The AM Forum
March 29, 2024, 06:04:26 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 ... 4 [5] 6 7   Go Down
  Print  
Author Topic: Looking for ideas for a ~25 watt plate modulated homebrew Pissweaker rig  (Read 62256 times)
0 Members and 1 Guest are viewing this topic.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #100 on: December 02, 2020, 10:57:00 PM »

*** BE SURE TO LOOK BACK  AT THE LAST  FOUR POSTS TO THE BEGINNING OF THESE NEW PICS ***  (Page 4)


* DSCF0006.JPG (322.15 KB, 1280x960 - viewed 386 times.)

* DSCF0027.JPG (324.95 KB, 1280x960 - viewed 352 times.)

* DSCF0029.JPG (319.78 KB, 1280x960 - viewed 374 times.)
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.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #101 on: December 02, 2020, 10:58:47 PM »

More:

I plan to get real glass for the window.  Also a nice cabinet.


* DSCF0022.JPG (325.12 KB, 1280x960 - viewed 340 times.)

* DSCF0002.JPG (311.41 KB, 1280x960 - viewed 345 times.)

* DSCF0003.JPG (322.15 KB, 1280x960 - viewed 366 times.)
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.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #102 on: December 02, 2020, 11:01:02 PM »

More:


Last pic... who could resist his request?


* DSCF0004.JPG (321.48 KB, 1280x960 - viewed 389 times.)

* DSCF0005.JPG (319.19 KB, 1280x960 - viewed 362 times.)

* DSCF0041.JPG (344.07 KB, 1280x960 - viewed 380 times.)
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.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #103 on: December 02, 2020, 11:02:48 PM »

You named it after who?



And remember this clean canvas a few months ago?


* DSCF0040.JPG (328.06 KB, 1280x960 - viewed 358 times.)

* DSCF0003.JPG (310.79 KB, 1280x960 - viewed 361 times.)
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: 1201


This ham got his ticket the old fashioned way.


WWW
« Reply #104 on: December 02, 2020, 11:25:40 PM »


At 50 watts carrier there is not as much heat as I expected. The class C  813 is just loafing along dissipating about 20 watts. I figger the two 813 modulators are dissipating  125 watts total at dead carrier - and they are rated at 250 watts diss.  There is a quiet Variac controlled muffin fan cooling the 813s.

Sure it is not very efficient, but there's no other tube system that is cleaner than a series modulated, class A, transformer-less AM transmitter.    

I still have some minor bells and whistles to optimize, but this rig is by far the cleanest, hifi AM tube rig I have built to date. To produce huge audio and at the same time having DC to light performance - while the highs stay crisp and the lows rock the bass speakers by a tap on the boom is exciting.

Kudos, Tom!  I have been wondering how this project was going, and it sounds like the results with three 813s is spectacular.   I would not bang on the mic stand if I had any mod iron in the path, but series modulation can be so forgiving when subjected to massive transients! 

I can't wait to see more of the details and circuitry.  Now that my processor project is wrapping up, I will be focusing on the prototype 8000 rig, series modulated by a pair of 810.  This is in preparation for the 3CX3000 series modulated rig; I think it will be wise to get the bugs out of a smaller prototype first.  I wonder if Frank or John published the specs on the driver board, or will you be sharing that data? 

Congratulations on another superb rig!
73, Rick
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
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #105 on: December 03, 2020, 12:42:25 PM »

Hi Rick,

Thanks for the comments!

I will send you the latest ver 1.1 audio board. No charge.  

The ver 1.1 regulator board works. I had it in operation but because I needed to change the voltage for experiments and also because it was fragile without any protection circuitry, I used the board as a simple unregulated C-L-C    + - 300V supply.     Since this is a class A MOSFET driver, there's no need for dynamic regulation. The hum is better suppressed with the regulator, but using a C-L-C filter brought the hum to near zero.

I did have some early regulator audio parasitics with the 4D32 version, but they disappeared after I went with the 813s.  

 I'll axe John to post the latest schematics and you can go from there when the audio board arrives.

Bottom line is I have plenty of the 1.1 audio driver boards that are correct except for a capacitor polarity marking..    But I have no extra 1.1 regulator boards.  The 1.0 is unusable because the tracks are too close.  John said he had some extras, so we'll see when he posts here.

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: 1201


This ham got his ticket the old fashioned way.


WWW
« Reply #106 on: December 03, 2020, 03:39:16 PM »

Thanks a MILLION, Tom.  You sure have made some great progress.  I will be listening for you on HF, I assume you will be active on mostly 75 and 40, but with 15 opening up you might be doing some 15 meter AM DX. I can point my yagi your way if your are up there.

My interest right now is getting the driver and modulator section of my prototype working.  I believe you placed the RF deck just under the HV supply, and the modulator between the RF and the minus side of the HV supply,  that is what I will do initially, so seeing how you interfaced the 813s to the driver will be key.  I assume you are running them in triode configuration, with grids tied together?  Or are you using a screen supply?

I really have no need for a regulator board for the driver, as I prefer to use a variac, and I may use either bipolars or FETs just as additional filtering, so just the driver board would be a great help.

Once I get things working with the RF deck "on top", I want to switch things up and put the modulator on top, running as a cathode follower, because that is the way I will configure the 3CX3000 modulator.  It will be more complex to drive, but it also let's me use my dad's 1949 P-P 304T-TL RF deck without modification.  That will be a real treat to mod that deck with a single series tube!  So at that point I may pass the driver on to another who would like to try the series mod rig.

Being in the frosty north, I believe you should turn up the wick, run the 813 series mod rig at max input, without a linyear afterburner, and see how that inefficiency heats the shack.  Inquiring minds want to know.  Perhaps you can turn off the wood stove or other combustion mode shack heaters.  That is where the series mod version is 100% efficient, in the IR band of radiation!
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 #107 on: December 03, 2020, 04:29:31 PM »

Here are the schematics on both boards.

I have a lot of the regulator boards. 20+. They would work well for the other bigger MOSFET board. Contact me by pm or email if you want one. These are a +/- 200 to 300v MOSFET regulator. Fully self contained. Just add a transformer.

On the 1.1 Audio mosfet board, C6 silkscreen polarity is incorrect. Make sure you get that errata or you will have an exciting firecracker bringup. I updated the schematic and the gerbers for v1.2 but we never went for any more boards. We can have more made easily if demand warrants.

John

* GFZ-REG Schematic 1.1.pdf (58.5 KB - downloaded 174 times.)
* Small Mosfet Driver Schematic 1.2.pdf (52.37 KB - downloaded 171 times.)

* board-top.png (64.81 KB, 2452x1503 - viewed 307 times.)
Logged
AG5UM
Member

Offline Offline

Posts: 114


« Reply #108 on: December 03, 2020, 07:29:11 PM »

Tom,
When you say you were not happy with the "linearity and response" and got it working "barely",
Does this mean the 4D32/6LF6 combination does not work? or just that the 813's are so much better?
I really liked the idea of the 4D32/6LF6 combo, does this mean that it was a bad idea?
Congrats on the 813's your test results seem fantastic, congrats.
AG5UM
Logged
w8khk
Member

Offline Offline

Posts: 1201


This ham got his ticket the old fashioned way.


WWW
« Reply #109 on: December 03, 2020, 07:57:56 PM »

I have a lot of the regulator boards. 20+. They would work well for the other bigger MOSFET board.

John, very nice job on both circuits and boards.  At first I was under the impression that the additional board was just a regulator, but since it also has the rectifiers and filters, it would make the project simpler and much easier to implement. 

Are you driving the 813s in triode mode?  If so, are control and screen grids tied together?  I will be in touch regarding your board offer. 
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
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #110 on: December 03, 2020, 08:38:08 PM »

Tom,
When you say you were not happy with the "linearity and response" and got it working "barely",
Does this mean the 4D32/6LF6 combination does not work? or just that the 813's are so much better?
I really liked the idea of the 4D32/6LF6 combo, does this mean that it was a bad idea?
Congrats on the 813's your test results seem fantastic, congrats.
AG5UM



Hi Don,

That's a tough question...   I think the 4D32 X 6LF6s would have made a beautiful rig once optimized.  But at that point in the journey with series modulated rigs I was still trying to find out what worked and what did not.  My first waveforms had bad audio oscillations and the waveform quality was below par.  I had poor high end response and the triangle had wavy lines.  The 6LF6s are fragile and the sockets are weak and easy to ruin the pins after lots of trials and rebuilds.   In contrast I know the 813 well and the sockets are heavy duty and robust for big voltage. Four sweep tubes were a lot to deal with compared to a pair of 813s.  The parts count sucked. The filament power for the whole group was almost the same though.

In hindsight, I'm glad I went with the 813s cuz they seem to handle the abuse of class A better and the waveforms are exquisite.  From what I've read, getting a series modulated rig to work well, especially picking the right tubes is not easy due to lack of info. Maybe I just lucked out with the 813s.  

I plan to increase the plate voltage to 2KV soon.     1500V is fine, but I have been slowly getting higher and higher peak power and cleaner performance the more I play around with parts optimization, tube currents, grid leak values, etc.   There are always sweet spots that we find out after hours of testing.  Even when I first fired up the 813s I had trouble getting a clean waveform above 7 KHz. But after a lot of work I can  confidently set the sinewave frequency to 20KHz, 30 KHz, 40 KHz and higher and see the same perfect waveform. I can drop the  carrier power down and then modulate 300%+ and maintain perfect waveforms. That was my goal - to be able to go anywhere in frequency and hammer the audio and maintain near perfect performance. The RF final, being in class C tunes easily without fuss, just like a plate modulated rig. IE, once the rig is set up, there's very little to worry about in tuning, unlike say a linear amp or screen modulated rig.

But all in all, I think if you went ahead with the 4D32 / 6FL6 rig and have a lot of patience, you will obtain similar results.

Rick:  Yes, I am driving the 813s in class A with the grid and screens tied together. They are being fed with Fran's MOSFET board with about +10 volts DC bias (adjustable) and 160V peak to peak single ended audio. (with + - 300V on the audio board) I went this route because my 813s plate modulated rig uses this method very successfully. To get nice straight triangle waveforms shows that this was the right choice for linearity.  I was willing to use a regulated screen supply and drive the grids, but didn't have to.   My 4-1000A plate modulated rig uses regulated screens and is grid driven in AB1 for modulators, so both methods appear to work well. I've gotten great waveforms from both of those rigs.

My total NFB ladder worked out to be about 230K using (22) 47K 3 watt resistors in series parallel tapped directly off the modulator plates.  With the NFB pot turned all the way up I get a good dose of NFB.

I might add that NFB is very important if you want to beat the audio to achieve the highest peaks possible with minimum distortion.  Without NFB you will just have to lower the audio and accept a little less linearity. But just a touch, like minus 4-6 DB is perfect.

Yes, the regulator board is a nice compact way to get the rectifiers, caps and FETs on one board.

Right now I'm working on trying to stop a key-up surge that is related to the HV coming on and sending a pulse thru the NFB into the driver board.  I've tried a few things and Frank is on the job figgering it out.  I also want to get rid of my HV Variac that has smoked a few times. I plan to add in a 40V boost transformer in the HV xfmr primary to see if I can get nearer to 2KV.  

At one time I had the audio driver and everything on one chassis.   I decided to rebuild the rig with the 813s with the audio driver outboard to eliminate any possible audio feedback. It worked and I see none now. Lots of nuts and screws filling old holes in the back.  But given enuff room and shielding, it would have been fine all in one.

I'll be curious to what you guys come up with.

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: 1201


This ham got his ticket the old fashioned way.


WWW
« Reply #111 on: December 03, 2020, 11:58:22 PM »

Rick:  Yes, I am driving the 813s in class A with the grid and screens tied together. They are being fed with Fran's MOSFET board with about +10 volts DC bias (adjustable) and 160V peak to peak single ended audio. (with + - 300V on the audio board) I went this route because my 813s plate modulated rig uses this method very successfully.

Tom, thanks for the details.   Looking at my options, I believe I will install parallel sockets for 813s and 810s/8000s on my breadboard.  I have the tubes, so I might just test using 813s first, then swap to 810 modulators with a single 8000 final.  That way I can use your configuration as a starting point for verification, then switch to my triodes and compare bias and drive requirements, as well as results.  Interesting here is the fact that the 813 and the 810/8000 require the same filament voltage, and similar current, so just adding sockets allows the comparison to be made back and forth as needed.  

This will all be on a pine board - open construction, we could call it the WidowMaker.  I will be sure to take some pics for the photo gallery of the safest rigs of 2021!  Playing with it a a decent "social distance" will be healthy.  I have a 4000 volt supply with a UTC transformer, variac, and PC board with the rectifiers, filter caps, and bleeders, so it will go together quickly.  It should light up and heat the room nicely!


I wonder if you tied the suppressor grids to the cathode, or to the control and screen grids?  I would assume all three are tied together, but I would really like to be sure.


After all the testing is completed with the RF on top, I will put the RF stage at the bottom, and see if I can achieve similar linearity running the series modulator on top as a cathode follower, possibly using a 4D32 or 4-65 as a driver.   I expect I will probably need a 4-250 as a driver for my 3CX3000, considering the voltages involved.

Once all this testing and data collection is finished, I will strip down the WidowMaker, and design a cooling plenum and duct work for the 3CX3000, with the goal of making it very quiet.  Then I hope I can use it to modulate the Push-Pull 304-TL deck, saving the time and expense of building up another RF amplifier.  These tubes play well with lower voltages and higher currents, keeping the unmodulated HV down to 2KV or lower and still being able to run the limit.  But I hope to put the modulator "on top", and the RF deck down below, so that I do not have to make any modifications to Dad's 1949 RF deck.

Since I will not be putting this on anything but a dummy load until the final rig is completed, I will probably use a 1 MHz rock oscillator and buffer/driver, and monitor it with my Fisher 90T anniversary-series tuner.  This one was a 1957-vintage AM/FM receiver/tuner, and included narrow, medium, and wide IF bandwidth filters, so it will be the audio monitor, along with an SDRplay receiver for looking at the spectrum.  I will also look directly at the RF output with a pickup device, allowing either trapezoid or envelope monitoring directly at the dummy load.  This is smack dab in the middle of the standard broadcast band, but I do not have any notions of playing "pirate radio", unless I have the urge to hear some Steppenwolf "Born to be Wild" or possibly "Magic Carpet Ride".  These bring back some lucid memories from my days in the USAF in Tucson, we were a pretty wild bunch!

For the initial tests, there will be no T/R switching or NFB, just basic circuit testing to move on to the big rig.  Is there anything else I should know about your modulator circuit, or is it all straightforward following the 'GFZ driver board?
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
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #112 on: December 04, 2020, 01:50:05 AM »

Yo Rick -

Sounds like a good plan.  Putting the modulator on top is appealing to get rid of the scary floating HV stuff for sure.  I've gotten used to it from building up several tube PDM rigs over the years.

I connected the suppressors to the cathodes on all three tubes. I figure the rig works so well (very linear)  that I have nothing to gain or measure. It can only get worse, so the cathode it is.

Frank came up with a modification that you will need to use. When I keyed up the rig the HV put a voltage surge thru the NFB ladder into the driver board. The solution was to add a small relay to keep a bias (47K referenced to the -300V supply) on the output FET gate until the key up.  No relay delay needed.   Before the modification I would key up and the rig wud overshoot to 80 watts carrier and then settle down to 40W.  Now it comes up smoothly with barely any sign of overshoot. Wherever I set the idle pot, the carrier stays solid.   He will probably make the mod for the schematic.

I was able to get 80 watts carrier out at 125% modulation. Or 40 watts carrier at 250%.  All clean as a whistle at whatever frequency your heart desires... Wink  The modulators are hardly warm under full tone modulation cuz all the voltage is across the RF final. Even at dead carrier status, the two 813s handle the heat very well. I just have to talk fast and often.  

I found I can run the audio much higher with little distortion due to the NFB.  I's all coming together.  A little here and a little there eventually adds up to a fine machine.

It's good you have a big HV supply to play with. I started with 1500V for the 4D32 plans. Watts are hard to get when the modulator is running class A. I am going to try to bring it up to 2KV and go from there.  90-100 watts of carrier at 130% would probably be the max for 2KV barefoot.   But it will be fun to try it at 10-15 watts driving the big linear at 300%. It will sound like DSB unless someone uses a sync detector. Years ago everyone had diode detectors and you couldn't run that kind of audio without complaints. I know - I ran an upside down 4X1 plate modulated rig.  But it can be done nowadays as long as you find a way to keep the signal contained for bandwidth. I recently played around with Steve/QIX's low level negative peak diode clipper from the PDM board and may again.

I'll bet your 810s will be as good as the 813s.  It all depends on how much time you spend refining it.  Just like a good dog, they are what you make them. Spend a lot of time on them and they both show FB results.

*** Be sure to use non-inductive resistors, no wire-wounds,  for the NFB ladder. Don't axe me how I know.  ***

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: 1201


This ham got his ticket the old fashioned way.


WWW
« Reply #113 on: December 04, 2020, 02:04:21 AM »

Tom, thanks so much for the additional information.  I am glad you are getting such great results using series modulation.  I only hope I can achieve the same clean signal with the 3CX3000 and 304-TLs.

Good news on the relay fix.  I will keep that in mind.  I will probably wait on the NFB until I get the final rig built, as I want to know exactly what is going on with the prototype, and not "hide" anything with NFB that I should be aware of as I move to the next level.

I bet Yaz is proud and honored to have your cleanest rig named after him.  And he also can't wait until warmer weather and sailing season!

I will soon have some photos of the WidowMaker to share with you and the gang.  Will look very old-fashioned, I am sure!

73, OM!
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 #114 on: December 04, 2020, 07:45:10 AM »

On the regulator, you may see a line to the right of the big caps. We made it so you could cut them in 2 and just use the regulator part if you already had a source of unregulated dc.

The FETs mount under the board and those big hole are to access the hold down screws. How I build one is to lay the board on the HS, mark the mounting screws and the center of the big holes with a sharpie. Then go drill and tap all holes. Mounts are 6-32 and fet holes are 4-40. Then bend the fet leads and spread them so they friction fit on the back of the board. Lay the board on the HS, screw down the mounts (on 3/8 or 1/2 nylon spacers). then screw down the fets. Now you can solder the fets in place. After that you can remove the whole board from the HS and the fets will still be in proper relation to the board. On final assembly I add sil-pads for the fets. Keeps it nice and clean.

Here is a pic of the audio board. Note that C6 plus side is connected to -Vdc. That is the error. Just reverse that cap when building. Also note that this one laid out with Tom's preferred FET orientation with the FETs off the side of the board. You will need a correspondingly larger HS for this.

I will touch up the BOMs to make sure they are accurate and post them this weekend.

John


* board-top.png (54.46 KB, 2526x1501 - viewed 354 times.)
Logged
w8khk
Member

Offline Offline

Posts: 1201


This ham got his ticket the old fashioned way.


WWW
« Reply #115 on: December 04, 2020, 01:43:55 PM »

John, I can see a lot of planning and effort went into the driver and regulator boards.  Thanks for all the hard work!
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 #116 on: December 04, 2020, 03:00:36 PM »

I switched to using Kicad. Much better than what I have been using - EasyEDA. More work to find parts but the layout in general is much easier. Plus I have all of my work product on my storage, not a Chinese cloud. Sometimes I have to make my own footprint or import one, not hard to do.

Logged
w8khk
Member

Offline Offline

Posts: 1201


This ham got his ticket the old fashioned way.


WWW
« Reply #117 on: December 04, 2020, 03:23:03 PM »

YES!  810s and 8000s are like bro and sys, but to be politically correct, they are siblings, they do not exhibit any gender attributes.   Hi Hi.

That said, all the specifications and ratings of these tubes are identical, except for the Mu, or amplification factor, and interelectrode capacitances.  The 810s are much higher Mu, (36)  excellent for audio (modulators) while the 8000 is lower Mu, (16.5) better suited for the RF amplifier, as in the modulated final.  The lower Mu 8000 requires less drive, therefore is easier to make cleaner modulation peaks as a plate-modulated class C final.  (Specifications for these two tubes are attached below.)

RCA used to advertise both on the same page of QST back in the mid to late 40s, as the best watts-per-dollar option of any pair of tubes available at the time.

Yes, I believe these should work very well together for a moderately-powered AM rig.  My choice for the prototype will be the pair of 810s (for the increased dissipation in class-A) as series modulators, with one 8000 for the final.  This is prior to moving on to the 3CX3000 modulating a push-pull pair of Eimac 304-TLs.  My brother is considering the build of a series modulated rig with 811s and 812s.

By the way, RCA advertised the higher-Mu 811 along with the lower Mu 812, for the reasons stated above.  In the earlier RCA Ham Tips (available online from my friend, Brian Page, N4TRB) RCA provided a wealth of sample circuits for modulators, finals, and entire vintage homebrew rigs.  

You may find the Ham Tips repository here:  http://n4trb.com/AmateurRadio/RCA_Ham_Tips/rca_ham_tips.htm

John, I will send an email to you shortly.....

* 810.pdf (299.98 KB - downloaded 150 times.)
* 8000.pdf (572.95 KB - downloaded 158 times.)
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
AG5UM
Member

Offline Offline

Posts: 114


« Reply #118 on: December 04, 2020, 09:48:30 PM »

Tom,
Thanks for explaining, I'm still interested in the 4D32/6Lf6's idea, so, thanks for explaining.
Great job, success is a good thing...
The way you label everything on the chassis, really helps someone like me trying to follow along.
Rick,
I'm anxiously waiting to see your pics of "The WidowMaker", sounds cool..
This series modulated transmitter project for some reason has really peaked my interest, interesting transmitter.
I'm following along as best I can....
Donnie
AG5UM

Logged
AG5UM
Member

Offline Offline

Posts: 114


« Reply #119 on: December 05, 2020, 11:57:14 AM »

A couple little questions:
I would probably be using push-pull tube audio (although I see the mosfet driver board is beautiful and the logical choice),
question, how much of the beautiful cleanwave test results are due to the wonderful mosfet board? (Compared to other audio)
Question: Do I understand correctly, that its important to first get the unit "Right", and then work on the NFB to perfect it?

Also, thanks for answering my simple questions..I'm still back trying to understand/learn the series modulation design.
I'm also trying to keep up with everything your doing to get those great waveforms.....
peak limiters,courtesy filters,neg feedback board, mosfet driver boards....etc....
I'll try to keep quiet and just follow along.
Donnie
AG5UM


Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #120 on: December 05, 2020, 01:17:19 PM »

A couple little questions:
I would probably be using push-pull tube audio (although I see the mosfet driver board is beautiful and the logical choice),
question, how much of the beautiful cleanwave test results are due to the wonderful mosfet board? (Compared to other audio)
Question: Do I understand correctly, that its important to first get the unit "Right", and then work on the NFB to perfect it?

Also, thanks for answering my simple questions..I'm still back trying to understand/learn the series modulation design.
I'm also trying to keep up with everything your doing to get those great waveforms.....
peak limiters,courtesy filters,neg feedback board, mosfet driver boards....etc....
I'll try to keep quiet and just follow along.
Donnie
AG5UM


Hi Donnie,

Good questions...  

Every component in the audio chain is important, as well as having the class C stage set up right.   You can get great waveforms out of a tube driver too, and I almost did it that way. But the MOSFET audio driver is used in all of my AM rigs here, so I built a custom one for each. (except for the Hammond/External Amp driver in the 4D32 Summer Breeze rig)   It either works or it doesn't, already perfected, whereas a tube driver wud take a lot of time to perfect the first time out.

Besides cleaning up audio parasitics or RF instabilities, the hardest part is really making the modulators put out big peak power and stay linear with good harmonic distortion numbers - and also keeping the RF stage accepting this modulation and producing good IMD.. and having a great audio freq response to show it.  It's really a matter of running the basic tests over and over, trying different parameters and parts values until it is tweaked in.  But I do think the MOSFET driver will outperform most tube audio drivers unless the tube driver has extra work done like class A, p-p, NFB and close attention to linearity, regulation, etc. No transformers - thus R/C or DC coupled.

Yesterday I got tired of my smoking HV supply Variac. It was fine at 1200 Volts and the original peak power ratings, but I removed it and now have 1750 VDC and all is well.  I use a step start in the HV so there is a measure of safety having voltage only when keyed.  I tend to build a rig and then demand 50% more power from it later on.  I was even toying with the idea of adding a second 813 to the final and upping the air flow on the two modulators, but so far have sat on my hands... Wink

Yes, the rig should first be optimized without the NFB. Make it as good as possible. Then use NFB to extract the last bit of clean heavy modulation and low/high end freq response at the end of optimizing. NFB is not a band-aid, rather more a way to enhance performance that would not be there without it.  

You may be disappointed in the carrier power output if you are looking for 250% audio. It's all about peak headroom available.   I am now running this rig at about 90-100 watts carrier out at 130% modulation, and about 50 watts carrier at 200%.   In comparison, think in terms of a plate modulated single 813 having X2 voltage  (X4 power) when plate modulated.   The 1750V on this series modulated  813 final is all we have to work with, so there may be only 600V across the RF final and 1200V across the modulators during dead carrier   and 1750V across the RF final during full modulation.    A plate modulated rig with a mod xfmr can supply  4KV+ under the same situation.  But with series mod there's no mod xfmr, mod choke and mod cap to worry about -  so, no free lunches in the end.

The audio driver is up to you, but I prefer to focus on the main mod and RF final components.  Just like using an FT-1000D as the 5 watt RF driver.  I could build a tube VFO and RF driver, but decided not.

Keep the questions coming. We all learn from them.

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.
w9jsw
Two shots of Whisky
Contributing
Member
*
Offline Offline

Posts: 592



« Reply #121 on: December 05, 2020, 03:44:09 PM »

BOMs for both boards...

I called it a PW Mosfet. If you add another 813 it will need to be renamed as it will no longer be Piss Weak...   Grin   Roll Eyes

* gfz-reg_bom_1.0.xls (8.5 KB - downloaded 158 times.)
* PW Mosfet_bom_1.0.xls (9 KB - downloaded 149 times.)
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #122 on: December 05, 2020, 06:51:45 PM »

Rick axed me some email questions about the 813 parameters I run.  I thought it would be helpful to some builders here.



Where applicable, with one 813 RF tube in the final and two in the modulator at 1750 volts HV, dead RF carrier at about 60 watts out:


I’m using about 6 – 10 watts max of drive with a 10K 20 watt WW resistor for grid leak.  No fixed bias.   I key the 1750V HV on and off using a step start.

The grid current (controlled by the input RF drive)  is set to about 15-20  mA no matter what carrier power I am using.  You need the heavy grid current to reach maximum audio peaks regardless of the carrier level.  I usually run a little more grid current than the data sheets.

** I'm using a 3K non-inductive 40 W swamping grid resistor thru a .001 cap to ground.  It's a small chip that  is heatsinked to the chassis.  Nice.

Screen dropping resistor: I tap off the RF final 813 just before the RF plate choke (DC HV) with a 12K  30 watt WW resistor for the screen voltage.   About 40 mA, depending on the HV level, loading, etc.  With a fixed screen resistor, there is no changing it on the fly.    I have no protective bias cuz when the plate voltage is gone, the screen is gone too.  (using keyed HV step start)

I run about 100 - 175 mA plate current (at dead carrier) most of the time.  It depends where the audio board DC bias is set which sets the carrier level. (called the idle pot)  This audio board feeds the modulator grids and screens tied together.  I run the carrier output between 20 – 90 watts or so with maximum peaks reaching the same level.   IE, for 20 watts carrier I can easily modulate over 300% positive.  As the dead carrier setting gets set higher, I reach the same peak power on modulation  -  and  the % of modulation is lower.

1000V across the  modulators  at 150 mA = 150 watts dissipation in the 813 modulators.  So a pair modulating a single one is a good choice for all around use.

Use the GFZ driver board rather than a solid state amp and a transformer.  Since you have a choice, might as well choose transformer-less throughout the whole rig.  The NFB will haunt you later if you don’t.

When using my 813 design, continue to remind yourself  that there are many points in the rig at HV potential including the metering, HV fil xfmr, grid leak, screen resistor and assorted circuitry.

I'm staying with ONE 813 in the final.

John/JSW:  Great on the BOMs! It makes it so easy when logging into Mouser.

T


* DSCF0018.JPG (324.68 KB, 1280x960 - viewed 344 times.)

* DSCF0023.JPG (324.61 KB, 1280x960 - viewed 329 times.)

* DSCF0033.JPG (328.24 KB, 1280x960 - viewed 327 times.)
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: 1201


This ham got his ticket the old fashioned way.


WWW
« Reply #123 on: December 05, 2020, 07:21:10 PM »

BOMs for both boards...

I called it a PW Mosfet. If you add another 813 it will need to be renamed as it will no longer be Piss Weak...   Grin   Roll Eyes

Thanks for the BOMs, John...

I know you call it PW now, but with either one or two 813s, it is certainly not Piss Weak.  So maybe you might want to differentiate by calling it the SE version, for Single Ended.  The AudioPhools will certainly know what you mean.  If it turns out I can use it for the 3CX3000 modulator, I am sure that one will NOT be PW!

No need to rename the BOM file!
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
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


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


« Reply #124 on: December 06, 2020, 02:01:31 PM »

I put Yaz into the lineup. Everything is working.  Here's a review of the rig harem.   I plan to cycle thru them each day to refresh my memory.  It's easy to forget how to tune and operate each one, so I have no choice... Grin


You will see Fabio II  the 4X1 X 4X1s plate modulated,  Hollywood the pair of 813s X 813s plate modulated, Yaz the 813 X 813s series modulated, Summer Breeze the 4D32 plate modulated by a Hammond xfmr and SS amp, Baby Blue a single 3-500Z linear, Rico Suave II the pair of 4X1s in linear,    Fabio I the single 8877 linear and Dr. Love.

All working on the same PTT system and coax switching to select any rig in seconds.

T


* DSCF0003.JPG (322.6 KB, 1280x960 - viewed 334 times.)

* DSCF0001.JPG (329.58 KB, 1280x960 - viewed 301 times.)

* DSCF0022.JPG (327.12 KB, 1280x960 - viewed 312 times.)
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.
Pages: 1 ... 4 [5] 6 7   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 0.097 seconds with 19 queries.