The AM Forum
March 29, 2024, 09:56:25 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: "AC-DC Filter Choke" in Cathode Modulator  (Read 15041 times)
0 Members and 1 Guest are viewing this topic.
Carl WA1KPD
Contributing
Member
*
Offline Offline

Posts: 1629



« on: July 22, 2014, 11:01:53 PM »

Building the cathode modulator below from QST into a very sad, abused looking Glob Chief I got at Nearfest. Pair of 807s. The circuit calls for L1 to be an "AC-DC Filter Choke" (?). In the text they refer to it as an audio filter choke. Will a small power supply choke work?
Thanks
Carl
/KPD

* SimplestModulator.pdf (905.5 KB - downloaded 422 times.)

* Untitled.jpg (156 KB, 1080x460 - viewed 1025 times.)
Logged

Carl

"Okay, gang are you ready to play radio? Are you ready to shuffle off the mortal coil of mediocrity? I am if you are." Shepherd
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #1 on: July 22, 2014, 11:31:43 PM »

Carl

Since the resistor that is in series with the choke is a 2 watt, 2000 ohm resistor ... that implies that the DC passing through it is less than 32mA. It's probably less than 20mA... if the designer of the circuit built in some margin.

So the choke should be rated to handle 32mA to be on the safe side.

The choke won't have much effect unless it's impedance is more than 2000 ohms (the value of the resistor it is in series with) at the lowest frequency of interest. At 32Hz, that implies a 10H choke.

If the modulation resistance of the modulated RF stage is (for example) 5000 ohms... and keeping in mind that the plate resistance of the tube is specified as 9300 ohms when the average plate voltage is 135V...  then a 20H or 25H choke would be better (if you want a low frequency cutoff of 32Hz).

If you can accept a low frequency cutoff of 128Hz, then a 5H choke would be sufficient.

Note that 5000 ohms of modulation resistance (of the modulated RF stage) would correspond (for example) to an average plate current of 60mA and a plate-to-cathode DC voltage of 300V (at carrier)

Separately:

The biasing of the 6Y6G seems a little strange. The nominal (Class A1) biasing of a 6Y6G tube would be with: around 60mA of plate current, around 3mA of screen current, and around -13.5V of grid-to-cathode bias.

http://www.mif.pg.gda.pl/homepages/frank/sheets/127/6/6Y6G.pdf

But the 50 ohm cathode resistor (R9) would only provide 3.2 volts of cathode bias with 63mA of cathode current.

The tube is going to be operating with no grid current, except on the largest positive grid modulation peaks (if then).

Therefore the cathode resistor (R9) seems to be too small.

As another consistency check on R9: a 2 watt resistor would be able to handle a maximum of 200mA... if R9 is really supposed to be a 50 ohm resistor. But a 2W resistor would be able to handle a maximum of 100mA... if R9 were 200 ohms.  Therefore, the use of a 2W resistor suggests that R9 should be 200 ohms or maybe 270 ohms.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
W3RSW
Contributing
Member
*
Offline Offline

Posts: 3308


Rick & "Roosevelt"


« Reply #2 on: July 23, 2014, 08:20:00 AM »

Remember that The frequency "cutoffs" of simple LC and RC circuits are not absolute. They are by no means brick wall cutoffs.  3 Dbv down at the knee for a specified freq. still yields usable bass below cutoff. Roll off is 6 Dbv per octave or 20 Dbv per decade.   For the simple communication circuit shown, preceding amp stages will have already taken their toll even if the previous stage coupling caps of .005's, etc. are increased to .01's or even .05's.  

 5 henries is quite sufficient, particularly since even with it your rapidly approaching 120 HZ full wave hum from a full wave AC supply. Again I'm assuming for such a simple transmitter circuit your not trying for HIFi audio and highly regulated B+; are you not?

Also much of the current, both quiescent and modulated, to the RF final runs through the last AF tube, not just the choke/2k resistor combo.

Please also note that R9 and R11 have "see article notes" attached to their value so I'm curious, Carl, what the article states?  -probably related to what the RF tube type and class C current draw requires.  That may answer the seemingly low R9 cathode resistor value compared to the usual audio amp value for a class A stage into an output / speaker transformer.
Logged

RICK  *W3RSW*
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #3 on: July 23, 2014, 09:56:29 AM »

Rick

I agree that, for this purpose (i.e. a low frequency 3dB cutoff, at 128 Hz, is probably acceptable), using a 5H choke would be good enough.

Yes, the principal purpose of the 2k ohm resistor and the choke (and the 8uF filter capacitor, C6) is to supply low voltage B+ to the low level stages of the modulator, and to the screen of the 6Y6G... not to carry the average plate current of the modulated RF stage.

Of note:

If one used a separate low voltage B+ supply for the low level stages of this modulator and for the screen of the 6Y6G, one could eliminate this choke (and its effect on the low frequency rolloff of the modulator) entirely. These days, capacitors and diodes are very inexpensive; and there may already be a low voltage B+ supply for the lower level RF stages of the transmitter.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
k7mdo
Contributing
Member
*
Offline Offline

Posts: 420


« Reply #4 on: July 23, 2014, 09:57:40 AM »

I have an old AT-1 Heathkit modulated with this circuit.  The R-9 is initially to be adjusted so that final rf plate current is 1/2 of its value at resonance. In the AT-1 this ends up around 22 ma.

I have actually used the "now AM" AT-1 for local contacts on 75 meters.

Tom
Logged
Carl WA1KPD
Contributing
Member
*
Offline Offline

Posts: 1629



« Reply #5 on: July 23, 2014, 10:11:23 AM »


Please also note that R9 and R11 have "see article notes" attached to their value so I'm curious, Carl, what the article states?  
Hi Rick,
A copy of the article is attached. It has to do with the plate voltages and current in the system.
So a regular rcvr power supply choke will do which is what I was going to start with?
Although I am not going for hi-fi, more just playing with the concept, I may play with some of the caps after I get it working.

Thanks to both of you for the input.
Carl
WA1KPD

* SimplestModulator-1.pdf (905.5 KB - downloaded 450 times.)
Logged

Carl

"Okay, gang are you ready to play radio? Are you ready to shuffle off the mortal coil of mediocrity? I am if you are." Shepherd
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #6 on: July 23, 2014, 10:49:55 AM »

Carl
Rick

Having, now, browsed the article in which this modulator was presented...

It is apparent that this is not a "cathode modulator" of the type that comes to my mind when I use that terminology.

This is actually intended to modulate the grid voltage of the RF output tube... while only incidentally modulating the screen-to-cathode voltage and the plate-to-cathode B+. I.e., the RF amplifier's grid to cathode bias will be the difference between the grid-to-ground bias supply voltage, and the cathode-to-ground voltage produced by this modulator.

It is implicitly assumed in this design that the output RF tube is self biasing. I.e. the RF signal between the grid of the output RF tube and ground will cause some grid current to flow through into RF output tube... leading to a self adjusting DC grid bias of the RF output tube. It is further implicitly assumed in this design that the RC time constant of the RF output tube's grid self-biasing circuit is sufficiently large to allow the grid-to-cathode voltage to be modulated at the desired audio frequencies.

As such, the impedance that this modulator looks into (i.e. the change in the current passing through the modulator divided by the associated change in voltage at the output of the modulator) is much lower than it would be for the type of cathode modulator that came to my mind initially.

The required peak-to-peak audio frequency voltage swing is relatively small.

As a result, the value of the inductor is not critical. One could even eliminate the inductor (but not the 2k ohm resistor), and just increase the value of C6 to remove any modulation of the low-level B+ being supplied to the modulator via the 2k ohm resistor and C6.

As to the current passing through the 6Y6G, and the required cathode biasing resistor:

As Tom points out, the actual value of the cathode biasing resistor is adjusted so as to result in a 50% drop in the RF stage plate current when the modulator is inserted between the RF stage's cathode and ground. If the average cathode current, flowing through the output RF stage, when it is adjusted for full output, is I amperes... then the value of the 6Y6G cathode resistor will be approximately 2 x (13.5V/I). If I is (for example) 50mA, then the required cathode resistor value is approximately 26V/50mA = 520 ohms. [The required value of this resistor will be somewhat larger than 520 ohms, because some the RF stage's cathode current will flow through the 10k ohm resistor.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #7 on: July 23, 2014, 11:12:24 AM »


The circuit as shown will be a combination of control grid and cathode modulation. If the grid leak resistor, or the grid bias supply were to be returned to the RF tube cathode, then the modulator becomes a real cathode modulator. This circuit illustrates the point:

http://www.amwindow.org/tech/htm/wb9eckseriesmod.htm

I would't expect much more then 50% clean modulation with the circuit this thread is about. YMMV.

Jim
Wd5JKO
Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #8 on: July 23, 2014, 11:34:49 AM »

Carl
Jim
Rick
Tom

Thinking about this even more:

The self biasing circuitry of the output RF stage will probably have a short enough time constant to cancel out the effect on the grid-to-cathode bias associated with the modulation of the cathode-to-ground voltage.

Example: with a 100pF grid input coupling capacitor and a 30k grid-to-ground resistor, the time constant will be 3 microseconds. Therefore, the self biasing circuit will automatically readjust to keep the grid-to-cathode bias fixed, provided the cathode-to-ground voltage doesn't contain frequencies higher than 53kHz (i.e. the 3dB rolloff frequency for cancelling out the grid-to-cathode modulation). Below 53kHz, there will be some residual grid-to-cathode modulation... but the grid-to-cathode voltage modulation index will drop off at 20dB per decade as the audio frequency decreases.

Therefore, on further thought... the modulation of the screen-to-cathode voltage and the plate-to-cathode B+ (of the RF output stage) will not be incidental... it will be the principal cause of the modulation.

To the extent that there is residual grid-to-cathode voltage modulation... the associated modulation index will increase 6dB per octave as the modulation frequency increases.

In other words:

Back to the earlier discussion (including the necessary value of the RF choke, etc.). This circuit will behave as a "cathode modulator" (as we normally interpret that description), provided the RF output stage employs a grid self-biasing circuit with a time constant of a few microseconds (or less).

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
W3RSW
Contributing
Member
*
Offline Offline

Posts: 3308


Rick & "Roosevelt"


« Reply #9 on: July 23, 2014, 11:43:26 AM »

Boy are we having fun. You guys are deep into VYZ territory, cathode vs grid modulation. I'll try to assimilate time constants, but there is a constant grid leak voltage across the typical CW transmitter final of the era, thence with superimposed audio modulation across the grid leak resistor. A lot of assumptions are made as to what the grid circuit of the typical CW transmitter looked like.

Really a neat plug in for a smallish, cathode keyed CW rig. But I don't think many were built for large transmitters.

But back to R9

Let's say the 50 ohm r9 is not a typo.
It's rated at two watts, so back then that means it was expected to dissapate about 1 watt. Assuming 50 ma from what is nominally a 100 ma CW transmitter, And using  (P = E^2/R) , or ( E = Sqr PR) yields about 7.1 volts across it.  I think the R9 50 ohm spec. was about right for the typical 100 ma CW loaded little rig of the era.  

Of course the reduced from CW current criteria would determine it. Too bad we don't have a modern, color pix of the underside of the rig to see the resistor's color bands.  Grin
Logged

RICK  *W3RSW*
W3RSW
Contributing
Member
*
Offline Offline

Posts: 3308


Rick & "Roosevelt"


« Reply #10 on: July 23, 2014, 11:54:04 AM »

Maybe we should just build one.
Say a 6cL6 crystal EC Osc. And 807 final, about 450 on the plate, simple grid leak, rf choke lifted bias, simple PI net out and cathode key both the Osc. 

Build up the modulator beast as shown, plug it in and fire it up with a d104. Then play with the resistor / cap values and watch the scope.
Logged

RICK  *W3RSW*
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #11 on: July 23, 2014, 11:59:00 AM »

Rick

50mA DC flowing through a 50 ohm resistor:

Voltage across the resistor= 50mA x 50 ohms = 2.5V (But what is needed for a 6Y6G, drawing 50mA, is 13.5V)

Power dissipated by the resistor = 2.5V x 50mA = 0.125W (Why use a resistor with a 2W rating?)

The author mentions applications with as much as 200mA per (parallel) 6Y6G. 200mA x 50 ohms = 10V. That makes sense if you really plan to run that much current through each 6Y6G tube. However, running that much cathode current per 6Y6G tube, key down, seems (to me) to be a bad idea.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #12 on: July 23, 2014, 01:25:12 PM »

  File image is a quote from original QST article.

Jim
Wd5JKO


* Quote_From_QST_Article.PNG (27.77 KB, 374x111 - viewed 916 times.)
Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #13 on: July 23, 2014, 02:58:38 PM »

Jim

The author's article contains the text you quoted... and the article was published in QST.

Nevertheless, I believe that the author's statement is simply wrong; and that my analysis, in my post above (for the case of a self-biasing RF stage), is correct.

Note that the author says that:  

"To use the modulator, first tune up the transmitter for c.w. operation and load it to normal input. Then connect the modulator into the amplifier cathode circuit --- it can be plugged in place of the key if there is a key jack in the cathode---and the plate current should drop to about half of its c.w. value. Then talk --- that's all there is to it."

The author says nothing about readjusting the grid-to-cathode bias (which would become much too negative, unless the R.F. output stage is self biasing). This implies (at least to me) that the author is assuming (but not explicitly starting) that the RF output stage employs self biasing.

For the reasons I stated, in my post, above, the self-biasing would cancel out the grid-to-cathode modulation ... at audio frequencies of interest.

As a reviewer of technical articles submitted for publication in "refereed" journals, and as a technical journal editor... I have seen many articles submitted for publication whose principal results/conclusions were simply wrong. As a reader of articles published in a variety of well-respected technical publications, I have seen articles published whose principal results/conclusions were simply wrong. Recently, I joined the Ph.D. advisory committee of a student who had published a technical paper that had been reviewed (prior to submission for publication) by the student's Ph.D. thesis advisor and four additional "technical experts" who were members of the student's advisory committee. This took place prior to the time I agreed to become an additional member of the student's advisory committee. The conclusion of the published article was a key element (at that time, but not any more) of the student's ongoing thesis research. But, the conclusion was simply wrong. I knew that the conclusion was wrong when the student presented it to me (for the first time)... and it took me a while to find the actual error in the student's underlying mathematical derivation of the result. What the student had "shown", and also had published in a technical paper was (believe it or not) that you could find the exact carrier frequency associated with a perfect single sideband signal (100% carrier suppression and 100% unwanted sideband suppression) without any prior knowledge of what the characteristics of the modulating signal were. I also have seen completed Ph.D. theses (not by my students) whose main result was simply wrong. We all make mistakes. Sometimes our mistakes end up in publications in technical journals.



Stu

Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
Carl WA1KPD
Contributing
Member
*
Offline Offline

Posts: 1629



« Reply #14 on: July 23, 2014, 03:47:12 PM »

Maybe we should just build one.

I'm 90% of the way there. 6AG7 feeding 2X 807s- working 160-10
All additional switches tube sockets and mic input mounted in recovered real estate in teh xmtr chassis.

Hope to finish wiring and start the trial run in a couple of weeks.
Will let you know, or hear the results.
Carl
/KPD
Logged

Carl

"Okay, gang are you ready to play radio? Are you ready to shuffle off the mortal coil of mediocrity? I am if you are." Shepherd
Carl WA1KPD
Contributing
Member
*
Offline Offline

Posts: 1629



« Reply #15 on: July 23, 2014, 03:50:18 PM »

I have an old AT-1 Heathkit modulated with this circuit......  

I have actually used the "now AM" AT-1 for local contacts on 75 meters.

Tom
Something is working...... Smiley
What did you use for the choke?

Carl
WA1KPD
Logged

Carl

"Okay, gang are you ready to play radio? Are you ready to shuffle off the mortal coil of mediocrity? I am if you are." Shepherd
W3RSW
Contributing
Member
*
Offline Offline

Posts: 3308


Rick & "Roosevelt"


« Reply #16 on: July 23, 2014, 05:20:21 PM »

Stu,
Quote
50mA DC flowing through a 50 ohm resistor:
Voltage across the resistor= 50mA x 50 ohms = 2.5V (But what is needed for a 6Y6G, drawing 50mA, is 13.5V)

um, yeah, you can't be much more right than that. -you stated 2.5 volts at least twice.  Grin
Why I used a more complicated formula with wrong answer twice escapes me too. Sleuthing and assuming with too much coffee this morning.
Wet noodle time.  Oh, and I didn't sleep in a Holiday Inn last night, but was watching the Tour D'France this AM.
Logged

RICK  *W3RSW*
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #17 on: July 24, 2014, 10:03:56 AM »

Carl

(et al.)

As an additional comment:

When you lift the cathode of the RF output tube off of ground to put this modulator in series:

You need to add an RF bypass capacitor from the RF output tube's cathode to ground. You don't want to make it too large a value (or it will roll off the high audio frequencies in the modulation). A 0.005uF capacitor (-j6366 ohms at 5000Hz) would be a good choice. If the cathode was already (directly) connected to a key jack with an RF bypass capacitor... then you need to make sure that the value of that capacitor is not too large.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
Carl WA1KPD
Contributing
Member
*
Offline Offline

Posts: 1629



« Reply #18 on: July 24, 2014, 10:33:34 AM »

Hi Stu,
I think I have that covered with the 2 .005 caps that are there.
Attached is the xmtr schematic. I have circled the caps I am thinking of, as well as where I plan to insert the modulator.
Thanks
Carl
WA1KPD


* Chief Sche.jpg (182.63 KB, 1106x814 - viewed 955 times.)
Logged

Carl

"Okay, gang are you ready to play radio? Are you ready to shuffle off the mortal coil of mediocrity? I am if you are." Shepherd
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #19 on: July 24, 2014, 10:43:33 AM »

Carl

That should work, but you may find that the pair of 0.005uF bypass capacitors, in parallel, is too much. You may have to remove one, or change them both to 0.002uF, if the modulation response rolls off at too low an audio frequency.

In the schematic that you posted, there is a combination of "fixed" bias (R2 in parallel with C1), whose value (voltage) depends on the average current being drawn from the power supply; and self-bias (C6 in series with R5).

It will be interesting to see if the RF grid drive signal from the output of the 6AG5 oscillator has a large enough amplitude (peak voltage) to drive the 807 grids properly when the series modulator raises the cathodes of the 807s by a few hundred volts (positive) relative to ground.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
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 0.103 seconds with 18 queries.