On Reading a Scope

[WA3VJB]

So the latest issue of the broadcast engineering publication RadioWorld has an article about restoring tube-type AM transmitters. One of the points the author makes includes two images from an oscilloscope. The first shows a nice sine wave, modulating at  or just below baseline 100%.

His second image purportedly shows the ability of that same transmitter to modulate at 125%.  Okay, yes, the baseline at crossover is extended laterally, so that there's a space between the two sinewaves.  The vertical deflection is the same.  It simply looks like overmodulation.

He has since written to me that indeed, the 2nd image shows overmodulation in both directions, which apparently illustrates the transmitter is capable of being overmodulated.

He then says applying processing would limit negative peaks while allowing positive peaks to come through at greater than 100 percent.

NO image was included that actually shows assymetrical modulation.

SO, my question is still out there -- how does the ability to hit baseline measure the presence and extent of assymetrical, positive (desired) modulation in an AM transmitter ?

And what part of the second image actually shows that capability?

Is it the length of the crossover spacing at baseline, combined perhaps with the continued shape of the sinewave (not getting ugly), or some other way that I am missing from these two images.

Article here:

http://rwonline.com/article/rebuild-that-relic-of-an-am-transmitter/24672

Thanks

[K5UJ]

SO, my question is still out there -- how does the ability to hit baseline measure the presence and extent of positive (desired) modulation in an AM transmitter ?

I don't think it does.  Off the top of my amateur head, the baseline only shows whether or not you are exceeding 100% neg.

I guess if you are going over 100% pos., the pos. peaks would be higher but to see them you'd probably have to have a stable scope and sharp eye and graticule and memory of where the 100% pos. peaks had reached.  Unless you go wildly positive, like 200%. 

Speaking for myself, I have only seen the asymmetry in the positive direction at my shack by looking at my trapezoid display.   I look at where the carrier rests, the no. divisions over to the right where the negative point lands, then the no. divisions to the left where the wide open part of the positive trapezoid ends.  Any shooting over the neg. no. of divisions is excess positive peaks over 100.  He could have shown the supermodulation much better with a trapezoid as far as I know (if he wanted to use a scope).

[K5UJ]

I should have looked at the article first (I'm never going to learn to check facts first).   I love the first sentence:  Many people think old AM transmitters are worthless.  I sure wish that were true!

I see that in the second scope pix the positive peak is no higher than in the first one, so I guess I'm mistaken about what I posted a few minutes ago.  The author is Mark Persons.  Mark is a good guy.  I have purchased a number of RF items from him. 

[Steve - K4HX]

It's pretty easy to measure the percentage of positive modulation on a scope, especially using a sinewave input. It's all detailed in the link below.


http://www.amwindow.org/tech/htm/modulation.htm

[W7TFO]

I'm probably going to get a lot of flak for this view. but here goes.

Any time the envelope goes over 100%+, a normal detector will induce distortion over the inherent.

High levels of + peaks look good on a scope, but can be hard on the ears with a BA receiver.

73DG

[WA3VJB]

Yes, when I saw it was written by Mark Persons I had all kinds of ignorant self-doubt.

And Steve, yes, that's how I set up my scope, as described in your link. But the chart does not explain how to determine whether a transmitter is capable of excursions that rise higher in the positive direction than negative. It only tells me how to measure it, if the transmitter can do it.

To Rob -- exactly !  The second image does not seem to show either the potential for, or the successful result of assymetrical modulation. Maybe it wasn't supposed to.  But the discussion needed to tell me that.  Or else there might be people out there who start slamming baseline saying WHEE looka MEE  125% modulation, caw maw !

Dennis, I agree with your point but that's not the question here.  The merits of assymetrical modulation include apparent loudness up to a certain point. Increased distortion will eventually become unacceptably audible where the carrier is reduced too far in relation to sideband energy.

My question is how the second image shows the potential to achieve a judicious amount of positive modulation without hitting baseline.  I bet it's because the interval between sinewaves shows no artifacts, but I'm not sure.

[K1JJ]

My question is how the second image shows the potential to achieve a judicious amount of positive modulation without hitting baseline.  I bet it's because the interval between sinewaves shows no artifacts, but I'm not sure.

The first pic shows a standard 100% negatively modulated waveform with a 6 division positive peak p-p excursion.  This could be called a standard AM 100% modulated waveform.

However the second pic shows the carrier in cut off (>100% negative) but the pos peak excursion is STILL 6 divisions p-p.  This tells me either the positive peaks are limited or the scope's vertical input sensitivity was adjusted down.  The positive peaks should have increased beyond 6 divisions if the transmitter was operating linearly.

Either way, this example does not show the desired difference in positive modulation capability of 125%, as the author suggests.

BTW, the interval between the peaks (the thin baseline)  is simply carrier cut-off and does not necessarily give an indication of how clean and high the positive peaks are, or its positive peak capability. It simply shows that the negative peak voltage exceeds the carrier voltage in the negative direction.

I would venture to say that his negative peak limiter is phased backwards, limiting the wrong side of his audio. Maybe when he modulates with his voice, his own phase is correct. (does this make sense?)

The positive peaks do look clean in this example however, as shown by the smooth look of the sine wave.


I've seen some boatanchor AM rigs with problems that could cut off the carrier like that but flat-topped at 80% positive peak audio and sounded terrible.  The ability to cut the carrier off does not necessarily mean good positive peak audio results. It is an indication of, but not proof.

T

[WD8BIL]

I think your second suspicion is right, Tom. With the negative peaks driving the carrier to cutoff and the positive peak still rounded at the top it does show the TX could possibly be moded 100%+ to some extent.
He'd been better of leaving the vertical gain on the scope alone.

[K5UJ]

yeah good call Tom.  I never thought of that.  Last year some time I made a change in my audio chain and everything looked goofy.  the scope envelope looked strange; the trapezoid didn't look right, --I could not put my finger on it due to my inexperience but I felt something was not right.  I spent hours fiddling, adjusting this and that --no luck.  then I  happened to look at the pos and neg peak LEDs on the PMC450 and saw the Pos flashing all the time and the neg. never lighting up.  I immediately knew what was wrong and gave myself a dope slap.  Then I flipped the phase switch on the back of the mic preamp and all was well. 

the one thing about that article that bums me out is all these small station owners will read it and hang on to their tube rigs.

[KM1H]

Any time the envelope goes over 100%+, a normal detector will induce distortion over the inherent.

If you mean diode detector, most will crash at around 80% and 125% just sounds like CB

[w1vtp]

Low distortion in AM modulation depends on the ability of the modulator to reproduce the AF content that is imposed on the input of the modulation circuit. I like Stu AB2EZ (s) approach best.  Set up an dual channel scope so that both the input AF and the modulation envelop is shown such that the representative gain is the same. The resultant waveform of such a setup will be that the AF waveform will track perfectly with the modulation crest and null component.  Any departure is distortion.

The asymmetry of the human voice along with the proper phasing of the modulation in a properly adjusted transmitter will usually result more than 100% positive peaks whereas the negative nulls may approach 100%  Any artificial imposition of huge percentages of positive modulation is by definition distortion.  We would do well to set up our transmitters so that the input AF tracks the modulation component perfectly -- then we can apply our processing to increase talk power should we find it necessary.

For reduced distortion of high AM percentage, I like enhanced carrier or AM Sync detection.

Al

[WA3VJB]

Thanks, Tom for what is possibly the most likely answer -- he adjusted down the vertical deflection, obscuring the point he was trying to make.

He, the author, did write and say there was NO processing on the applied 1000hz tone, and that he increased it 2db past 100% to show 125%, as Steve's chart delineates.

But as a test for a transmitter's capability to then accept negative limiting and continue to produce 125 percent positive ?

I will wait to see if he can show how.

[k4kyv]

Looks like he's been reading or listening to too much of Ryan C. Inman's drivel:

...if the transmitter you are going to work on has PCB-filled capacitors or PCB-filled transformers, you really want to replace them before turning the power on. The same goes for mercury vapor rectifier tubes...

[Opcom]

I'm probably going to get a lot of flak for this view. but here goes.

Any time the envelope goes over 100%+, a normal detector will induce distortion over the inherent.

High levels of + peaks look good on a scope, but can be hard on the ears with a BA receiver.

73DG

Yes you are right because the negative portion of the waveform is compressed. If the classic sinewave was displayed at 100% - -we all know what that looks like, but with +125% and -100%, the negative side will start to look squished. If I'm right it will make the carrier shift positive.

Remember when super and turbo modulation was discussed as a means to increase the + modulation, some simulations were run by members showing the waveform results and they often were not pretty.

My guess to make it 'sound less bad' the negative side should not look 'clamped' to some (keep alive) value but should look as rounded as possible at its peak.

As for sounding like CB, maybe, but there is "BC programme" material such as top 40 oldies, which should be of good fidelity, and then there is "voice communications" material like hams use in which some distortion is acceptable trade for more 'talk power'. Pick yer poison!

[Opcom]

Looks like he's been reading or listening to too much of Ryan C. Inman's drivel:

...if the transmitter you are going to work on has PCB-filled capacitors or PCB-filled transformers, you really want to replace them before turning the power on. The same goes for mercury vapor rectifier tubes...

Maybe concerned about environmental and safety issues. Do we consider those components in our equipment that will be sold or inherited by the next big-iron enthusiast after we kick the bucket?
Opinions on this board are that new generations are dumbed down. Eventually, someone may own the antique who was never trained or been mentored about component safety.
A further step for conscientious owners of big rigs might be to add a sheet near the front of the manual for the piece of gear advising future persons how to care for any hazardous components. Not making them afraid of them but suggesting how to work with what is there. It is just a kindness going forward.

[KA3EKH]

That’s generally why I don’t waste too much time with trade publications, articles tend to go one of two ways, some are so technical I cannot make head or tales of them and others are so simplistic as to be a waste of time. If you’re an AM broadcaster you’re going to have a modulation monitor that will read both negative and positive peaks, you will have some form of processing in front of that transmitter that will limit negative peaks and drive positive peaks. Every AM broadcast station is required to perform annual NRSC compliance test of bandwidth and while your doing this its always better to look at modulation characteristics with a spectrum analyzer, sound wrong but I have set many AM transmitters by finding when they splatter and backing down from that point. Using a scope is a good tool but not the only tool, with many of the new digital scopes if you’re not careful you can make things worse. Until I learned here about using the audio source for the trigger on my Tektronix digital scope thought I had all types of modulation issues when using that.

[flintstone mop]

What I typically see on a 'scope from an AM station is that it is rarely close to 90% neg. And the newer transmitters can easily do 125% pos.

Fred

[k4kyv]

I have never noticed distortion from positive peaks over 100% that result from natural asymmetry of the human voice. OTOH, asymmetrical peaks resulting from deliberate distortion, such as ultra-modulation, do cause the signal to sound distorted.  I once experimented with the classic ultramodulation circuit taken from the 1956 QST article.  Most people told me the audio wasn't any louder than before, but that it sounded more raspy when the ultramod was switched in.

Peaks in excess of 100% are not supposed to contribute enough sideband power to make the signal sound louder, but their purpose is to allow enough headroom to accommodate the modulating signal without flat-topping.  Since the waveform of the human voice tends to be naturally asymmetrical, if the positive peaks are limited to 100%, that means you can't modulate at or near 100% in the negative direction without overmodulating in the positive direction.  The extra headroom makes the signal sound louder simply because you are able to fully modulate the carrier.  The portions of the waveform that lie at intermediate percentages of modulation are still what contain most of the average sideband power. Those peaks in excess of 100% positive are of short duration and contribute little to the recovered audio at the detector. Therefore, any distortion they might cause is of short duration and contributes a minuscule percentage to the total sideband power, so such distortion is generally unnoticeable if it exists at all.  The headroom peaks serve mainly to prevent even more severe distortion that would result from sharply clipping the positive peaks, and extended positive peak capability makes for a clean signal, since the modulation capability of the transmitter is not being exceeded.

The distorted sound from signals with up to 150% or more positive modulation is more likely to come from something being overdriven in the audio chain, or from excessive audio processing, resulting from the effort to push the peaks up that high, not from the peaks themselves.

If your natural voice results in 130% peaks, it is advantageous to have a transmitter capable of a little more than that to include some margin, but there is no advantage to having 150% capability in that case, and deliberately pre-distorting the signal to achieve peaks that high will result in noticeable distortion with the diode detector, both from the distortion itself, and from the fact that dwell time in excess of 100% positive has likely been considerably increased.

[K5UJ]

When I first started operating AM regularly, I ran about 200 w. carrier and up to 150% positive peaks.  I could do it driving a leenyar with a s.s. plastic radio.  Eventually I worked up the amp to where it could coast along at 300 w. and I quit doing the super high modulation.  Why?

1.  I concluded it is a pointless thing to do with a few limited exceptions:  running pretty low power, like 50 watts, or trying to work transcontinental.  But the vast majority of QSOs with 100 w. or more don't need it.

2.  Hams with old receivers have a tough time getting decent sounding audio when it goes that high.

3.  In a group of hams with 9 of 10 on frequency running typical ham rigs and the audio that goes with them, the guy who has super high positive peaks is LOUD and causes everyone to grab the RF gain or AF pot and back it down.  Having to adjust that every time the loud ham transmits is a PITA.

4.  As has been stated already peak audio isn't really important--average level with compression is important. 

I changed to more carrier/less audio with  just my natural voice asymmetry and a high average level.