new project: Less common form of screen grid modulation delivers high quality au

[PA1JO]

Hi,
I start a new project trying screen grid modulation on a rig for 1395 kHz. A local music station. Legal with license to max 100-watt PEP. The aim is a transmitter with superb AM sound and 40-50Watt power in the antenna.


Fig1 shows the PA with two 6146B tubes. On point A regular screen grid modulation should apply.
I found an appealing modulator schematic. In fig 2.  Option 2 shows an adjusted schematic from a 200-Watt ARRL design published in 1967.
Point B in fig 1 (the PA) is suggested in an ARRL book dated 1964 as a small additional audio input that compensates disturbances occurring when thresh passing the zero voltage on de screen grid.
But unfortunately, I did not found a practical application of this point B audio supply.   

Fig 2 (option B) shows the modulator. 


The design is simple. The Zener diode makes sure that the screen voltage stays the same independently of screen current. The load on the transformer (12k) makes sure screen current has no effect on the audio input supply.
Have you guys experienced such a schematic and such modulation system?

Tips and suggestions would be very welcome.

[DMOD]

125V DC + 270V/2 = 260Vpeak which exceeds the max SG voltage rating so you may want to change some values.

Phil-AC0OB

https://frank.pocnet.net/sheets/079/6/6146B.pdf

[DMOD]

Hi,
I start a new project trying screen grid modulation on a rig for 1395 kHz. A local music station. Legal with license to max 100-watt PEP. The aim is a transmitter with superb AM sound and 40-50Watt power in the antenna.


Fig1 shows the PA with two 6146B tubes. On point A regular screen grid modulation should apply.
I found an appealing modulator schematic. In fig 2.  Option 2 shows an adjusted schematic from a 200-Watt ARRL design published in 1967.
Point B in fig 1 (the PA) is suggested in an ARRL book dated 1964 as a small additional audio input that compensates disturbances occurring when thresh passing the zero voltage on de screen grid.
But unfortunately, I did not found a practical application of this point B audio supply.    

Fig 2 (option B) shows the modulator.  


The design is simple. The Zener diode makes sure that the screen voltage stays the same independently of screen current. The load on the transformer (12k) makes sure screen current has no effect on the audio input supply.
Have you guys experienced such a schematic and such modulation system?

Tips and suggestions would be very welcome.

Just last month I designed for a fellow op a transformer SG modulation system for an EICO 720. The transformer was a tapped version from his junk box and reversed for this circuit. Reports were that the audio was superb.

Here are some rules we applied (per 6146 tube):

Quiescent Screen Voltage of > or = 75V,
Quiescent Screen Current of 5mA,
transformer audio SG feed voltage of at least 180Vpeak-to-peak
Transformer driving Impedance as low as feasable.

Since the op wanted variable power control, a potentiometer was added and the zener voltage was established at 100V.

Phil-AC0OB

[KD6VXI]

40 to 50 watts at the antenna is 160 to 200 pep.

You planning on 50 pct modulation?


--Shane
WP2ASS / ex KD6VXI

[K8DI]

I'm not going to comment on the screen grid modulation part of this, just a reminder on your schematic point A for audio signal to the control grid -- as drawn there's -100v, the bias, on point A.  You'll need a DC blocking capacitor there, or your audio source will be damaged or at the least pull down the bias which will put the tube into some other conduction angle up to and including one that burns it up.  A transformer will not do the trick -- it's essentially a short at DC.

Ed

[PA1JO]

125V DC + 270V/2 = 260Vpeak which exceeds the max SG voltage rating so you may want to change some values.

Phil-AC0OB

https://frank.pocnet.net/sheets/079/6/6146B.pdf

You are right Phil, but the old ARRL books say that I have to modulate on the screen 110% for reaching 100% modulation.

Then with 260 peak I do a modest job, only 4% above 250.

Can you elaborate on this??

[PA1JO]

40 to 50 watts at the antenna is 160 to 200 pep.

You planning on 50 pct modulation?


--Shane
WP2ASS / ex KD6VXI

I think you are mistaken Shane. With plate modulation you would be right but with screen modulation RMS = PEP

[KD1SH]

   An RF measuring instrument at the output of the transmitter knows nothing about how that modulated RF envelope was created. A carrier, when modulated to 100%, will show a peak of 4X that carrier, regardless of whether plate modulation, screen grid modulation, control grid modulation, suppressor grid modulation, or cathode modulation is employed.

I think you are mistaken Shane. With plate modulation you would be right but with screen modulation RMS = PEP

[steve_qix]

First of all, I like screen modulation.   But, it is very inefficient.

If you expect to get 40 watts out of a screen modulated transmitter, and also expect to have headroom for positive peaks up to at least 150% positive modulation, you won't be getting 40 watts of carrier out of a pair of 6146B tubes.

The pair of tubes represent a total of 70 watts of plate dissipation.  A screen modulated transmitter AT BEST is 33% efficient, and you adjust things to have a lot of headroom, it's even less efficient.  But, for the moment, let's stick the design at hand.

Let's figure you run the most amount of power you can with the tubes in the design.  We know that 33% of the carrier power goes into the matching network, and the rest is dissipated in the tube plates.

So, figure how much power you can run...... take the plate dissipation and divide by the percentage of power being dissipated in the plates, and you get the power input you can run.  That's 70 W. of plate dissipation divided by .6666666, equals 105 watts.

Now figure the output:  105 watts input times .3333333 = 35 watts of carrier power delivered to the matching network (which is not lossless).  So, you could get something less than 35 watts if everything is pushed to the max, and you don't have a lot of headroom for positive peaks.

If you really want ultimate fidelity, get the audio transformer out of the design.  Use something like this:

[DMOD]

I think you are mistaken Shane. With plate modulation you would be right but with screen modulation RMS = PEP

The PEP caclulation is primarily for the SSB mode. The definition of Peak Envelope Power (PEP) by the Collins Radio Corporation is; “Peak Envelope Power (PEP) is defined as the Root-Mean-Square (RMS) power developed at the crest of the modulation envelope”.

For AM, the PEP is generally taken as 4Xresting Carrier power at 100% modulation, regardless of how that AM envelope was created.

See the caclulations on pages 4 and 5. For single tone modulation the Peak Envelope Power PEP = the Average Power. For two tone modulation the Peak Envelope Power PEP = 2X Average Power.

For speech, many complex and different frequencies are present, so what is the real PEP?

http://rfcec.com/RFCEC/Section-3%20-%20Fundamentals%20of%20RF%20Communication-Electronics/04%20-%20AMPLIFIERS%20-%20RF%20POWER%20AMPLIFIER%20BASICS/RFPA%20-%20Peak%20Envelope%20Power%20(PEP)%20(By%20Larry%20E.%20Gugle%20K4RFE).pdf

Phil-AC0OB

[KD6VXI]

40 to 50 watts at the antenna is 160 to 200 pep.

You planning on 50 pct modulation?


--Shane
WP2ASS / ex KD6VXI

I think you are mistaken Shane. With plate modulation you would be right but with screen modulation RMS = PEP

I believe you are mistaken, as others have pointed out.

I'll bow out now, no need to muddle things.

--Shane
WP2ASS / ex KD6VXI

[PA1JO]

First of all, I like screen modulation.   But, it is very inefficient.

If you expect to get 40 watts out of a screen modulated transmitter, and also expect to have headroom for positive peaks up to at least 150% positive modulation, you won't be getting 40 watts of carrier out of a pair of 6146B tubes.

The pair of tubes represent a total of 70 watts of plate dissipation.  A screen modulated transmitter AT BEST is 33% efficient, and you adjust things to have a lot of headroom, it's even less efficient.  But, for the moment, let's stick the design at hand.

Let's figure you run the most amount of power you can with the tubes in the design.  We know that 33% of the carrier power goes into the matching network, and the rest is dissipated in the tube plates.

So, figure how much power you can run...... take the plate dissipation and divide by the percentage of power being dissipated in the plates, and you get the power input you can run.  That's 70 W. of plate dissipation divided by .6666666, equals 105 watts.

Now figure the output:  105 watts input times .3333333 = 35 watts of carrier power delivered to the matching network (which is not lossless).  So, you could get something less than 35 watts if everything is pushed to the max, and you don't have a lot of headroom for positive peaks.

If you really want ultimate fidelity, get the audio transformer out of the design.  Use something like this:

Please explain this Steve.

I get easily 70V in a 50 ohm dummy . That is 70 what RF power. If I can modulate this with a screen I would say I get 0,707 times 70V = about 50V RMS. In 50 ohm that would be 50 watt.

What is wrong with my thinking

Thanks in advance and have a great day,

Koos   

[DMOD]

"You are right Phil, but the old ARRL books say that I have to modulate on the screen 110% for reaching 100% modulation.

Then with 260 peak I do a modest job, only 4% above 250.

Can you elaborate on this??"
 
I would be happy to but I am not sure that's what the ARRL books say and I think you need some additional understanding of the AM mode.

In the AM mode, a carrier is modulated by some form of modulation system to form a Modulated envelope that varies in Amplitude. Frequency spectrum-wise, this method produces an RF bandwidth with two sidebands whose bandwidth is dependant on the modulating (audio) frequency.

With respect to screen modulated tetrodes or pentodes any method, whether it be transformer modulation, Cathode-Follower modulation via something like a 6DE7, or Shunt tube modulation, can produce good audio if attention is paid to keeping the modulating voltage within the screen's linear characteristics. Presenting clean audio to the modulator is also paramount.

In the DX-60 and other Novice transmitter designs, a little known fact is that they were using the RF Power Amplifier and Osc. and RF Power - Class C FM Telephony curves/specifications in order to squeeze out as much power as possible for SGM.

Again see: https://frank.pocnet.net/sheets/079/6/6146B.pdf

For the typical operating conditions the Vsg was specified as 200V with the max Vsg at 250V.

A basic premise for modulating a screen is to have the modulator's output impedance as low as possible, hence the reason the Cathode-Follower (CF) modulation system was popular since the CF system has a low output resistance of < 150 ohms.

Two 6146's in parallel will increase your power output by about X1.9.

Even with increased output power at MW it is important that you have as many ground radials as possible spoked around the base of the tower whether they be buried or sitting on top of the ground.

I hope this background information helps with understanding SGM.


What equations are you using to determine power and voltage and what intruments are you using to measure voltages, modulating percentages, and powers?

Phil-AC0OB

[KD6VXI]

If you can make a maximum of 70 volts at 50 ohms you'll need to set your carrier at 35 volts if you want to ever see 100 percent symmetrical modulation.

The method of impressing the carrier with information doesn't matter to what the final waveform looks like.  Math supports this.  Period. (if everything is working properly).

70 volts at 50 ohms is 98 watts.  35 volts is 24.5 watts.  Magic, huh...  1/4 the max power you're producing.  This is am.

Incidentally, I'd set it up for appx 30 volts at carrier.  That's 20 watts of carrier and allows for 150 percent positive peaks almost.


Taking this information and using the information DMOD gave you now choose how many tubes are necessary to not be running full bore all the time in the rf section.

Then figure out your modulator.  That's the easy part since it has been done a million times.

Pet Peave.....
Rms is a measurement of current or voltage.  There is no such thing as rms power outside of the audio world, introduced because manufacturers blatantly lied about power output in the audio amps 40 plus years ago.  Other than that marketing gimmick it doesn't exist.  You have avg power and peak envelope power, in the realm of am / slop bucket.

--Shane
WP2ASS / ex KD6VXI

[DMOD]

"I start a new project trying screen grid modulation on a rig for 1395 kHz. A local music station. Legal with license to max 100-watt PEP. The aim is a transmitter with superb AM sound and 40-50Watt power in the antenna."

What is the output power described in the station license or are you making an assumption?

Any license should state the TPO or transmitter power output in Watts with no modulation. Whether or not you get to reach your peak power is dependant on the efficiency of your modulating system.

Shane has given you some good data.

Let's look at the math for a single 6146. If you are providing 600V B+ then let's use the tube data I pointed out earlier and run the plate current at 125mA to be on the safe side, although I would most likely run at 100mA for extended tube life.

DC input Power is = 600VX0.125A = 75 Watts.

SGM Theoretical RF Output Power = DC input Power X 0.33 = 25 Watts RF maximum carrier with no modulation.

RF AC Power in Watts is = Vrms^2/R; that is, the rms voltage squared divided by the antenna resistance.

Now manipulate the simple algebraic equation and solve for Vrms.

The Voltage RMS or Vrms = SQRT(PowerXAntenna resistance).

RMS Voltage for a 25 Watt carrier into 50 ohms = Vrms = SQRT(25X50) = SQRT(1250) = 35.35 Volts RMS.

Peak AC RF voltage = 35.35Vrms X 1.414 = 50volts peak.

On an oscilloscope with oscilloscope probes across 50 ohm load, you will see 2 X Vpeak =100Volts peak to peak.

The message is when reporting voltages and currents for RF which is AC, you must give us those voltages in terms of Vrms or Vpeak, all of which can be calculated on a simple hand-held calculator.

Phil-AC0OB

[PA1JO]

Thank you Guys, especially Phil and Shane who took the trouble to understand my incorrect thinking and explain the correct thinking. 

Until next time,

greetings from Holland,
Koos

[DMOD]

Thank you Guys, especially Phil and Shane who took the trouble to understand my incorrect thinking and explain the correct thinking.  

Until next time,

greetings from Holland,
Koos

Hi Koos,

I would still be interested in the answer to these questions:

What is the output power described in the station license?

Any license should state the TPO or transmitter power output in Watts with no modulation. Whether or not you get to reach your peak power is dependant on the efficiency of your modulating system.

Undoubtedly you have a power meter in line (or should have) so what is it showing without modulation?

[PA1JO]

I would still be interested in the answer to these questions:

What is the output power described in the station license?

I'm sorry Phil, but I'm afraid I'll have to disappoint you.

The license states that we must adhere to the RED guideline. That our frequency is 1395 kHz, that the maximum bandwidth is 9kHz and the maximum power is 100 watts PEP, AM modulated signal.

Furthermore, we must not cause any disruptions or nuisance. If we use factory equipment, it must meet CE standards.

If we build a transmitter and equipment ourselves (we could do that says the license) the RED directive applies. This RED is lots of red tape. There is no heavy enforcement. Only if you cause disruption or nuisance will the installation be checked and measured.

Attached is link to the RED guideline.

https://ec.europa.eu/docsroom/documents/33162

[DMOD]

I would still be interested in the answer to these questions:

What is the output power described in the station license?

I'm sorry Phil, but I'm afraid I'll have to disappoint you.

The license states that we must adhere to the RED guideline. That our frequency is 1395 kHz, that the maximum bandwidth is 9kHz and the maximum power is 100 watts PEP, AM modulated signal.

Furthermore, we must not cause any disruptions or nuisance. If we use factory equipment, it must meet CE standards.

If we build a transmitter and equipment ourselves (we could do that says the license) the RED directive applies. This RED is lots of red tape. There is no heavy enforcement. Only if you cause disruption or nuisance will the installation be checked and measured.

Attached is link to the RED guideline.

https://ec.europa.eu/docsroom/documents/33162

Thanks Koos, no dissapointment all.

Just to set the stage, my primary interest was to compare what we call our FCC part 73 broadcasting rules to yours.  I could really care less how much power you're transmitting as I am not the spectrum police.

It sounds as if the community broadcasting rules for you are rather liberal. For the US, there are really no rules to allow for AM community broadcasting and if we do want to transmit AM, we are limited to 100mW "input" power to the transmitter and coax length and antenna sizes are severely restricted.

If your license says 100 Watts PEP then your RF output carrier with no modulation is 25 Watts or 35.35Vrms across 50 ohms or 50V peak or 100V peak-to-peak as seen on an oscilloscope. For a 100 Watt PEP with 100% modulation, you would observe 70.7 Vrms, 100V peak, or 200V peak-to-peak across 50 ohms.

That being the case, I would concentrate on antenna efficiency, such as a large number (> 45) of ground radials at least 1/4 wavelength long (~ 55 meters), the pi-net efficiency, tuner efficiency, coax length, and modulation to get as close to 100% modulation as possible. Audio quality is important as well in order to keep listeners 'tuned-in.'.  

In your second schematic, the quality and bandwidth of the audio passed through the transformer will be your limiting factor, so I would try to find a quality audio transformer with the widest possible audio bandwidth (frequency response) AND with the lowest possible impedance for the winding that feeds the screen and in series with the DC screen bias. A transformer winding that can carry 50mA should be sufficient.

And Lastly, for reliability reasons, I would suggest paralleling the 6146s plates and feeding the screens in parallel as well so each plate current is ~63mA at 600V. The resistance component of your final stage impedance is then about 2200 ohms for pi-net considerations.

Below is a suggested schematic for a Dual 6146 BC MW transmitter.

 Good luck with your community radio.

Phil-AC0OB

[PA1JO]

Thank you for the schematic and explanation Phil.

On liberal regulations: I don't know. I think it's more commercial.
It works this way:

At the beginning of this century, the Medium Wave band here was something like a deserted desert. That free space was getting occupied by Radio 'Pirates'. Pirates who didn't do anything crazy and didn't cause disruptions could often do that. The understaffed radio control service had other priorities (chasing FM pirates e.g.)

However, the music industry was unhappy about this because they received no money for the music that the 'pirates' broadcast.

Now that it has been regulated you can legally broadcast with a license. But this has become an expensive hobby. Music broadcasting rights fees have to be paid (BUMA STEMRA) and the costs are not small.

What you see is that radio stations are not making it financially because they have too little advertising income. Some stations are surviving, but you see commercial management operating there and no radio hobbyists.

Hobbyists are now fleeing to the 1600 kHz area and become pirates again. 1600kHz is an area for which no licenses are issued and which is now undeveloped.


Finally,
We are going to start a screen grid modulation project and try to get 35 watts of PEP with nice sound from two 6146Bs.

The limit of 35 watts is based on the given 35% efficiency at 17.5 watts output per tube, which means that 50 watts of input power per tube is necessary.
The Plate dissipation (Pp) per tube is 50-17.5/1.1 = 29.5 watts. (Pp for FM is max. 27 CCS and max. 35 ICAS).
We expect that we can use the FM conditions as a reference because with screen grid modulation, unlike with Plate modulation, we do not increase the plate voltage through modulation.

By the way, we still have to recover from the disappointing knowledge that with Plate modulation we can produce up to 70 watts of PEP with an 6146B and with screen grid modulation only 17.5 watts.
Plate modulation: Pi = 70/0.75 = 93 watts. Pp = 93-70= 23/1.1 = 21 watts (CCS max = 27 watts)
Okay, that is how it is.

Have a great day

[DMOD]

Finally,
We are going to start a screen grid modulation project and try to get 35 watts of PEP with nice sound from two 6146Bs.

which means that 50 watts of input power per tube is necessary."

No, and where are you pulling these numbers from?

I don't know what you said above but I assume you didn't believe anything stated in the previous post?

As explained above, for both tube's the total DC input power is 600VX0.125 mA = 75 Watts or 37.5 Watts per tube, so taking into account the efficiency 0.33X75W = 25 Watts resting carrier for SGM. As explained above you get 100W PEP with 100% modulation, which you said was stated in the license.

But Do what you like.

[PA1JO]

No, and where are you pulling these numbers from?

Excuse me Phil, The calculations come from "The Radio Handbook, William I. Orr, W6SAI" page 15-8.

I applied them on your schematic and saw it was only 25 Watt. My radio friends did not like that.

That the resting carrier 25 watts would meen 100 watts PEP didn't dawn on us yet. But I see you said it in your explanation. Shame on me.

Even worse, our current so-called 30 watts transmitter produces 120 watts PEP.

We all been trained and did exams with this (I see now it is a faulty) rule. https://www.iwab.nu/007_028.html

We were confused.

Thanks for your tenacity :-)

[steve_qix]

First of all, I like screen modulation.   But, it is very inefficient.

If you expect to get 40 watts out of a screen modulated transmitter, and also expect to have headroom for positive peaks up to at least 150% positive modulation, you won't be getting 40 watts of carrier out of a pair of 6146B tubes.

The pair of tubes represent a total of 70 watts of plate dissipation.  A screen modulated transmitter AT BEST is 33% efficient, and you adjust things to have a lot of headroom, it's even less efficient.  But, for the moment, let's stick the design at hand.

Let's figure you run the most amount of power you can with the tubes in the design.  We know that 33% of the carrier power goes into the matching network, and the rest is dissipated in the tube plates.

So, figure how much power you can run...... take the plate dissipation and divide by the percentage of power being dissipated in the plates, and you get the power input you can run.  That's 70 W. of plate dissipation divided by .6666666, equals 105 watts.

Now figure the output:  105 watts input times .3333333 = 35 watts of carrier power delivered to the matching network (which is not lossless).  So, you could get something less than 35 watts if everything is pushed to the max, and you don't have a lot of headroom for positive peaks.

If you really want ultimate fidelity, get the audio transformer out of the design.  Use something like this:

Please explain this Steve.

I get easily 70V in a 50 ohm dummy . That is 70 what RF power. If I can modulate this with a screen I would say I get 0,707 times 70V = about 50V RMS. In 50 ohm that would be 50 watt.

What is wrong with my thinking

Thanks in advance and have a great day,

Koos   

No doubt you can get 70V into a 50 ohm load.  The problem comes about when you start modulating.  If you want good, LINEAR (clean) modulation, you HAVE to leave headroom in the design and the adjustment of the transmitter.  Otherwise the audio quality will suffer, and you also the positive modulation will be limited - perhaps to less than 100% before distortion starts showing up.

The more headroom you leave, the less carrier power you will be able to run without exceeding the tube ratings.

Everyone is different.  Personally, I don't design or use anything (at least on a regular basis) that cannot modulate to at least 150% positive modulation without any waveform distortion or clipping.

A nice way to test a transmitter is to use a 400Hz triangle wave.  If there are any transmitter problems, they will show up immediately when you look at the modulation envelope on an oscilloscope or modulation monitor.  Ideally, the slop of the triangle waveform will be a straight line up and down.

[PA1JO]

No doubt you can get 70V into a 50 ohm load.  The problem comes about when you start modulating.  If you want good, LINEAR (clean) modulation, you HAVE to leave headroom in the design and the adjustment of the transmitter.  Otherwise the audio quality will suffer, and you also the positive modulation will be limited - perhaps to less than 100% before distortion starts showing up.

The more headroom you leave, the less carrier power you will be able to run without exceeding the tube ratings.

Everyone is different.  Personally, I don't design or use anything (at least on a regular basis) that cannot modulate to at least 150% positive modulation without any waveform distortion or clipping.

A nice way to test a transmitter is to use a 400Hz triangle wave.  If there are any transmitter problems, they will show up immediately when you look at the modulation envelope on an oscilloscope or modulation monitor.  Ideally, the slop of the triangle waveform will be a straight line up and down.

Thanks Steve, we start a new design

[DMOD]

"Excuse me Phil, The calculations come from "The Radio Handbook, William I. Orr, W6SAI" page 15-8."

What was the application and equations?

"I applied them on your schematic and saw it was only 25 Watt. My radio friends did not like that. That the resting carrier 25 watts would meen 100 watts PEP didn't dawn on us yet. But I see you said it in your explanation. Shame on me.

Even worse, our current so-called 30 watts transmitter produces 120 watts PEP.

We all been trained and did exams with this (I see now it is a faulty) rule. https://www.iwab.nu/007_028.html

We were confused.

Thanks for your tenacity :-)" [/quote]

Do you have a schematic of your so-called 35 Watt transmitter that you can post as a PDF? Is that 35 watts carrier with no modulation?

Your friends didn't like 25W carrier and 100 Watts PEP? But that is what you said your license required.

The schematic I presented is from a working transmitter so I know its capabilities and the audio is superb.

"I get easily 70V in a 50 ohm dummy . That is 70 what RF power. If I can modulate this with a screen I would say I get 0,707 times 70V = about 50V RMS. In 50 ohm that would be 50 watt."

NO, you're still not understanding what was previously explained to you.

70V what, Peak-to-peak, RMS, or peak? You have never stated the units. What did you measure this with; oscilloscope, some kind of meter, what? Please tell us these things because they are important if you want help with this and you were asked these questions  previously but have never answered them.

Let's take 70V peak-to-peak: 70Vp-p/2.818 = 0.355X70Vp-p = 25Vrms.  Power = Vrms^2/R = 625/50 = 12.5 Watts carrier no modulation.

Let take 70Vrms: Power = 4900/50 = 98 Watts carrier.

SO what is it? 70Vp-p or 70Vrms?

Phil - AC0OB