The AM Forum
March 28, 2024, 08:51:01 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Russian Tubes For Series Mod.  (Read 4012 times)
0 Members and 1 Guest are viewing this topic.
KD1SH
Member

Offline Offline

Posts: 737



« on: March 20, 2019, 07:19:09 PM »

  I've been thinking about series modulated rigs - big ones in particular.  A while back there was some talk - see  http://amfone.net/Amforum/index.php?topic=44700.0 - about throwing dissipation to the wind (pretty much literally) and just blowing off the excess heat.
  I had a notion once to build a legal limit series modulated rig, got all excited about it, and then got out my calculator and punched in some numbers.  Calculators can be such killjoys.
  It's not that tubes with gonzo plate dissipation are made of unobtanium - they're definitely available - but for a modest budget operator like myself, they're made of unaffordium!
  And then I was looking at  http://qro-parts.com and got to wondering if anyone has considered using some of those big Russian tubes in a series modulator.  A tube like the Svetlana 4CX400A would seem like quite a bargain at $99.50 for a modest rig, and a pair of GU-74B's can be had for $399.50, giving you 1600 watts of plate dissipation.  Usually when I think of these tubes I'm thinking of RF PA applications, but is there any reason why they couldn't be used at audio frequencies?
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #1 on: March 20, 2019, 10:07:02 PM »

Hi Bill -

If price is the main factor then the 3CX-2500F3  (F3 = fil leads, low mu) is a very common triode tube that was used in many BC transmitters over the years, as both modulators and RF finals.  Two would make a great final and series modulator. Put as much HV on them as you can and a KW carrier will be no problem.

They are plentiful and usually found at cheap prices. I even have a pair here with chimneys that have been in many rigs over the years. They last so long a used pair should be FB. I know there are some guys on this BB who would part with a pair for you.

I've built a PDM rig with a pair as well as a plate modulated rig with them. Even built a linear amp using them years ago before I started using tubes designed for linear service...  

They are low mu triodes that would need neutralization in RF service.  The series modulator could be driven nicely with the GFZ audio driver with some driver board range widening. No screen supplies to float or fool with. Just add a fixed protective and grid leak bias and you're golden. Maybe a zener in the fil CT could be the fixed bias OR maybe none is needed due to the series tube's constant control.   (The fil current at 51A / 7.5V is stiff though)

They are indestructible work horses with a plate diss of 4KW. You will NEED that with a series modulator. :-)   This is a case where the "2500" nomenclature doesn't match the diss.

https://www.cpii.com/docs/datasheets/79/3CX2500A3-8161.pdf

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
w8khk
Member

Offline Offline

Posts: 1200


This ham got his ticket the old fashioned way.


WWW
« Reply #2 on: March 20, 2019, 10:45:57 PM »

Tom and I were typing at the same time, so I went back and read his post.

William, I will share some of my thoughts on the development of a series-modulated heavy-metal rig.  I would be interested in hearing of your plans as well.

Thanks Tom, for saving me lots of typing!  Yes, I am using a mix of 3X2500F3 and 3CX2500F3, basically the same tube, one glass, one ceramic.  The nice thing about the flying leads for filament and grid connections is that you do not need any expensive socket.  Just cut a hole in a sheet of G-10 fiberglass (circuit board material, with no copper foil) and blast some air through it.  Make sure you cool the seals around the grid wires.  

Filament power  - especially the current, is extreme, but that equates to enormous cathode emission, a welcome friend on both the series modulator and the modulated final stage!

In addition to massive dissipation, you must also have a tube that can withstand a great deal of voltage between cathode and plate.  This is actually more important than the dissipation rating!  Your power supply will be typically 2.5 to 3 times the voltage you would normally use on a plate-modulated rig, and the series modulator drops it down to the nominal voltage, allowing the final supply to swing from zero to more than double the nominal voltage, allowing headroom for positive peaks greater than 100 percent modulation.  

As you plan your rig, you want to think carefully about the topology.  Is the modulator going to be above the RF final, (between the positive power supply lead and the plate circuit), or below (between the RF final cathode circuit and ground)?  Above works as a cathode follower, and you will  need to swing the modulator grid very high, using another transmitting tube as a driver.  If the modulator is "below" the final, it still performs as a plate modulator, NOT a cathode modulator.  The plate of the modulator tube will then be the effective DC return point for the final, and thus grid and cathode circuits are returned to this point instead of ground.  

If the RF final is above the modulator, link coupling could be used for the input grid drive circuit, and the link winding can be formed in a way that gives the required HV DC isolation.  Standard Pi or Pi -L can be used in the plate circuit.

One of the greatest challenges will be driving the series modulator in a manner that provides a linear amplification of the audio waveform, staying in the linear portion of the characteristic curve of the tube.  This is another reason to plan for ample (HIGH) power supply voltage.  Since you will not see any significant phase shift or distortion that is common to modulation transformers, it is possible to add feedback to the driver, sourced from detected audio from the RF output.

I have made arrangements to test with several 810s in parallel for a modulator, and a couple more 810 triodes for the RF stage.  This will run at much reduced voltage from the same 5KV 1A CCS supply, throttled down using either 120 instead of 240 volts, or a Variac, or both.  I will start with just the modulator, and a string of 40 or 60 watt incandescent light bulbs to simulate the RF deck as a load.  This way I can test the modulator and driver circuits, evaluate linearity, headroom, etc, then plan the RF deck accordingly.  When I am happy with the topology, I will scale it up to the full KW input capability with the pair of 3X3000 tubes.  

As an aside, I decided to limit my project to triodes for the mod and RF, mainly to avoid the additional complexity of managing screen supplies, and the challenge of properly modulating the screen voltage.

I have thought about using grounded-grid in the final (even though it would result in incomplete modulation) to avoid the need to neutralize, but will probably end up using a grounded cathode in the RF final.

One very time-consuming phase is expected to be the cooling system.  Variable speed squirrel cage DC blower, with insulation-lined duct-work.  Input air will enter the cabinet under the power supply through many small vents, muffling the intake draft.  Several labyrinth baffles will quiet the exhaust airflow.  All this comes after open-air breadboard testing.  This will be a rather long development process to make it shack-friendly, but I really wished I had it this past winter.  Looking at my chart of heating gas invoices, this electric heater will be a welcome appliance, focusing the heat where it can be enjoyed for hours on end!

Regarding tube availability, I have seen the 3CX3000F series (good broadcast pulls) go for 250 to 300 dollars each, much more economical than the Russian triodes or tetrodes.

Logged

Rick / W8KHK  ex WB2HKX, WB4GNR
"Both politicians and diapers need to be changed often and for the same reason.”   Ronald Reagan

My smart?phone voicetext screws up homophones, but they are crystal clear from my 75 meter plate-modulated AM transmitter
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #3 on: March 21, 2019, 12:12:51 AM »

The 3cx3000 is an amazing tube.  I've successfully run them at 9kv (unloaded, 8500 volts loaded).

Taking that into account, you should be able to run it in a series mod netting you 4.xx kv modulated.

The 3cx3000A7 hi pots to over 18kv, the limit of the old tester.

The cathode is pretty amazing as well.

I look at the 3000 as a ceramic sweep tube with a LOT more dissipation.  The power levels I've seen doing stress testing would make your eyes pop!  At 6kv they REALLY sing.  I have posted videos of my 3000 projects here and elsewhere (YouTube)


--Shane
KD6VXI
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #4 on: March 21, 2019, 01:29:37 AM »

Another informative post, Rick.

Yep, I agree with you both that the 3CX-3000 family is a great tube, though very expensive these days. What are they going for now-a-days?  That's the only reason I would go with the very cheap, used 3CX2500. More effort to get working, agreed.  I'll bet a good used 2500 could be found in a box at the flea mkts for under $30 each.  The 3X is the glass base and the 3CX is the ceramic base. Go with the 3CX because it is a much newer version.

Another good tube, which I also have is the YC-156 / YC-179  which is really a 3CX-5000B7. No socket required - just bolts to the chassis with the grid flange. They were once plentiful as MRI pulls but have dried up due to MRI machines going all solid state. They used to cost about $300 each as used pulls.  Possibly much more now.  They are a VERY linear tube, maybe the best out there in the whole whirl. The fil is indirectly heated so that the heater requirement is zip, like an 8877.

My dream choice might be a pair of YC-156s in RF / series mod with only about 5KV on the whole thing.  They don't need a lot of voltage, unlike the 3000/2500 series, to make serious power. Amplification factor of 200, like an 8877. What a tube.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #5 on: March 21, 2019, 02:24:38 AM »

Tom,

New 3cx3000 tubes are about 875 or so.  Used / pulls can be had for less, but of course you're rolling the dice there.  Getting one rebuilt is also an option, but it's about the same price as new.....  Depending on your own relationship with econco.

They generally come in around 50 bucks more than an 8877.

The Yc156 is pretty much unobtanium now.  I've seen people spend a grand on one.  Rediculous if you ask me.  Of course, they are easy to drive and have superb imd ratings.  If you really look around, you can find other B7 variants, up to 15kw (the Yc156 is a 3CX5000B7).

The 3cx3000A7 also does not require a socket.  A 3 inch chassis punch will captivate the grid flange.  4 to 6 8x32 screws hold it in place, I ground a washer with a flat side to hold the tube down.  The center pin collette can be made with the outer shell of a PL259 connector with a penny silver soldered inside.  If you get it hot enough to melt the silver solder, you lost cooling air or something else.

For chimneys a trip to home depot can solve that.

But, if you can get flying lead 3cx2500 tubes that cheap its a no brainer.

Time for bed.

--Shane
KD6VXI
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #6 on: March 21, 2019, 03:33:23 PM »

Tom,

New 3cx3000 tubes are about 875 or so.  Used / pulls can be had for less, but of course you're rolling the dice there.  Getting one rebuilt is also an option, but it's about the same price as new.....  Depending on your own relationship with econco.

They generally come in around 50 bucks more than an 8877.

The Yc156 is pretty much unobtanium now.  I've seen people spend a grand on one.  Rediculous if you ask me.  Of course, they are easy to drive and have superb imd ratings.  If you really look around, you can find other B7 variants, up to 15kw (the Yc156 is a 3CX5000B7).

But, if you can get flying lead 3cx2500 tubes that cheap its a no brainer.

--Shane
KD6VXI


Good info, Shane.

Wow, I've tripled my investment return in YC-156s if they now sell for $1K each... :-)

It's amazing that the YC-156 / YC-179  (3CX-5000B7) has the same cathode/filament structure as the 3CX-15,000B7. Just that the anode/plate structure is bigger on the 15K. And the fil takes so little power and is EZ to drive. Super clean IMD, suited to exquistite pulse reading MRI machines. They are the pinnacle of tube development.  Just like the R-390 was for tube receiver technology development.

It appears the 8877 can take the required grid current to run in hard class C.     2400V plate modulated = 1KW out. 45 Ma grid current.    

Here's an 8877 data sheet that talks about class C, plate modulated, but using cathode drive.  The YC-156 grid is rated only 35 watts, so may be linear operation only.   The 8877 is 25 watts.

www.tubecollectors.org/eimac/archives/3cx1500a7(75).pdf


A new YC-179 for $2K...
https://mgs4u.com/product/eimac-yc-179/


T


Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #7 on: March 21, 2019, 04:09:51 PM »

Their are quite a few tubes that 'share' structures.  It would be interesting to see just how many do.

Also, if you increase the air holes on a 3x3 chassis and use a larger blower you can also increase PDiss.  I believe VE7RF did some tests and was able to increase the dissipation of the 3x3 to 6kw and the 3x6 to (iirc) 9 kw.  Verified with tempalac or however its spelled.  I know a 3x3 will do 15kw out the snout. I'm sure imd suffers, but predistortion! Lol.

He used a square chimney made of etched pc board (leave edges with copper to solder seal it into a square box).  That gives you enough space for 3/4 inch holes instead of the Itty bitty holes you have to use with a stock chimney.  Then the only backflow is the tube, instead of tube and chassis.

Anywho.  Yup, you made some dollars on the yc tubes.

I contemplated building a personal amp with one but then found a forgotten 4x5k and socket.  Maybe pwm with a 3cx3000A1 KD0HG sold me years ago...

So many tubes, so little time.

--Shane
KD6VXI
Logged
w8khk
Member

Offline Offline

Posts: 1200


This ham got his ticket the old fashioned way.


WWW
« Reply #8 on: March 21, 2019, 10:51:12 PM »

The 5 kilovolt 1 amp CCS supply was completed long ago, and is awaiting the opportunity to inconvenience a multitude of electrons.   

While I have had the requisite 3CX3000 bottles for some time, I have been awaiting delivery of several 810 triodes for the lower power prototype.  I have received word that they are now on the way.  I still need to make a decision on the HV driver tube for the prototype modulator testing.

I will put the modulator and final deck on a single pine board, or possibly two pine boards stacked.  This will not need anything more than convection cooling, and it will all be out in the open so some pictures of the rig should be outstanding and as interesting as some of the "Clip Lead Special" rigs seen on this forum previously. 

Too bad there will be no glowing 866s or 872s.  All solid state power supply this go-round, my friends.  But several big bread-slicers will grace the rustic wooden chassis.  And maybe some coiled copper tubing to bring back the memories of prohibition and the still in the back woods.......

As I have done in the past, I will provide photos as the project approaches fruition.
Logged

Rick / W8KHK  ex WB2HKX, WB4GNR
"Both politicians and diapers need to be changed often and for the same reason.”   Ronald Reagan

My smart?phone voicetext screws up homophones, but they are crystal clear from my 75 meter plate-modulated AM transmitter
KD1SH
Member

Offline Offline

Posts: 737



« Reply #9 on: March 22, 2019, 07:04:19 PM »

   Wow - tons of great info here - thanks to all for coaching a relative AM noob.  My high powered TX project is far from shovel-ready at this point.  Don't even have the power-iron to feed such a beast right now.  I do have a power supply that will put out around 3500 VDC @ at least 1.5A; I suppose I could run that big transformer into a doubler.

   Right now, just to get my feet wet (bad analogy - who works on high voltage gear with wet feet?) I'm gathering parts to start off with a much lower powered series mod rig, something along the lines of the WB9ECK 807/6LF6 rig.  Might try it with 6146's, though, because sweep tubes aren't the plentiful bargain they used to be.

   Some of my basic assumptions, possibly erroneous:

  Series mod tubes could be configured in a series "stack" to get more voltage handling ability, but biasing and driving them would require some fancy footwork, because all the cathodes would be floating at different potentials.

  Putting the series mod tube below the RF PA and letting the cathode "sink" through it would seem to present less trouble in biasing and driving the modulator than putting the mod tube above the RF PA and "sourcing" through it.  Seems in the latter topology you'd be obligated to "float" your driving circuitry.  I've never liked floating stuff; only boats should float.

   On the other hand, putting the series mod tube below the RF PA would, I think, make using a fixed/regulated bias source - to eliminate the need for a clamp tube or some other such drive-loss protection - more problematic, since the cathode of your RF PA would be jumping all over the place with reference to ground.

  I can think of a ton of noob questions - some maybe inane - but you guys have given me a lot of great info to build on.  Thanks -


Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
w8khk
Member

Offline Offline

Posts: 1200


This ham got his ticket the old fashioned way.


WWW
« Reply #10 on: March 25, 2019, 01:57:42 AM »

As you start to work on your series modulated rig, do not push it up to 3000 volts or more.  Start with 1000, and work up to 2000.

It does not matter if you "float" the RF or the modulator section.  Either one has its complications, but none are insurmountable.  

I will start with the series modulator ABOVE the RF deck, and use a standard RF configuration.  I will do triodes, so that no clamp or screen supply is required.  With the modulator between the HV B+ and the RF deck, you simply have a cathode follower.  The grid must swing the entire range from no plate voltage to plate voltage times two, plus a bit extra for positive peaks.  The good thing is that the cathode follower will not draw grid current, so it is easy to drive.  A grounded cathode transmitting tube that can withstand high voltage, using a plate resistor to the HV, can be directly connected to the grid of the series modulator.  Bias the grid of this driver tube to provide the desired resting plate voltage for the RF deck, and vary it up and down with modulation.  Look at the transfer curves to select a tube that will give you linear response with the audio input.  The only challenge with floating the modulator is the need for a filament transformer that can handle the extreme voltage swings - it must have good isolation/insulation, and it is best if it has low capacity between the filament winding and the primary and core.

If you float the RF deck and put the modulator between the cathode and ground, the only concern is the HV insulation and isolation between the components and ground.  Main challenge is again the filament transformer for the final, and possibly a clamp tube.  Again, this is where I prefer a triode for circuit simplicity.  Using link coupling for the RF input provides driver isolation.  Pi or Pi EL network for output provides isolation via the plate blocking capacitor between the plate and the Pi tank.  All circuits that would otherwise go to ground will be connected to a common point and to the plate of the series modulator tube.  Insulating couplings for the tuning controls, and grounded panel bearings for the knobs, and YOU are now isolated from the floating ground on the RF deck.

The key to having a clean, linear setup is to review and study the tube curves, decide on an approximate operating point, and test before putting all of it together.  I plan to fully test the modulator before starting on the RF deck. I have not yet decided whether to put the modulator or RF deck on top.  If you put the modulator on the bottom, you could easily use a screen grid tube here.  To test, I will substitute a string of incandescent light bulbs  (20 or 25 each sixty watt bulbs to test the modulator for the KW input rig) for the RF section.  That way, the steady-state dissipation and voltage drop across the modulator can be determined.  Then a scope can be attached via a voltage divider and a high-voltage probe, to look at the modulation linearity.  The modulator driver, and possibly feedback, can be designed and tested using the light bulb string.  Once this is finalized, the RF deck can be inserted in place of the light bulb load.  Remember the light bulb load will be seeing only DC and audio modulation, not RF, so the only issue here is that the light bulb steady-state dissipation equates closely to the DC input E and I from the modulator, and the audio then modulates the light bulb signal.  

I plan to test at lower levels, using about 2000 to 2500 volts, with one or more 810 triodes for the modulator, then several more triodes for the RF deck.  I will use "pine-board" breadboard construction, because it is not permanent.  This configuration will allow me to test the audio driver circuit, and I will probably use a tetrode or pentode sweep tube here, or maybe a 4-65 or 4-125 for the high-end testing at higher voltage.  I may even need to go to a 4-250 or 4-400 just to get the driver gain and voltage capability for the high end modulator.  The screen grid tube will provide much more of the gain needed than would a triode for the driver.  

When I scale it up to the 3CX3000 bottle, I will likely run 5KV on the modulator, and shoot for around 2 KV at 500 ma, or 1 KW DC input on the RF deck, allowing for positive peaks greater than 100 %.  The final will likely be a classic grounded cathode amplifier, with link coupling on the input, and pi output feeding an antenna tuner.  No vac caps, just bread slicers.  Lots of air with a carefully designed plenum and insulation/baffles to muffle the air noise to a minimum.  A DC squirrel cage blower from an automotive air conditioning system, with temperature feedback to adjust for the required airflow will be implemented before the 3CX3000 tubes are used.  

Hopefully this will give you some insight as to how to get started with your design, testing the various sections before settling upon a final design.
Logged

Rick / W8KHK  ex WB2HKX, WB4GNR
"Both politicians and diapers need to be changed often and for the same reason.”   Ronald Reagan

My smart?phone voicetext screws up homophones, but they are crystal clear from my 75 meter plate-modulated AM transmitter
KD1SH
Member

Offline Offline

Posts: 737



« Reply #11 on: March 26, 2019, 06:08:52 PM »

  Good advice all the way around.  I've been copying all the great info from this thread into a Word doc in my "series mod" folder for future reference.
   A few questions: 
   Link coupling seems like a good way to maintain isolation between driver and PA, but couldn't we get the same isolation by capacitive coupling?  Obviously your blocking capacitor has to withstand whatever potential difference exisits.  Would link coupling, assuming proper tuning, give better energy transfer?  Maybe better harmonic attenuation?
   As far as floating the RF deck - since we're using a blocking capacitor between the plate and the tank circuit, do we still need to isolate the tune and load caps?  If we isolate the tank, does that imply that we're also tying the shell of the output coaxial connector to that same common point, and keeping it isolated from chassis ground?  In my station, all coax runs are grounded to a common bus here in the shack and also out at the base of the tower where my lightning arrestors are located, so that as soon as I connect my coax to the rig I'm grounding that floating tank circuit. 
   Also, if we assume that the RF voltage generated by the PA tube is actually appearing between the plate and the cathode of the PA, and we bring our tank circuit common "floating ground" point back to the plate of the series modulator, how can this be a "return" path for RF, since there's going to be an RF choke between the plate of the series modulator and the cathode of the RF PA?  Wouldn't there have to be an RF path back to the cathode, not just DC?
   Looking at the schematic for the WB9ECK 807 series mod transmitter, Mr. Stout shows the plate and load "bread slicers" tied right to chassis ground, as well as the shell of the coax connector.  Of course, this is a much lower voltage/lower output transmitter than what we're talking about here.
   Sorry if these seem like dumb questions - I'm still looking at the AM world through my wet-behind-the-ears noob lenses.  So much to learn, so little time.



If you float the RF deck and put the modulator between the cathode and ground, the only concern is the HV insulation and isolation between the components and ground.  Main challenge is again the filament transformer for the final, and possibly a clamp tube.  Again, this is where I prefer a triode for circuit simplicity.  Using link coupling for the RF input provides driver isolation.  Pi or Pi EL network for output provides isolation via the plate blocking capacitor between the plate and the Pi tank.  All circuits that would otherwise go to ground will be connected to a common point and to the plate of the series modulator tube.  Insulating couplings for the tuning controls, and grounded panel bearings for the knobs, and YOU are now isolated from the floating ground on the RF deck.


Hopefully this will give you some insight as to how to get started with your design, testing the various sections before settling upon a final design.
Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
w8khk
Member

Offline Offline

Posts: 1200


This ham got his ticket the old fashioned way.


WWW
« Reply #12 on: March 26, 2019, 10:41:43 PM »

The output Pi network is coupled from the plate circuit by a blocking capacitor, and the B+ is applied in shunt feed by the RFC, so there is no need to float the tuning and loading capacitors, they should be grounded to the chassis.  It is important to have the RF circuits referenced to ground, while the DC supply is floating.   Don't forget the safety choke in parallel with the loading capacitor!

The RF choke between the plate of the modulator and the cathode of the RF Final may not be necessary.  The RF is effectively grounded at the cathode by the bypass capacitor at the top of that RF choke.   This capacitor value should be chosen to have a very low impedance (reactance) to RF at the lowest (RF) operating frequency, but it should not be so large as to attenuate the audio frequencies, or the highs will be rolled off.  We want to avoid this, or why would we use series modulation instead of Class-B and a modulation transformer?

The input RF drive may be either inductively or capacitively coupled, it matters not.  I chose link coupled, and the only reason was convenience.  When using a triode, it is necessary to neutralize, and the link-coupled input with a split-stator capacitor addresses this function.  Only the DC operating potentials (grid, cathode) need be returned to the common point at the top of the modulator.  This is what differentiates plate modulation (series) from cathode modulation.  It does not matter whether the modulator is above or below the final, as long as we are modulating the plate voltage to that stage.  A modulator in series with just the cathode is simply cathode modulation.

In summary, the only challenges in floating the final are the insulation ratings of the filament transformer, tube base/socket, and the components employed in the grid bias circuit.  The DC isolation must also be provided in the RF coupling, either by link coupling with sufficient spacing for the MAXIMUM DC potential applied to the plate circuit, or capacitive coupling meeting the same criteria.  I believe link coupling would be easier to accomplish, considering the potentials involved.

That said, I plan to use a standard grounded RF deck, and run the series modulator as a cathode follower, with a resistance capacitance coupled driver stage, achieving the required bias, and audio swing on the grid, which will never draw grid current.  A high-voltage tetrode transmitting tube should perform well here.
Logged

Rick / W8KHK  ex WB2HKX, WB4GNR
"Both politicians and diapers need to be changed often and for the same reason.”   Ronald Reagan

My smart?phone voicetext screws up homophones, but they are crystal clear from my 75 meter plate-modulated AM transmitter
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.063 seconds with 18 queries.