Print Page - Grid modulation

Dear friends ,

I lived in Indonesia ,I have s hobby in making small MW Transmitter,which is used as gospel Radio,here we must build our own modulation Transformer,with Iron-core thats not specially made for Audio transformer,(we can't find such iron core here) so the Efficiency is low.So to modulate A transmitter using three 813 tubes ,I used Four 813 tube triode Connected,The problem is sometimes the Modulation transformer broke.

So I'm looking for different kind of Modulation,after looking in the A.R.R.L handbook.and find Grid modulation tech'nic,At a glance ,Grid modulation has a low efficency for transmitter modulation ,But there is a big difference between Control Grid (G1) modulation and Screen Grid (G2) ,because when we modulate G1, the transmitter must be operated in Linear Fashion (class B) But in Screen Grid Modulation The Transmitter is operated in class C,with reduced
Screen Voltage (half DC screen voltage) and The Plate current is half the plate current of a Class C operated telephony transmitter,so the power output is half of Plate modulated tube.

Back to the top I used three 813 for the transmitter an four 813 for the modulator,so I used seven 813 to make the transmitter,So I was thinking ,why didn't I build the transmitter with six 813 ,screen grid modulation ,so the transmitter using plate modulation and screen modulation (they were using the same amount of 813 tube) have the same power output .

The benefit off screen modulation
1.We don't need modulation trafo
2.We don't need different High voltage Power supplly (when we used modulation tubes with needs differrent plate voltages).
3 Screen Grid is the most simplle form of grid modulation

Can someone check if i missed/wrongly wrote the Conclusion I made above.

Thanks

Gito Nugroho

Gito,

I assume that your MW AM broadcast transmitter does not need to be high efficiency. I am no expert on grid modulation (never tried it), but it is my understanding that any form of grid modulation (G1, G2 or G3) is an 'efficiency' modulation, and is therefore limited to about 25-35% plate efficiency when we have headroom for 100% upward modulation. Running AM linear has the same disadvantage of low efficiency. A rule of thumb for any of these is to take the CCS plate dissipation of the final tube(s) and divide by two, and that will be the maximum carrier output you should run. Maybe then take 75% of that to keep the tubes cooler. So six 813's gives you 600 watts Pd, which is good for 300 watts carrier. De rate that 75% gives you 225 watts. This extra de-rating will give you headroom for positive peaks in excess of 100%, but will likely lower the % eff to 25% or so.

If you wanted to look at screen modulation, you will need a low impedance drive to the the screen since the loading is non-linear over the audio cycle. I would look at how Central Electronics used an 812 triode for a g2 screen supply on the 600L amplifier (single 813). Download the 600L schematic on the "Bama mirror" web page. Figure a way to adapt that to modulate the screen grid.

An alternative here is cathode modulation. Cathode modulation is a combination of plate and G1 modulation where the circuitry details dictate the percentage of G1 and plate modulation. As the % favors G1 modulation, the % eff drops to about 33%, and as the % favors plate modulation the % eff rises to about 66%. A compromise might be around 50% when the modulated power need is still quite low compared to plate modulation. The modulation transformer (if used) will require taps, and will be Low impedance, and not need extreme HV insulation like with plate modulation. In fact using a large solid state audio amp as the modulator is practical here since the turns ratio will not likely be large. Using cathode modulation might bring you up to at least 500 watts AM with the same six 813's.

Good Luck,
Jim
WD5JKO

Thank you Jim for your attention/response.

Jim ,what bothers Me,is when I read ,about screen modulation in A.R.R.L handbook(Old one),is .....In alternative adjustment method not requiring an oscilloscope ....tuned the RF Amplifier with rated d.c screen voltage for c.w operation...note the plate current,and if posible the r.f output current......then reduce the screen voltage until the plate current is one-half of the previous value . So the RF amplifier is tuned for c.w operation ,which has higher output then phone operation,when reduce the screen voltage ,where the plate current is one half of the previous setting (about 100 ma) so the input power is 200 watt.when the Rf amply is plate modulated (ccs operation) the power input is 1600 x 200ma = 320 watt,assuming the efficiency of both amply is 70% the screen modulated has 140 watt output and the amplitude modulated has 224 watt output,so if I used six 813 with SG modulation the power output is 840 watt without modulation,when screen modulation the plate voltage stay the same,so maximum plate current up swing is 200 ma so the output .7 x 2000 x 840 = 1600 watt.
with three 813 telephony (CCS operation) ,we have 0.7 x 3 x 224 watt =470 watt,with plate modulation the peak out put is 4 x 470 watt = 1880 watt.
Please correct/help me if I'm wrong.

Thank You again.

Gito.n

Gito,

When you take a class C RF amp at maximum output, and reduce the screen voltage till the plate current drops in half, the RF output drops to about 1/4 what you had before. The reason for this is that the plate efficiency is also cut in half (1/2 X 1/2 = 1/4). So lets say you had 100 watts RF out before (CW) with 65% class C efficiency. Drop the screen voltage till the plate current is at 1/2 the prior value, and the RF output will drop to something around 25 watts. If you modulate the screen now, the plate current will remain steady, and at 100% modulation you will peak 100 watts PEP. If you plate modulated the same tube at 100W carrier output, the PEP at 100% modulation will rise to 400 watts PEP.

Regards,
Jim
WD5JKO

Jim

Thank you ,for your explanation.it's much help to me,but Jim looking at 813 sheet table when it works on CW it may have 2000 vdc on the anode with 200ma plate current (CCS operation)so the output power is 275 watt(according to datasheet) ,so in SG modulation the power output is +/- 70 watt.,with SG modulation the power output is 275 watt
When it works on plate modulation we can only have 1600 vdc(CCS operation on the plate so the power output is 180 watt 100% modulation the output power is 720 watt,so the efficiency between 813 SG modulated and 813 Plate modulated is 275/720 x 100%= 38% ,Am I right Jim.

n.b. I'm sorry if my eng'lish is not good

Thanks

Gito

Gito,

I think you are pretty close. The efficiency in actual operation might be a little lower. I still stand by my earlier comment though that for linear AM, or grid modulated AM (both variable efficiency modulation), that a rule of thumb is to divide the plate dissipation of the tube by two, and with about 35% efficiency, Pd/2 = max carrier out per tube before modulation.

With my grid driven class Ab2 813 linear amp (Central electronics 600L) I can achieve 100 watts carrier output and still modulate upward 100%. I am not sure of the % efficiency, but that 813 is RED, and dumping a lot more than 125 (ICAS Pd) watts. To be more conservative, with an 813 in variable efficiency modulation (linear or grid modulated), take data sheet CCS Pd / 2 = 100/2=50 watts per tube. So six 813's give you 300 watts AM without the need of a big plate modulator, or transformer. You can get more of course, but at the expense of shorter tube life.

Do you have any 803's? These are similar to 813 except they have a suppressor grid. Suppressor grid modulation is easier to do than screen grid modulation because the loading (I think) is minimal whereas with screen grid (G2), the loading is non-linear over the modulation cycle. This non-linear loading means that you must supply the audio to G2 with a low impedance source such as a cathode follower. Remember I mentioned an 812 tube?

I still think cathode mod (50% G1 + 50% plate) is a better choice since you might get upwards of 50% efficiency and only need to have 1/2 the audio power needed for full plate modulation. I stand by my earlier statement that six 813's in cathode modulation should be good for 500 watts carrier AM. Think 1KW DC input, 500w carrier output, and 250 watts RMS to fully modulate.

Maybe some others will offer their perspective on this issue. Like I originally said, I have never played with screen modulation, or cathode modulation.

Regards,
Jim
WD5JKO

I had very good results with cathode modulation of a Class-C 250TH amp on the 40 meters band.

I used a large power transformer rated 120VAC input (this was put in series with the cathode) and 36V 10A secondary (this was connected to a big audio amp's 32 Ohm output).

Even though using a power transformer, the voice clarity and frequency response tests were very good, and we swept the audio frequency from 50Hz to 7000Hz and it was pretty flat.

Part of this was because we used a 175-watt tube type audio amplifier (Altec 1570B), but my friend later switched to a 100 watt solid state public address amplifier with the same good results. The amps had alot of impedance taps, so we had 4/8/16/32 Ohms and 25V, 35V, 70V line as well to help the matching. We did not really know what we would be matching.

Part of this result was also because the power transformer was 10X the size it needed to be. I do not suggest you need a 2500VA transformer. The one we used was merely what we had lying about. Maybe you could find a 500VA or 1KVA one around.
If using a power transformer for audio this way, it is important to use one at least 2X the required size for good audio quality and also because you will have DC in the winding and you should avoid saturation. You might want to find one with a dual 120/240 primary, so you can match it easier. Telephone companies sometimes use high-current 48V power supplies, and they have quite big transformers.

The carrier was about 100 watts, so we would only have needed 25 watts, but we chose to use an "overkill" audio amp so it would not introduce additional distortion.

We had to increase the bias and the RF drive voltage to about twice what was used for CW because the extra audio voltage in the cathode circuit would cause the bias to be cancelled out on modulation peaks thus driving the RF tube's grid-cathode circuit positive with audio.

Once we adjusted it properly the results were very good. You might face a problem finding a 250-500W audio amp so the audio quality does not suffer due to any lind of impedance mismatch between the audio amp and the RF stage cathode.

Our goal was to do this with 'junk' parts on-hand, and for low/no cost, so it worked out well.

True you may need only 250W of audio, but a bigger amp will make a good difference in the sound quality, since you are broadcasting that is important to your listeners.

Here is my 'report' on my little experiment wit my friend Dennis W5FRS and his transmitter. Maybe it will suggest some helpful ideas.
http://www.montagar.com/~patj/cathmod01.htm

Jim

It's true that cathode modulation is more efficient than grid modulation,but as I wrote before I used seven 813 ,three for the transmitter and four for the modulation,if I used all the seven 813(not six as I mentioned before) in S.G ,the power output with modulation will be 7x 275 watt=1925 watt .
using three 813 tubes plate modulated ,with modulation is 4 x 540watt =2160 watt,
so the difference is not too great.

why I used four 813 to modulate three 813 transmitter,because from my experience ,as I wrote before I have a low efficient modulation transformer ,because it's a broadcast transmitter which operated 12 to 18 hour a day ,so we don't have to used the full power off the modulator to fully modulate,so the quality of the audio is quit good,and the tubes are running "cool" .

So why S.G modulation.in the ARRL handbook it states .... the beam tetrode have very good modulation characteristic when the modulating voltage is superimposed on the d.c Screen grid voltage .... in many way screen modulation is more satisfactory than control grid modulation.

I forgot to tell you

I used fixed D.C power supply (not with dropping resistor from The Anode power supply) so to modulate it I'm figuring to use a transformer and modulate it with a small solid state audio amplifier ,from the ARRL handbook...The most satisfactory way to apply the modulating voltage to the screen is through a transformer......

some year ago my friend which have a B.C station using two 4-400 tubes run into trouble when his modulation transformer broke.So to put it on air(while repairing the modulation transformer) I tuned the transmitter at full output,than drop the screen voltage to half ,and than I modulate the S.G ,and the quality off the audio was quite good.off course the power output off the transmitter is smaller than plate modulation.

And why SG modulation,because its the simple way off modulating.

Why I m thinking to used S.G modulation ,because I used seven 813 for the plate modulation,than it crossed my mind Why not S.G modulation? so I begin to calculate the power output and find out the output has not much difference .

If my calculation is true,I don't have to used Modulation transformer for Plate modulation, but a small audio transformer for SG modulation.with a small solid state amplifier to drive it

Jim ,here in Indonesia we have difficulty in buying transmitter tubes now,so the only tube we can find here is 813,807,833,so we have a little choice to used transmitting tubes,

To Patric and Steve thank you for your advice,and guidance .

Sorry my english is not good

Thanks

Gito.N

Gito,

You might contact Patrick Turner at Turner Audio in Australia about transformers. He has a large stock of iron, and can custom design and build transformers.

http://www.turneraudio.com.au/

Remember those power levels you quoted around 2 KW are PEP power, so we're talking about 500 watts carrier.

Having seven 813's in parallel will create some unique challenges:

The Plate to Grid capacitance (Cgp) of 0.25 pg X 7 = 1.75 pf. You will likely need to neutralize even at MW

The 14 pf plate to G2, G1, and filament capacitance X 7 = 98 pf. You may need to adjust the L/C ratio in your plate tank circuit.

Each tube will need to carry 1/7th of the load. How do you insure each tube is doing it's job? Maybe have a filament transformer for each tube, and a cathode current shunt from each filament center tap to ground.

Your screen modulator will likely need to provide 5 watts RMS per tube, and since the load is non-linear, swamping the mod transformer secondary with a resistor will help. You will need an amp capable of at least 50 watts RMS, and maybe even 100 watts RMS.

Regards,
Jim
WD5JKO

Jim

thank you for telling me to re tuned the transmitter if I used seven 813,fortunately ,I used Pi-section output coupling,so it's not so difficult to change the LC ratio.
Yes it needs to be neutralize ,to stable the transmitter .

It 's true Jim how to find out if all the tubes works the same,using three 813 for the transmitter,I used 3 M ampere meters per tube to monitor the Cathode ,the Grid current,and the screen current for each tube,and its true that are differences .So actually I don't think off that before,(using seven tubes),

I admitted ,that I only calculate and try to find the difference / the efficiency Plate and Screen modulation ,at least in my transmitter .So maybe I tried one or two big tubes like 4-1000 ,if I can find such tube here.

to Patrick

Thanks for giving me /writing to me for your experience in Cathode modulation,The problem is ,you must have enough skill to tune such transmitter,and also a lot instrument/tools to monitor the operation of the transmitter.
here to find a 1000 watt dummy load(for continuous/long period testing) is out off the question.
Fortunately , I worked on medium wave transmitter,so I tried to build a dummy load from ten 100 watt 220 v ac light bulbs,I connected them in parallel ,when I calculate the resistance it was close to 50 Ohm.
So when I tested my transmitter,the S.W.R is 1/close to 1,and the lamps were glowing brightly/also showing the relative power output.
So I build transmitter using plate modulation/the easy way to modulate.
As I wrote before ,my Modulation transformer sometimes broke.So I'm trying to find another way to modulate my I transmitter,which needs to be easy.

Anyway thanks for Jim,Patrick,Steve for giving me such valuable lesson/inputs to me and it's a great help to me.

Thank you

Gito.N

Hello Gito,

If you are not in a position to do alot of adjustment, and also the people operating the transmitter every day are in no position do do alot of adjustments, then you are right to try screen grid modulation, it is simple enough to do as you described and it's low cost. I hope all those 813's will work together without oscillations.

About the stove elements, they work at low frequencies, but above a few MHz, it has more inductance since it is shaped like a coil. The oven element would be a good choice.

What is your frequency range? 530 to 1650 KC?
Also, it would be interesting to know, for our education, if your antennas are about 50 ohms or another impedance?

All:
SG modulation has been discussed, what are the pitfalls? I recall complaints about some ham rigs that used it and had poor audio. Was that just a bad design?

Patrick

I managed to bring the SWR of my antenna to 1 /nearly 1,but is the antenna efficient comparing to full design quarter wave, with full radials , that is the question?
It' a top loaded vertical using part the guy wire as the top loading.
And the impedance is 50 Ohm.

Yes ,these transmitter was operating in remote areas ,....Km(6 to 8 hour by plane from my home town },and there's no qualified technician there.
So I used the simplest /standard transmitter design,In case theres some trouble in the transmitter,I can tell them how to fix these transmitter by phone or fax.

Patrick,because I worked at low frequency ,by using light bulbs ( ten 100 watts 220 vac in parallel connection)as dummy load, I can have good SWR nearly to 1 .at least at these frequency .
And you can see the lamps glow brightly ,indication that the power from the transmitter is out and good./relative output)

Patrick ,I don't know that there are claims that S.G is not good?
But I have once using S.G modulation for two 4-400 in parallel(in emergency because the modulation transformer broke),and the result is quit good.
Maybe because I used a separate power supply for the screen,so I can change the voltage easily and modulate it through a transformer and drive it with a solid state amplifier.
In my opinion Solid state amplifier is better to drive ,as the load off the Screen varies with modulation.
Solid state amplifier can drive any load,for example you can put
an 4 Ohm speaker or an 8 Ohm speaker or an 16 Ohm speaker,and still the audio output is good ,of course with the power output differs ,so its not dependable of the load impedance ,to produce good audio,and of course the wattage must be enough
Amplifier using tubes must have a certain Impedance for good operation,like 6l6 in push-pull amplifier must have about 6 k ohm impedance plate to plate load and 4 0hm output( Output transformer) ,i if you put an 16 Ohm,or the 2 ohm speaker,The reflected load /the Tubes see is not 6 k ohm,So in my opinion Amplifier using tubes is not good for varying load,Yes you can minimize
that by loading with resistor.
So i choose to used solid state amplifier.
To S.G modulation is for seven 813 is a thought that crossed my mind,and I have to consider The + and - ,off this transmitter,how to neutralize, how to avoid oscilation.(because I used seven 813 ,for my plate modulated transmitter,3 for Rf final,4 for the modulator
Maybe I used one bigger Tetrode tube

Thanks

Gito.N

Gito,
the mod transformer that I use is from an old WW2 surplus ARC-5 "command set" transmitter. It is designed to plate modulate a pair of 1625s with another pair. (1625s are 807s with 12v filaments) It has one secondary winding to modulate the plates of the 1625s with a tertiary winding that applies around 25% of the modulating voltage to the screens. It is about 4" x 4" x 5". I used the plate mod secondary to modulate the screen of the 4-1000 and the tertiary winding to apply some of the modulating voltage in phase to the control grid bias. This eliminates the crossover distortion so often found in screen modulated transmitters, and alllows for better than 100% modulation peaks.

I did wring out the transformer to see what the turns ratio was, but it was many years ago. If I remember correctly the turns ratio for full primary to main secondary was somewhere around 2:1, with the tertiary winding being 1/4 of that. Depending on the choice of tubes you use and the voltages you run them at, it shouldnt be all that critical. As long as you apply some of the modulating voltage to the control grid bias.

If you cant fnd a mod transformer with a tertiary winding, you have to play with the control grid bias and drive levels to get it to cut off cleanly at the negative modulation peak, otherwise you will get nasty crossover distortion. Also, depending on the power level you are running, you may have to load down the main secondary winding with some external resistance (I did) to help match the screen grid impedanceto the P-P impedance of the modulator tubes. Glas all of this has been some help.

The Slab Bacon

To Patrick

I'm sorry I missed one of your Question,I worked at 1062 Khz to 1440 Khz.

To Wb6iYh

Thank you for your input Cathode modulation is good way to modulate a transmitter,but it needs a skilled person to tune it,not as easy as it seems,If you read carefully at the article you gave me,it stated Cathode modulation
has not proof satisfactory for use with Tetrode tube .....,in means that if you want to cathode modulate a Tetrode/there are many parameters that you must full fill. Jim (Wd5JKO)has told me his good result in Cathode modulated transmitter.but after reading his input/explanation ,you can tell he have good skill and knowledge to make a cathode transmitter works good.The point it it's not easy at least to me. The purpose off me here is to find an easy way to Modulate a transmitter,and how easy to repair it.
In S.G modulation circuit (4-1000) that Slab Give me,it needs only a 25 Wat audio amplifier to dive it for full modulation ( 1800 Peak power),and to tune transmitter is Easy.

To Slab

Thanks again ,what's the color of the 4-1000 plate ,does it turned red/orange red wit full modulation? did you cooled it with a fan/blower to keep the temperature down .
Because as B.C transmitter ,you must run it 12 to 16 hour a day

Thanks to You all

Gito,N

n.b sorry my eng'lish is not good

Gito,

Grid modulation is a form of "efficiency modulation" where the efficiency is around 33% unmodulated, and rises with modulation. So if the tube is RED without modulation, and you add a tone (say 1 Khz sine wave), and adjust the level upward to 100% modulation, the tube will cool down (less RED) as the static power decreases as sideband power increases. The % efficiency will approximately double at 100% modulation.

If you have audio processing to maintain a high level of modulation, then the tube will run cooler, and therefore last longer. Conversely with plate modulation the tube is stressed more during modulation, and is cooler with no modulation.

Jim
WD5JKO

Gito,
the older ARRL handbooks are a great reference source, but they are a bit stodgy and stuffy in their designs and design theory. You might try looking for some of what we refer to as the "West Coast Handbooks". They were simply called "The Radio Handbook" Written by Bill Orr. They are much more comprehensive in theory and offer many more circuit ideas than the ARRL handbooks. It would be well worth your time to try to find some of them. He went into great detail for screen and other alternative forms of modulation, and interesting antenna systems.

Also if you can find any old copies of "The Radiotron Designers Handbooks" they can also provide much useful information for you. It would be well worth your while to try to dig up either one for a reference source.

The Slab Bacon

The AM Forum

THE AM BULLETIN BOARD => Technical Forum => Topic started by: Gito on March 15, 2009, 11:14:16 PM