The AM Forum
March 28, 2024, 02:56:34 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1] 2 ... 5   Go Down
  Print  
Author Topic: 811 bias for Mod  (Read 48348 times)
0 Members and 1 Guest are viewing this topic.
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« on: December 28, 2018, 07:06:12 AM »

Picked up a sweet pair of tested 811's from a guy in last ER issue (Sylvania) for a modulator I am building. 811,not 811A. Can run them at 1700 or 1500 (have multiple taps on the S-49 transformer). I know that you can push 811A's to 1700, lots of folks do. Not so sure about the 811's. These are going to drive a single 8000 tube a.la. W8ACR/K9ACT circuits (triodes are nice). For the 811A the bias is -4.5V at 1500V. For the 811 the bias is listed at -9V at 1500V.

1. Should I try to run it at 1700 or be happy at 1500?
2. I have a free 6.3V tap on the grid trans that can give me 8V easy. Is that enough bias or will I really need the extra volt?
3. Should I skip the 6.3V hardware and just use a diode string or Zener for the bias?

John W9JSW
Logged
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #1 on: December 28, 2018, 09:19:05 AM »

John,

The bias point depends on the plate voltage applied to the 811's. 0v@1250v, -9v@1500v, -??v@1700v, etc. The unused 6.3v winding should run a bit high, being unloaded by filament current. It will probably give you all the voltage you'll need with the light load of the bias. I think you'll just have try it, and see what you get. A small variable power supply would be handy for testing, or even just a 9v battery.


Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #2 on: December 28, 2018, 09:52:00 AM »

If you use a voltage doubler, you’ll be able to get 12v from that 6.3 winding and be able to use a pot to set whatever bias you need.

Ed, KB8TWH
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
KK4YY
Member

Offline Offline

Posts: 565


Your best isn't as good as you can be.


« Reply #3 on: December 28, 2018, 11:21:17 AM »

If you use a voltage doubler, you’ll be able to get 12v from that 6.3 winding and be able to use a pot to set whatever bias you need.
Good call, Ed. I did that in my 813 amp. Worked fine.

Don
Logged

All your worries won't add a day to your life, or make the ones you have any happier.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #4 on: December 28, 2018, 11:47:00 AM »

3. Should I skip the 6.3V hardware and just use a diode string or Zener for the bias?
John W9JSW

Hi John,

Will the modulator grids be pulling any current? (AB2, B2, etc., )  If so, you will need to regulate the bias supply.  Even a few tenths drop will add to the distortion with a hard-driven high Mu tube.  IE, why add distortion, even a tiny amount, if you can avoid it?

That said, in every big rig I have ever built to date, (linear RF, audio AB2)  I have always used a string of diodes in the fil center tap and a rotary switch to select the exact bias I needed, no matter what plate voltage (or screen voltage with tetrodes) I chose.  Figure on ~0.5V drop per diode, diode arrow direction towards ground, forward biased. 20 diodes in a string should give enough range.

The diodes technique is rock solid self-regulated and adjustable. I love using fil CT diode bias for directly heated cathode tubes.

BTW, run as much plate voltage as you dare on the rig. Higher voltage helps efficiency and the tubes will be driven easier with less grid current for a given output. (cleaner)    Make them sing for their supper, but cleanly!  Add some audio negative feedback around the modulator and you're golden. 6-10 dB  NFB  - experiment.

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.
WD8BIL
Contributing
Member
*
Offline Offline

Posts: 4409


« Reply #5 on: December 28, 2018, 12:15:52 PM »

I'm with Tom Vu on this one. I've built many many amps with 811, 811As and 572Bs. All used Zener diode biasing. Simple, cheap and effective.

Logged
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« Reply #6 on: December 28, 2018, 05:28:35 PM »

Tom please explain how to do this - Add some audio negative feedback around the modulator and you're golden. 6-10 dB  NFB  - experiment.

1700v it is, with a diode string instead of the battery.

Attached is my schematic. Let me know if you see anything messed up. Been reading a LOT of old threads on how to get this simple design correct.

John - W9JSW

* 8000 Transmitter.pdf (912.09 KB - downloaded 366 times.)
Logged
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #7 on: December 28, 2018, 05:31:00 PM »

I have always preferred the center-tapped filament transformer with a zener because one can sample plate current in a simple fashion as shown in the file below:


Phil - AC0OB

* General Modulator with 811s.pdf (114.89 KB - downloaded 356 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #8 on: December 28, 2018, 09:07:51 PM »

OK John -

I looked over your schematic and it looks very FB.  

The only change I would make is to use a bigger "glitch" resistor in the plate HV, like 5-20 ohms.  If I read it right, there is only 0.82 ohms in there now?

Just ground the grid CT for the modulators and put the diode bias string in series with the modulator fil CT and you're all set.   Just to be sure you understand, use a rotary switch to short out a few diodes at a time to change the resting current. You can use a clip lead to find the best choices before wiring the rotary switch  in.


Audio negative feedback for the modulator:

This can be an advanced subject - I am covering only highlights here -  you might do some reading on the web. I'll get ya headed in the right direction.  

Basically what I did for my pair of 4-1000A grid driven modulators with regulated screen current:  Start with a HV resistor ladder from one modulator plate cap to ground. It might be 1-2 megs with a smaller 1K resistor at the bottom of the ladder to ground. Tap a small audio voltage (like 3 volts) from the 1K to ground and feed this audio voltage back to a low level 1 volt audio point in your audio chain.  Use a capacitor in series with this lead that will have to be experimentally determined... 500 pF - 0.1 or whatever works best doing audio sweeps.  You may also need a small HV cap, like 30-50 pF ACROSS the ladder to fine tune things. It's all about running audio sweeps and trying things to get the best scope picture.

For a while I tapped off the modulator transformer secondary but later found it better over all not to push my luck and stick with the tube plate cap.  Doing NFB through a mod xfmr with all its phase shift is not easy.

At the low level output at the ladder use a small value experimentally selected cap to ground to limit ultra-sonics.  The phase needs to be right - pick the proper plate cap so it will be negative feedback, not positive. If positive it will take off and may do damage.  Use a variable pot to adjust the level of audio into the 1 volt preamp input, just like a mixer. This pot determines the amount of NFB in DB.  Too much NFB and the system may become unstable and take off. I was able to get 12 DB after doing a lot of experimentation and settled for about 8-9 DB NFB in the end.

For 6DB of negative feedback you will have to have 6dB of excess system gain available because this gain will disappear once the NFB is working with 6 DB of NFB. There are surely some good 811A NFB designs out there on the web, so look around. When running right, the NFB will greatly smooth out the extreme lows distortion and increase your high end limit to reach extremes in a cleaner manner.  On my 4X1 modulators, I was able to get the response down to 20Hz and up to 12 KHz before it fell apart.  Without NFB it was not as clean. The 4X1 rig now approaches the class E rig I used to run, though the E rig had a wider and more transparent range for sure.

You should get the modulator working perfectly before adding the NFB. That's the good thing - you can add NFB at any time.


Phil:

Your schematic looks OK, except I'm not sure if I am correctly reading the fil CT 10 ohm metering resistor? Is that a safety resistor with a missing direct reading mA meter?   Excuse me if I am, but a 10 ohm resistance in the modulator CT will cause a voltage drop across the 10 ohms that will mean a couple of volts fluctuation as the class B cathode current flows up and down. This will result in distortion. There should be a low resistance current meter in place of the 10 ohms, just like John's circuit.   I just want to be sure -

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.
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #9 on: December 28, 2018, 10:03:24 PM »

OK John -

Phil:

Your schematic looks OK, except I'm not sure if I am correctly reading the fil CT 10 ohm metering resistor?   Excuse me if I am, but a 10 ohm resistance in the modulator CT will cause a voltage drop across the 10 ohms that will cause a couple of volts fluctuation as the class B cathode current flows up and down. This will cause distortion. There should be a low resistance current meter in place of the 10 ohms, just like John's circuit.   I just want to be sure - maybe the low-resistance meter is not shown and goes across the resistor which would solve the problem... or just replace the 10 ohm resistor with a direct reading 200 mA meter or whatever value.


T  

I usually meter everything with a cheap 10k/V voltmeter across a resistance since it is difficult to find decent meters of the value you need or inconvenient to drill 2 inch plus holes in thick aluminum.

I don't know about distortion or how much there would be but there would definately be negative feedback voltage at the cathode (filament).

As you suggested, one could simply monitor plate current with a current meter at the zener or drop the resistor to a 1 ohm value and use a DMM to determine plate current in which one would see 100 mV/100 mA of current.

In the new schematic below, I don't think 200 mV of cathode feedback voltage would cause any problems.


Phil

   


* General Modulator with 811s.pdf (115.95 KB - downloaded 294 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #10 on: December 28, 2018, 10:20:06 PM »

Phil,

Sure, 200 mV is much better compared to 2 volts of cathode fluctuation. Though 200 mV is still about 4% regulation which is "acceptable."  (using 4.5 volts of bias)    A direct reading meter might still be  a better solution.

A cathode resistor would work fine in a class A rig since there is no fluctuation, and use to be a common practice in low level tube audio amp circuits. But class B -  NG.

The reason I am so anal about this is because for a year I ran a pie-wound 15 ohm small wire, series cathode center-tap choke in a big indirectly heated cathode RF linear amplifier. (This was a "trifilar" choke, resulting in 30 volts fluctuation) The tube was known for its cleanliness and specially designed for linear service.  I could never understand why I could not get better than -25 DB 3rd IMD using it. Splatter problem.  I tried everything. Then one day it dawned on me and I wound my own choke using #18 wire. The choke was now < 0.25 ohms, 1/2 volt fluctuation. The IMD immediately shot up to -45dB++ where is should be. Clean as a whistle. HUGE difference on the airwaves.

I'm on a crusade not to let anyone else suffer thru what I went thru... :-)

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 #11 on: December 28, 2018, 11:02:02 PM »

As the circuit is drawn, plate current meter A2 will show the combination of plate current plus grid current.  In order to register only plate current through A2, connect the positive side of the bias supply and the bottom of the fixed bias adjustment potentiometer R14 to the center tap of the 810 filament transformer secondary.  With this change, grid current will no longer flow through meter A2.  An alternative would be to just ground the center tap of the filament transformer, and float the negative side of the high-voltage supply.  But when using a single high-voltage supply for both modulator and final amplifier, the first option is simpler, thus requiring no change to the modulator minus supply circuit.

i assume meter A8 is a voltmeter, intended to show total negative grid bias.  If that is your intention, you might consider moving the meter connection to the bottom (cold end) of RF choke L3.  You really do not want RF on the meter.  

Also, if you have any difficulty neutralizing the 810 final, you may consider reducing the value of C10.  This capacitor forms a voltage divider with the neutralizing adjustment capacitor C9.  5000 pf for C10 seems a bit large.  Another option to make it easier to neutralize is to use a split-stator capacitor for C5, grounding the rotor, and eliminating C10.  The handbooks discuss various methods of neutralization, with pros and cons of the individual circuits.

To reduce the possibility of parasitic oscillations, it is a best practice to ground all the bypass capacitors in the RF amplifier to a common grounding point on the chassis.

Good luck on your project!
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
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #12 on: December 28, 2018, 11:25:12 PM »

Quote
K1JJ: ...Sure, 200 mV is much better compared to 2 volts of cathode fluctuation. Though 200 mV is still about 4% regulation which is "acceptable."  (using 4.5 volts of bias)    A direct reading meter might still be  a better solution...

Did you mean to say 9 volts for bias? I was going by the Amperex spec sheet for Class B and a Vp = 1500V.

https://frank.pocnet.net/sheets/084/8/811.pdf


Phil
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #13 on: December 28, 2018, 11:57:46 PM »


Did you mean to say 9 volts for bias? I was going by the Amperex spec sheet for Class B and a Vp = 1500V.

https://frank.pocnet.net/sheets/084/8/811.pdf

Phil


Hi Phil,

I was going by what the guys mentioned earlier at 4.5 volts.  9V looks fine with 1500V as you said.
I usually use the switchable bias diodes, so don't worry much about bias until the rig is operating and gets the optimization dance in real time.

Good eye, W8KHK...  Yes the neutralization divider would be better at .005 - .001.  Good find on the grid meter/ RF connection change.   

The other minor thing is the grid bias is labeled screen bias... no screen in the 811A... :-)


Now if I can just find the motivation to get on the air. The 75M band has been so long it might be time to call for west coast.

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 #14 on: December 29, 2018, 12:07:33 AM »


Did you mean to say 9 volts for bias? I was going by the Amperex spec sheet for Class B and a Vp = 1500V.

https://frank.pocnet.net/sheets/084/8/811.pdf

Phil



The other minor thing is the grid bias is labeled screen bias... no screen in the 811A... :-)


Now if I can just find the motivation to get on the air. The 75M band has been so long it might be time to call for west coast.


As I learned long ago from Dennis, W7TFO, in a triode, you cannot hear it screen. 

Yes, get on the air, Tom.  I will be away from home for about ten days, so I will miss the Heavy Metal Rally.  I will take my SDRplay RSP2 so I can eavesdrop, but I will be back in time for the AM Rally.  Hopefully John, W9JSW will be ready for that event in February!  At the rate he is progressing, he will not have to push an AM button on a plastic radio!
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
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« Reply #15 on: December 29, 2018, 07:28:58 AM »

Thanks guys,

On C10, that was to be used across A8. If I move the meter down, the cap goes with it and solves the divider problem.

Just about ready to start bending metal. Is there a problem with having the modulator and the rf deck side by side? I have a short 9U (15in) encased metal rack and there is not enough room for a stacked config with the rf deck close to 12 in tall. I don't want to place the 8000 horizontal as I am leaving room for a second one (just in case  Grin ). My mod iron is the gating factor keeping me at 1 final. Hoping to score a better one eventually. I know I could put in a heising reactor but those are hard to find and will still leave me with a modulation transformer with a limited audio range.

So, my plan is to build the mod on a 7x12x3 chassis, and the deck on a 10x12x3 chassis. Work with them on the bench then join them together into a 17x12 unit and mount them on a 7U front panel. I plan to build my HV power supply as a separate unit and control it from a 2U power control that will have 115v switching, circuit breakers and soft start units. The plan for PTT is a dow-key sequencer. I just picked up a 12V one with an extra relay on the outside. I ordered the ER issue with Don's article on a simple sequencer.

My thoughts on PTT are to cut off the exciter and the audio and let the resting bias keep the 8000 in cut-off with the HV left on. Is this a good idea? Of course, I could cut off the HV if I want to if a really long span of RX is happening.

My exciter will be sandy. I have a digital VFO that I plan to run into one of those cheap QRP-Labs 20W amplifiers. I will have filters on the exciter to insure good clean signal into the transmitter. Later I may decide to replace this with a tube exciter. RX is an SDR.

For audio, I have 2 options. One is an old SS PA amp with mic inputs. Have to check it out more. It may be 70V output. The other option is that I bought one of those cheap Class D audio amps on Amazon. It is a 100W mono subwoofer and has instructions on how to remove the low pass to open up the audio range. I have a mic preamp that I can use if I go that direction.

Hope to start actual assembly real soon now. Been waiting long enough.

John W9JSW

Logged
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« Reply #16 on: December 29, 2018, 07:49:48 AM »

Do I need the clipper that Phil shows in his schematics? I read about them but do not understand how this one works. The Zeners add up to 580V. Does it just flat top the audio if over-driven? I see some transmitters that have splatter +/- 10kc or more and that is pretty bad. Will this prevent that? On my SDR transmitter (see QRZ for pics) the passband is brick wall sharp. Don't want this transmitter to be a band hog...

See attached revised schematic with all of the above suggestions (except NFB) implemented.

John W9JSW

* 8000 Transmitter.pdf (486.42 KB - downloaded 294 times.)
Logged
w8khk
Member

Offline Offline

Posts: 1200


This ham got his ticket the old fashioned way.


WWW
« Reply #17 on: December 29, 2018, 11:19:15 AM »

Thanks guys,

On C10, that was to be used across A8. If I move the meter down, the cap goes with it and solves the divider problem.


Yes, move the meter down.  With the meter where it is you probably will not be able to neutralize, and you will have RF on the meter.  After you move the meter, you may want to add an additional bypass capacitor across the meter which is now connected at the bottom of the RF choke.  But you must keep a cap where C10 is presently, however a smaller value, such as 500 to 1000 pf (not 5000 pf).  This cap and the neutralization variable are an RF voltage divider that enables neutralization.  You may remove C10 if you use a spilt stator capacitor for grid tuning, (grounded rotor) but without a spilt stator you must have C10 in its present location to neutralize.
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
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #18 on: December 29, 2018, 11:43:26 AM »

A few simple thoughts on negative feedback - neglecting phase angle.
The important thing to consider about feedback is that if it starts to become positive, you have
an oscillator.

That feedback resistor, if you put a cap in parallel then at some frequency that is a function of
the reactance of the cap, the cap "looks like" a wire, aka very low resistance. So that cap sets
an HF rolloff based on the value and so the reactance vs. the impedances. Usually this cap is
set to control HF overshoot, for a hifi type amp. It could be set to provide a 6db/octave rolloff
at some lower frequency, like -3dB@ 6kHz. for example. You'd be down 9db @12kHz.

There are other in circuit means by which bandwidth can be limited.

The point that you come back to with the feedback must be a point that is out of phase
with the point that you source it from - they need to cancel. Also the circuit needs to have
"excess gain". That means that if one needs to have everything running full out to reach clipping
of the 811 tubes, then adding feedback will not give you enough swing, not without increasing
gain, or increasing the input level, if that does not overload the input...

The small caps to ground are generally for the purpose of limiting or eliminating ultrasonics
and/or RF from the feedback and/or gain stages...

It's a good idea to look at something like a triangle wave once the modulator is built, with first
a resistive "dummy load" on the secondary, then with the actual final to be sure that it looks clean
and linear up into clipping...

                           _-_-

Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« Reply #19 on: December 29, 2018, 11:49:40 AM »

OK, Rick. I have it correct now.

On the grounded variable cap. I have a 10-100 breadslicer cap in the junk box. Will that work as a direct replacement in a single ended LC or will I have to recalculate the LC and also center tap the inductor? I have this 21uF Air-dux that just works out single ended. I did not want to go with a center tapped inductor due to ease of build for 4 bands. I will if it makes better sense. The input LC calculations were pretty easy using a LC calculator and 15k for RL. Not sure how to model the other circuit.

Pls review my rewiring of the bias. I forgot to bring the +V off of the supply. I think I have it correct now.

John

* 8000 Transmitter.pdf (922.3 KB - downloaded 301 times.)
Logged
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« Reply #20 on: December 29, 2018, 11:57:41 AM »

Thanks Bear. I understand.

I have a good scope and plan to take this stepwise starting with the modulator and testing it separate. That is why I wanted to have separate but joinable chassis. Easier to work with as I build it up.

Going to hold off on the feedback idea for now. Already a full plate.

Just realized I am going to need a bigger dummy load. My MFJ 300W unit is only good at that level for a few minutes. Wonder if I can strip it and insert it into a can of mineral oil and make it a cantenna?

John
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #21 on: December 29, 2018, 12:03:14 PM »

Hi John -

There's a lot to comment on, but I will chime in on two areas dear to my heart....

1) PTT Keying the modulators and final:  Yes, they need keying. Keeping the modulators running will burn up power for nothing - and might even generate hash into the RX if not stable. With modulators, even though they are biased into "cutoff" class B, they still need to idle at a recommended current so that the crossover point of the sinewave is clean - preventing crossover distortion.  To key, just stick a 30K - 50K resistor in series with the cathode circuit and use a set of contacts to short it out during key up. IE, 50K biases the modulators into deep-cutoff - and then this shorted resistor lets you idle the modulators anywhere you desire using the biasing diodes as described before. Do the same thing with the RF final.

See my 813 circuit for an example:   (There should be a .001 cap from the meter/RFC junction to gnd in the screen circuit)

https://www.amwindow.org/tech/htm/813/813.htm



Negative peak limiting and clipping:

There's a lot of opinions on this and I'll give you mine:  I feel all clipping and limiting (and processing) needs to be done in the low level audio stages. There was a time when I used high level clipping/limiting in all seven of my homebrew rigs. But one by one I removed these circuts due to side-grunge, splatter reports whenever I leaned into them. High level to me is more as a safety net.  Since then, I use just one master low level audio system and it works FB. (528E, PMC-300A, 31 band EQ...)   For hard negative peak limiting I use Steve/ QIX's clipper circuit taken from his PDM generator design. Just a simple diode and resistor circuit to limit the negative peaks to 95% and let the positive peaks soar to where the rig can handle... 140%, etc.

Add in a couple of 19" rack audio boxes to help audio processing and you are covered. If everything is running well in the one volt low level stages, (SDR is a good linear amp solution too) there is no reason to need splatter clipping in the high level mod transformer stages - except for arc-over protection, which is a valid reason.   Again, some use high level clipping and it works for them.  Be sure to have ball arc-gaps for mod transformer protection...  I use gaps on my Heising reactor too and even on the screen choke to quench collapsing fields when unkeying.

BTW, if you experience excessive arcing across the gaps during unkey, there are a couple of great solutions I will tell you about if needed later. The collapsing mag field has to be dissipated  somewhere and hopefully not in the form of damaging mod transformer turns arcing.

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.
WBear2GCR
Member

Offline Offline

Posts: 4135


Brrrr- it's cold in the shack! Fire up the BIG RIG


WWW
« Reply #22 on: December 29, 2018, 12:05:22 PM »

At 80m, you can use a variety of things for a dummy load - including appropriately sized heating elements, water or
mineral/transformer oil can be used. (if needed!)

A minor concern to me is that the secondary of the mod transformer is unloaded in "normal" operation, until
the grids start to draw appreciable current... I think I'd opt for a light resistive load on the secondary of the mod iron.
Dunno if this is standard ham practice or not...

                           _-_-
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #23 on: December 29, 2018, 12:22:27 PM »

Do I need the clipper that Phil shows in his schematics? I read about them but do not understand how this one works. The Zeners add up to 580V. Does it just flat top the audio if over-driven? I see some transmitters that have splatter +/- 10kc or more and that is pretty bad. Will this prevent that? On my SDR transmitter (see QRZ for pics) the passband is brick wall sharp. Don't want this transmitter to be a band hog...

See attached revised schematic with all of the above suggestions (except NFB) implemented.

John W9JSW

Hi John,

In the beginning of the thread I wasn't sure how much B+ was being run on the Mod transformer and the original circuit was a Soft Knee Limiter for supplies running about 700V.

The soft knee limiter was suggested by Bonomo in ER #111 to save the Modulation secondary from failure by loading it on negative peaks, reducing any high voltage
 V = - L di/dt inductive “kickbacks.”

I use it for any upgrades to EF Johnson and Heathkit type transmitters and a variation of it for homebrews'.

For your application at 1700 volts, the circuit has more 200 volt Zener diodes.


Phil

* Soft knee Limiter.pdf (148.33 KB - downloaded 285 times.)
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
DMOD
AC0OB - A Place where Thermionic Emitters Rule!
Contributing
Member
*
Offline Offline

Posts: 1768


« Reply #24 on: December 29, 2018, 12:37:33 PM »

OK, Rick. I have it correct now.

On the grounded variable cap. I have a 10-100 breadslicer cap in the junk box. Will that work as a direct replacement in a single ended LC or will I have to recalculate the LC and also center tap the inductor? I have this 21uF Air-dux that just works out single ended. I did not want to go with a center tapped inductor due to ease of build for 4 bands. I will if it makes better sense. The input LC calculations were pretty easy using a LC calculator and 15k for RL. Not sure how to model the other circuit.

Pls review my rewiring of the bias. I forgot to bring the +V off of the supply. I think I have it correct now.

John

One additional comment John on your latest circuit:

It appears your grid current ammeter is shunting the grid and protective bias.

I think it should be in series with L3 and R1 since you are measuring the current in a series circuit there.


Phil
Logged

Charlie Eppes: Dad would be so happy if we married a doctor.
Don Eppes: Yeah, well, Dad would be happy if I married someone with a pulse.NUMB3RS   Smiley
Pages: [1] 2 ... 5   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 0.087 seconds with 19 queries.