and perhaps IMD on a linear amp.

I did similar with an FT-101, replacing several hold-up caps in various parts of the circuitry, typically doubling the values. Compared to regulating voltages, adding more C is easy. Modern caps are physically smaller/uf and larger values can easily be placed in the vacated space. When re-capping an old rig, why not make it better than new?

As to the OP's original concern with carrier shift on an Icom IC7300, maybe its external power supply wasn't up to the task. I'd hate to see that radio get a bad rep for something that's out of the manufacturers control. Just sayin'.


Don

You are right.   There is a lot we don't know about the installation.  There have been numerous complaints about voltage drop across the blade fuses and the power supply cable.  We also don't know what kind of power supply is used.  We don't know the settings he is using on the radio, all of which could cause problems.

a better solution to voltage droop in dc supply situations is to use remote voltage sensing ... a well designed supply can dynamically regulate this to just a few millivolts

Steve points out that , "  I regularly modulate 150% in the positive direction, and 90% in the negative direction (at the same time) and there is no carrier shift."

Wouldn't  "asymmetrical modulation" ( +150  -90 )  be the definition of " non-linearity"  ( maybe "purposeful"  non-linearity ? )

At 150% positive modulation, I'm pretty sure that the average reading RF Ammeter  or the Bird 43 Voltage needle will increase in readings.   And I see a case  for zero 'carrier shift".       I think that would be a terrific accomplishment.   

I wasn't here to bash the Icom 7300 ....   My friend has the matching Icom power supply, but I can't speak for it's ability to supply sufficient current through it's power leads.   I just thought that "carrier shift"  was an interesting subject. 

Don W4DNR





   

As you can see Don, it is like discussing SWR.

... or PEP regarding the "legal limit".

I have only one plastic transceiver which is the Icom-7200 and use a 35 Amp Astron linear supply with the voltage set at 14.5V. No real voltage drop at max power with the supplied cables.

Again, it is the modulation method combined with the internal ALC that results in carrier shift.


Phil - AC0OB

The male voice (at most males) reproduced by a very good audio system (including the microphone) has a LOT of natural asymmetry.  

Asymmetry is quite natural and normal.  In my own transmitter, the output meter does not move under modulation, unless I modulated with such a low frequency that the meter actually responds to the modulation.  In most cases, the positive modulation is up around 150% and the negative modulation around 90%.

Non-linearity is another story, and usually DOES cause carrier shift  (often negative).

I should post some pictures.  It would help illustrate the behavior.

Yes, With commercial AM Broadcasting, we would start at the transmitter and swap output leads on every device between the transmitter and the microphone , looking for the combination ( phasing ) that produced the most asymmetry.

I worked on a Bauer 10J once that would do 110% positive 95% negative feeding it with an HP Audio Generator ... either polarity.  A 4CX15,000 modulated with a pair of 4CX5,000s.  All but one DJ could get that transmitter to modulate past 125% positive.       Later, some of the better AM processors like the Orban Optimod could even fix that.

Don W4DNR

That unit is well-worth having its schematic as one page for study.

I have an old General Radio mod monitor that is similar, as well as one of the Gates units shown in the schematic.

The design of these old units is fairly impressive, considering the limited technology available at the time.

Of course, these days we can do all of this using software and a small hardware interface to the transmitter, with much greater accuracy (particularly the meters) and at significantly lower cost.

However, it is always interesting to see what people were able to accomplish way back when, considering the limitations of what was available at the time for components.