The AM Forum
March 28, 2024, 07:38:48 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   Go Down
  Print  
Author Topic: zener diode for bias  (Read 11761 times)
0 Members and 1 Guest are viewing this topic.
KL7OF
Member

Offline Offline

Posts: 2313



« on: December 13, 2020, 12:49:08 PM »

could these zeners be used in the fil ctr tap on a 4-400  4-1000?  these things are 250W  and 13 cents each...https://www.digikey.com/en/products/detail/nexperia-usa-inc/BZX84-C8V2215/1156115
Ive used the 50W stud mount before...
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #1 on: December 13, 2020, 01:23:16 PM »

Steve,

Hmmmm... why does it first say 250W and then in the detailed description says 250 mW?

They look more like 250 mW to me.

T


DIODE ZENER 8.2V 250W TO236AB
Manufacturer Standard Lead Time
4 Weeks

Detailed Description
Zener Diode 8.2V 250mW ±5% Surface Mount TO-236AB


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.
W1RKW
Contributing
Member
*
Offline Offline

Posts: 4410



« Reply #2 on: December 13, 2020, 03:01:18 PM »

for a surface mount, it would be one hell of a zener if 250W. 

Mouser lists them as 250mW. Digikey misprint evidently.
Logged

Bob
W1RKW
Home of GORT. A buddy of mine named the 813 rig GORT.
His fear was when I turned it on for the first time life on earth would come to a stand still.
KL7OF
Member

Offline Offline

Posts: 2313



« Reply #3 on: December 13, 2020, 04:50:21 PM »

yeah I saw that after I posted...where do you purchase real 50W 5V zeners?   Is there a cheap source?   The hi voltage 50 watters cost waaay less than the 5 -8 volt ones.  
Logged
WBear2GCR
Member

Offline Offline

Posts: 4135


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


WWW
« Reply #4 on: December 14, 2020, 09:47:48 AM »


Maybe use a transistor or Mosfet + zener??

Would look like a shunt regulator circuit...

Or an adjustable 3pin Vreg (if the voltage isn't too high).

Or, a series of silicon diodes, stacked up, 0.6v drop each?
That's cheap, simple and easy.

                      _-_-bear
Logged

_-_- bear WB2GCR                   http://www.bearlabs.com
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #5 on: December 14, 2020, 11:02:06 AM »

I second the string of diodes.  I also put a largish cap across them to increase dynamic regulation.

I've also used a pass transistor and a TL431 with the associated circuitry.  A lot easier than 50 diodes in a string for the GS35b.

But, for ease of use, the 6A10 can't really be beat.

--Shane
KD6VXI
Logged
W1RKW
Contributing
Member
*
Offline Offline

Posts: 4410



« Reply #6 on: December 14, 2020, 12:49:48 PM »

I 3rd the string of diodes idea too.  K1JJ taught me that trick.
Logged

Bob
W1RKW
Home of GORT. A buddy of mine named the 813 rig GORT.
His fear was when I turned it on for the first time life on earth would come to a stand still.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #7 on: December 14, 2020, 01:16:20 PM »

The cool thing about a string of forward biased diodes in the cathode/ fil CT,  is to hook them to a rotary switch and by gradually shorting them out, have many plate idle current choices.  It is especially good when running class C linear using an AM carrier.   I add as much diode bias as I can while still mustering up enough drive on AM.

That AM class C linear mode really does work. I checked it the other day with the 4-1000A linear.  When in class C, the plates barely glow and the IMD  stays about the same.  Having this wide switchablity, IE, being able to run AB1 in SSB too, are quite the choices.

Of course this could be done with a string of big zeners, but the selections are more coarse and the zeners more expensive than rectifier diodes.

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.
KL7OF
Member

Offline Offline

Posts: 2313



« Reply #8 on: December 14, 2020, 09:29:02 PM »

The cool thing about a string of forward biased diodes in the cathode/ fil CT,  is to hook them to a rotary switch and by gradually shorting them out, have many plate idle current choices.  It is especially good when running class C linear using an AM carrier.   I add as much diode bias as I can while still mustering up enough drive on AM.

That AM class C linear mode really does work. I checked it the other day with the 4-1000A linear.  When in class C, the plates barely glow and the IMD  stays about the same.  Having this wide switchablity, IE, being able to run AB1 in SSB too, are quite the choices.

Of course this could be done with a string of big zeners, but the selections are more coarse and the zeners more expensive than rectifier diode
T
Hi Tom...How much do you really adjust the diode bias??? is it worth it for the associated wiring?   If you cud pick one setting what wud it be???
Logged
KL7OF
Member

Offline Offline

Posts: 2313



« Reply #9 on: December 14, 2020, 09:30:09 PM »

The cool thing about a string of forward biased diodes in the cathode/ fil CT,  is to hook them to a rotary switch and by gradually shorting them out, have many plate idle current choices.  It is especially good when running class C linear using an AM carrier.   I add as much diode bias as I can while still mustering up enough drive on AM.

That AM class C linear mode really does work. I checked it the other day with the 4-1000A linear.  When in class C, the plates barely glow and the IMD  stays about the same.  Having this wide switchablity, IE, being able to run AB1 in SSB too, are quite the choices.

Of course this could be done with a string of big zeners, but the selections are more coarse and the zeners more expensive than rectifier diode
T
Hi Tom...How much do you really adjust the diode bias??? is it worth it for the associated wiring?   If you cud pick one setting what wud it be???  for a 4-1000?

Logged
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 8308



WWW
« Reply #10 on: December 15, 2020, 12:39:24 AM »

Bias need not be fixed, fussy, or expensive. 

To reduce switch positions and use a common cheap switch, you can put 5 diodes in series for each step on a 6 position switch. 0 / 3.5 / 7 / 10.5 / 14 / 17.5V

If some fine adjustment is wanted, a second switch like it but with one diode per position. 0 / 0.7 / 1.4 / 2.1 / 2.8 / 3.5V

so, 0 to 21V in 700mV steps. Or whatever Vf is on the diode data sheet for the tube's current.

So simple!

The use of dirt cheap surplus bridge rectifiers such as found here: 
https://www.bgmicro.com/bridge.aspx

10A / 600V bridge $0.52
10A / 1000V bridge $0.79
12A / 800V bridge $0.48
15A / 200V bridge $1.93
25A / 200V bridge $0.99
I use them monthly for my parts orders, always low prices.

Each bridge having two sets of diodes in parallel doubles their current. Also the set is two in series so there is two junctions for one mounting screw. Overkill so they are less likely to blow up in a fault is real cheap with surplus. I like it anyway.
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #11 on: December 15, 2020, 01:40:54 AM »

Hi Tom...How much do you really adjust the diode bias??? is it worth it for the associated wiring?   If you cud pick one setting what wud it be???


Steve,

(Two 4X1 rigs, one linear and one plate modulated)   I have several HV taps  available that I use frequently to vary my 4-1000A power output. (2KV, 3KV and 4KV) So when the voltage is at 2KV, there are less diodes needed for the modulators.  When I ramp the rig up to 4KV, then there are more needed.  I use diodes in both the modulator and the RF final. The diodes in the RF final are really used as "fixed" bias so that if I lose drive, I have the final idle in class AB so there is no damage.  The modulator diodes are for normal biasing.  For a linear I usually have a class C position for AM linear that requires the most diodes and class AB1 for SSB.   The extra diode selections around these points will allow for the finer HV selections.

I usually put in a large string of them and use clip leads to determine the exact points. Then I  solder the rotary wires permanently.

So, bottom line is I select a HV position and then flip the rotary switch to get the correct idles. If the HV or mode is the same from the last time, then it all stays the same.  If used, you will need to find your own diode selections by experimentation and preference.  

For a linear, if I had to pick one setting it would be idling a 4X1 at about 80 mA at 4KV = 320 watts diss for a 1KW tube..   But for AM linear use, idle it below cutoff at zero current (no RF signal ) and then increase RF drive to let the carrier provide the "positive bias" to increase efficiency.


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.
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #12 on: December 15, 2020, 02:07:58 PM »

(Two 4X1 rigs, one linear and one plate modulated)   I have several HV taps  available that I use frequently to vary my 4-1000A power output. (2KV, 3KV and 4KV) So when the voltage is at 2KV, there are less diodes needed for the modulators.  When I ramp the rig up to 4KV, then there are more needed.  I use diodes in both the modulator and the RF final. The diodes in the RF final are really used as "fixed" bias so that if I lose drive, I have the final idle in class AB so there is no damage.  The modulator diodes are for normal biasing.  For a linear I usually have a class C position for AM linear that requires the most diodes and class AB1 for SSB.   The extra diode selections around these points will allow for the finer HV selections.

For a linear, if I had to pick one setting it would be idling a 4X1 at about 80 mA at 4KV = 320 watts diss for a 1KW tube..   But for AM linear use, idle it below cutoff at zero current (no RF signal ) and then increase RF drive to let the carrier provide the "positive bias" to increase efficiency.

Tom, please unconfuse me here...

You take a class AB1 linear amp, bias it up to class C, and then drive it with an already modulated AM drive signal? And it works and sounds good? Or something else entirely, like bias it up to class C, drive it with a CW RF signal, and modulate its power supply (ie plate modulate the amp)??

Kinda looking to do something like this. I have a linear, I wanted to turn it into a transmitter. As a linear, its AM carrier output is limited...

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #13 on: December 15, 2020, 03:35:53 PM »

Hi Ed,

Yes, you got it the first time...  Intuitively it doesn't make sense since we have always learned that a linear amplifier must conduct idling current to be clean on ssb and AM. Not so on CW.

But, take any regular linear amplifier like a pair of 3-500Zs that is biased into class AB or B.  Then increase the bias using say, diodes in the CT of the fil xfmr. (grid leak will not work - it must be a fixed bias source because the RF is varying)  Add enuff bias so that the amplifer is below cutoff.

It will take more AM carrier exciter drive now to get the carrier into the conduction zone. The carrier acts like a positive bias and the audio rides up and down on it in a linear fashion.  It's like the amplifier gets pulled back up into the linear curve without the penalty of class C distortion. You will notice the tube plates have less color for the same RF output from the final.  And, it will take more drive than when in AB1.  It can be quite dramatic.

I have run this idea past a number of guys smarter than me and a few were aware of this technique and agreed it works.  I think Buddly and a few others tried it a few years back based on a thread like this. They all saw good results.

The thing is it does not work on ssb since there is no carrier to bring the signal out of cutoff. SSB will splatter and have terrible crossover distortion when in class C.

I came across the technique by accident years ago when testing and continue to use it.  Maybe someone will give a more technical explanation of just what is happening. But try it and see for yourself and gain a good increase in AM linear amplifier efficiency, assuming you have the additional clean drive available.  I have tried biasing the amp way into class C using diodes and it took bigger and bigger drive, but never really found the limit.  I suppose class D using enuff drive to bring the linear stage down to 10 DB gain would be a reasonable limit, I dunno.

I see very little if no increase in IMD once the two settings are brought back up to the same output power.  IE, measure the IMD in AB1 at 500 watts out and then again in class C with the same 500 watts out. You will see the tube plate color darker and the IMD roughly the same.

I encourage anyone to try it and prove me wrong.  There are no free lunches (more drive needed) but in this case it seems the AM gods have smiled on us.

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.
KL7OF
Member

Offline Offline

Posts: 2313



« Reply #14 on: December 15, 2020, 04:39:35 PM »

Thank you Tom...good explanation....I plan to try this once I get the amp running a little smoother...still spits once in a while on hi bands...
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #15 on: December 15, 2020, 06:04:43 PM »

Tom,

As was discussed in the previous thread (Gosh, it must be a few years ago now) you also increase the "apparent modulation" on the signal with doing this.  IE, it's found that if you put a 100 pct modulated signal in, you may see an increase in peaks when driving the Class C amp.

Another thing to remember is:  You need to blush up those tubes at least once a year for a good session of gettering.

There are a few schools of thought on the reasoning behind why this works.  One person I spoke with told me the tube is self biasing due to rectification of the drive within the tube itself.  I don't know if I buy that, but.....  Sounds good Smiley

--Shane
KD6VXI
Logged
W2NBC
Contributing
Member
*
Offline Offline

Posts: 327



« Reply #16 on: December 15, 2020, 06:21:31 PM »

http://amfone.net/Amforum/index.php?topic=32635.0
Logged

Vintage Radio Pages- http://www.dealamerica.com
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #17 on: December 15, 2020, 07:21:31 PM »


Yep, Jessie, we certainly beat this horse to death back in 2012.

It appears enuff guys tried the test to suggest it works FB.  All my linears have it here.

Ya know, I'm presently playing with the idea of building a three or four tube 4-400A screen modulated rig.  But I'm wondering if this class C linear idea applied to a linear amp wud be more efficient than a screen modulated scheme.   Maybe in the end we will have the same results.  Bottom line is they are both running "class C"  with grid-like modulation...with no additional power applied to the plate as in plate modulation or PDM.   Series modulation is hopeless for a big rig efficiency-wise.    Cathode modulation uses about 85% grid and 15% plate modulation combined, so a 11N90 MOSFET in the cathodes of three to four 4-400As could be a good choice.

Shane, a good idea to get the anodes red once in a while for sure.  I also let my indirectly heated cathode tubes sit for 24 hours with just the fils running before firing up after a long rest. (> 6 months)

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.
VE7RF
Member

Offline Offline

Posts: 56


« Reply #18 on: December 16, 2020, 12:33:11 PM »

Installing  60 x 6A10's in series  works superb.  We used that on buddy's  YU-148 gg amp,  with a 20 position rotary switch.   The 1st 10 diodes are always in the circuit.  Then tapped every  2nd or  3rd  diode after that.  A spdt-center off toggle can be wired across the very 1st 3 x diodes for ultra fine adjustment.
On that setup,  7000 uf lytic wired across the entire string of  6A10's.   Bias V  does not  budge period.   Without the cap, the bias V will increase  by  12%, between  idle current..and normal full bore plate current.

No specs on the  YU-148/3x6  for  AB2  use for  ssb /am.    Just  60 vdc bias for  class  C   FM broadcast use.   We set the  idle at  200ma initially..then  experimented from there.  On the 3x3 amp,  bias is set for  150 ma.   Being able to dial up the bias/idle current...on the fly is a real treat.

On my latest bias setup,  I used a 43,000 uf  @  50 vdc  lytic..and mating  3"  bracket.   Does not budge.    12 position rotary switches are  easier to come by, and you can also feed the last output  (position #12),  into the com of a 2nd  12 x position switch..and  get another  12 positions if wanted to.  I can also shunt out the 1st 10 x diodes  with yet another toggle...to run zero bias....and do a load test on the B+  supply..and also test the cooling setup.

So far we have tried the above scheme on 3 x 6,  3x3,  YC-156/ YC-179,  3-500's  and a few other tubes...they all work.

Here's a  dedicated  bias  board, holds a max of  25 x 6A10/10A10 / 1N5408's   plus the 4 x holes for  mounting to standoffs  etc.  http://rfjunk.com/Bias-Board-25-Diodes-_p_299.html


This   bias board  holds  50 x diodes.  http://rfjunk.com/Bias-Board-50-Diodes-_p_300.html

It makes for a lot cleaner layout, vs  my  49 x diodes on vector board.

I gave up on zeners years ago.  Failure prone...and they always fail shorted.  Zener's  won't handle transients /glitches/surges/ spikes etc.
I tested strings of  1N5408's  on the  bench.   They ran warm with 1 A  CCS.    6A10's  run warm with 2A  CCS.   I also checked the V drop across each diode in the string.  They were all virtually identical..except a few that were .001  v  high...or  .001v  low.     Also tested  diodes in parallel..and yes the current splits  50-50..sometimes  49-51... or  51-49.

The  6A10/10A10  will handle a 400A  surge.  The 10AO7 is also rated at 10A  @ 1 kv.... put has a 600  surge rating..and a higher I  Rep  rating.   A  PM-600 is the same as a 6A10..and also works good.   If mounting to a PC board, mount the diodes well up off the board..since the leads are the heatsink.

We did find that the idle can only be so low on  CW mode....like  10-25 ma.   Much lower, and key clix will result.
Jim   VE7RF
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #19 on: December 16, 2020, 01:00:06 PM »

Jim,

FB on the rotary switch and grouped increments. I do the same here.

I have also had problems with zeners shorting in this service - as biasing in the CT leads, and they are always a short.  I thought it was just me.  Individual diodes in a string never crapped out. The BIG amp stress takes them out, whereas a docile 3-500Z amp may survive for years.

FB on adding the elec cap across the string.  I have done that on some amps here but will check to be sure I did it on all.  The stable key up is important.

From all the amps you tried, did you find a better efficiency using cut-off bias on AM as discussed?


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.
WD5JKO
Member

Offline Offline

Posts: 1996


WD5JKO


« Reply #20 on: December 16, 2020, 07:03:51 PM »

From all the amps you tried, did you find a better efficiency using cut-off bias on AM as discussed?

I took some data on this topic a few years back.

At the time, I was running four 572b's in a Gonset GSB-201 linear amplifier. The AMP was modified to add a pseudo filament centertap, and a big 12V zener to ground, or direct to ground zero bias. At the time three of them 572B's were soft, and one was stout. I could get 100 watts carrier output all day long with headroom for > 100% modulation.

The data matrix adds in ac filament power, and the HV DC supply losses. Doing that and the overall efficiency really suffers. Still the DC input/RF output was about 25.5% at zero bias, and 30% with 12v bias. That efficiency boost is not 4.5%, but really over 17%. My bet is that with good 811's, and 12v bias, my efficiency is up around 35%, maybe more.
https://percentagecalculator.net/

These days I have a stout set of Russian 811A's in the amplifier. Here I can run 160 watts carrier out on AM, with headroom for 140% modulation. No color at all on the 811A plates. For SSB I need 4.5v bias. I never repeated the efficiency matrix when using 811's.

Jim
Wd5JKO


* Efficiency Matrix GSB-201.jpg (232.96 KB, 1003x597 - viewed 412 times.)
Logged
K8DI
Contributing
Member
*
Offline Offline

Posts: 423


« Reply #21 on: December 16, 2020, 07:14:49 PM »

...take any regular linear amplifier like a pair of 3-500Zs that is biased into class AB or B.  Then increase the bias using say, diodes in the CT of the fil xfmr. (grid leak will not work - it must be a fixed bias source because the RF is varying)  Add enuff bias so that the amplifer is below cutoff.

It will take more AM carrier exciter drive now to get the carrier into the conduction zone. The carrier acts like a positive bias and the audio rides up and down on it in a linear fashion.  It's like the amplifier gets pulled back up into the linear curve without the penalty of class C distortion. You will notice the tube plates have less color for the same RF output from the final.  And, it will take more drive than when in AB1.  It can be quite dramatic.

One clarification or question left for me here...  My gut wants me to assume the linear amp in question needs to be grid driven. Is this the actual case or does this work in a typical grounded grid linear too?

Ed
Logged

Ed, K8DI, warming the air with RF, and working on lighting the shack with thoriated tungsten and mercury vapor...
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #22 on: December 16, 2020, 07:30:40 PM »

Good on all, Jim.   Your results seem to confirm our results too.  The color difference I've seen is actually quite dramatic on my pair of GG 4-1000As with 4KV.  (with regulated grid and screen voltages)

Ed:  Yes, it should work with both GG or driven driven amplifiers.  So far all of my tests have been with GG amps.  Grid driven results are just my assumption until  someone actually tests it.  There's not many grid driven linears out there these days.  Let's see... the old Hallicrafters  HT-33 with the PL172 -  and the Johnson Thunderbolt with 4-400As are grid driven I think.  There's a few more.

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 #23 on: December 16, 2020, 09:03:01 PM »


Ed:  Yes, it should work with both GG or driven driven amplifiers.  So far all of my tests have been with GG amps.  Grid driven results are just my assumption until  someone actually tests it.  There's not many grid driven linears out there these days.  Let's see... the old Hallicrafters  HT-33 with the PL172 -  and the Johnson Thunderbolt with 4-400As are grid driven I think.  There's a few more.

T

Yes, the HT-33 is grid driven, I have one that I restored, it is fully operational.  Other than a new oil cap, the only mod was to replace the 866As with 3B28s.  The bias is adjustable from the front panel, but the pot does not take it all the way down to cutoff.

That amplifier runs hot with a lot of dissipation even without modulation on sideband.  Would need to modify the bias circuit to make cutoff for use on AM, but that is a good idea.  It is one heavy beast, but with such a stout power supply, it would probably be a very good candidate for AM, especially if the resting plate current was cut way back.
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: 738



« Reply #24 on: December 16, 2020, 09:04:47 PM »

  It's things like these that make following a technical thread like this so rewarding. Wow - nothing like totally punching the reset button on a set of basic assumptions! Caution: folks with rigid thought patterns may feel lightheaded when their basic assumptions are shaken. They feel giddy. In extreme cases they fall over. Thud. I'm sitting down; I'm safe.
  I've been thinking of building a 3-500 amp - maybe a three-holer. If I do, I'm going to experiment with this concept.


Hi Ed,

Yes, you got it the first time...  Intuitively it doesn't make sense since we have always learned that a linear amplifier must conduct idling current to be clean on ssb and AM. Not so on CW.

But, take any regular linear amplifier like a pair of 3-500Zs that is biased into class AB or B.  Then increase the bias using say, diodes in the CT of the fil xfmr. (grid leak will not work - it must be a fixed bias source because the RF is varying)  Add enuff bias so that the amplifer is below cutoff.

It will take more AM carrier exciter drive now to get the carrier into the conduction zone. The carrier acts like a positive bias and the audio rides up and down on it in a linear fashion.  It's like the amplifier gets pulled back up into the linear curve without the penalty of class C distortion. You will notice the tube plates have less color for the same RF output from the final.  And, it will take more drive than when in AB1.  It can be quite dramatic.

I have run this idea past a number of guys smarter than me and a few were aware of this technique and agreed it works.  I think Buddly and a few others tried it a few years back based on a thread like this. They all saw good results.

The thing is it does not work on ssb since there is no carrier to bring the signal out of cutoff. SSB will splatter and have terrible crossover distortion when in class C.

I came across the technique by accident years ago when testing and continue to use it.  Maybe someone will give a more technical explanation of just what is happening. But try it and see for yourself and gain a good increase in AM linear amplifier efficiency, assuming you have the additional clean drive available.  I have tried biasing the amp way into class C using diodes and it took bigger and bigger drive, but never really found the limit.  I suppose class D using enuff drive to bring the linear stage down to 10 DB gain would be a reasonable limit, I dunno.

I see very little if no increase in IMD once the two settings are brought back up to the same output power.  IE, measure the IMD in AB1 at 500 watts out and then again in class C with the same 500 watts out. You will see the tube plate color darker and the IMD roughly the same.

I encourage anyone to try it and prove me wrong.  There are no free lunches (more drive needed) but in this case it seems the AM gods have smiled on us.

T



Logged

"Gosh, Batman, I never knew there were no punctuation marks in alphabet soup!"
—Robin, in the 1960's Batman TV series.
Pages: [1] 2   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 0.082 seconds with 18 queries.