Warning: Creating default object from empty value in /homepages/11/d132647312/htdocs/Amfone/mkportal/include/SMF/smf_out.php on line 47
Need Some Help With My Amp Project




 
The AM Forum
April 04, 2020, 02:06:13 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 [2]   Go Down
  Print  
Author Topic: Need Some Help With My Amp Project  (Read 1784 times)
0 Members and 1 Guest are viewing this topic.
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8104


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


« Reply #25 on: March 11, 2020, 03:27:10 PM »

Hi Glenn,

A few years ago I built up a 4CX-350J in class A/AB1 linear service.  It was part of a class A/AB1 linear chain that started with a tap off the FT-1000D at 100 mW, into a 1 watt lab amp, into the 4CX-350J, and into a pair of 8877s.   I was able to get -55DB 3rd order IMD numbers, which is incredibly clean for an analog amplifier system.

The "J" version of the 4CX-350 is at least 10-15 DB cleaner than the standard models, but costs more too. It is the optimized "linear" version, probably demanded by medical or military service when IMD was important.  BTW, if you are using an SDR radio with high-level-derived predistortion then the "F" tubes will work just as well.



** Some of the following may or may not apply to your situation, but in general:


What I did:  

I built up a simple adjustable current overload sensor - controlling a separate relay for the screen and grid that shut down these two elements instantly when either is overdriven.

I used a string of diodes in the cathode lead to supply grid bias... super stable voltage.

I used a well regulated (electronically-regulated) screen supply. You will want to run recommended screen current for best linearity, thus the need for exquisite screen voltage regulation.

Being in class AB1, there is no grid current, ever.

Watch your plate voltage for any sag and take appropriate measures to keep it stiff thru heavy plate supply filtering and a 240VAC feed, etc.

You basically need to follow all the rules of linear tetrodes to keep them safe from harm and clean for minimum IMD.  Run two-tone IMD tests thru until you are satisfied with at least -35 DB 3rd for "F" tubes -   and -45DB+  3rd IMD for "J" tubes..

Load the amplifier heavily (less C2 mesh in the loading cap) to get the best IMD numbers.

If you break any of these rules, it is very easy to have a splatter-maker -  or destroy the fragile grid or screen.

I realize you will be using the "F" version of the tube and not the "J", but I would put out feelers for the "J" tubes and ya never know what you may find cheap.... Grin  Keep ads out in the ham wanted listings.   Worst case, the "Nebraska" parts place had them new when I built mine a few years ago.

Good luck with the project, Glenn!

Tom, K1JJ



Logged

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

"You know my name"  https://youtu.be/noGjJyEDm5s?t=135

There's nothing like an old dog... a puppy... a dog in its prime... or ANY dog!
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2133


Making AM GREAT Again!


« Reply #26 on: March 11, 2020, 08:01:33 PM »

Tie the three grids together.

Run your rf drive to one side of the resistor.

Ground the opposite side.

Connect the rf drive side of the 50 ohm resistor to the Junction of the three grids with a couple .1 uF discs.

From the Junction of the disc's and the grids, you apply bias voltage via an rf choke.

Hope this makes sense.  I have a pair of 250B tubes doing this now. 

--Shane
KD6VXI
Logged
kc2we
Member

Offline Offline

Posts: 90


« Reply #27 on: March 11, 2020, 10:44:08 PM »

Look at the TMC PAL-500 amplifier. Uses two 4CX350A's. They use separate bias supply feed for each tube to balance idle plate current. Also, TMC used a parasitic suppressor in each grid. The grids are tied together for RF but actually dc isolated by blocking capacitor. This amp uses a 6CL6 and 6146 driver. That was because it was designed for a drive level of 100 mw for the military. Where the tube driving circuits are, use the 50 ohm load instead. I suspect that to 4CX350's tending to "take off", so they added the Parasitic grid chokes. Usually a few turns of tinned wire across a 1 W 47 or 100 ohm carbon resistor. In a Collins design with 4CX250's (TRC-75) they just tied the grids together. Maybe the 4CX250"s aren't as touchy.   ST
Logged

Seth Taylor
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8104


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


« Reply #28 on: March 12, 2020, 11:44:38 AM »

Look at the TMC PAL-500 amplifier. Uses two 4CX350A's.  This amp uses a 6CL6 and 6146 driver.


I used to have a TMC PAL-500 linear amplifier here.  I especially liked the RF negative feedback from the plates of the 4CX-350As back to the earlier driver stage. Nice way to clean up the 4CX-350s into something even better.  I even added a little more NFB by increasing the feedback capacitor value.  It became a little harder to drive but remained stable, so I knew it was working.  This will decrease IMD.   Of course, negative feedback requires a driver stage which is not planned in this thread project...

And yes, I used parasitic suppressors in the grid circuit too. (as well as in the plates)   Nice precaution.

T
Logged

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

"You know my name"  https://youtu.be/noGjJyEDm5s?t=135

There's nothing like an old dog... a puppy... a dog in its prime... or ANY dog!
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2133


Making AM GREAT Again!


« Reply #29 on: March 12, 2020, 07:16:17 PM »

The need for parasitic suppressors on the grids can be pretty much removed by grounding the screen at DC

You hook the positive side of the supply to the ground and the negative return to the cathode.

This does nothing for imd, you'll still need a regulated supply for that.  However, it sure does give a high gain tube a hell of a lot more stability.

I can't claim credit for this method, but it sure cured a partially stable 4cx5000 I built years ago.

That used a 3 turn 9:1 input.  140 in would net 14 large into the dummy load.

Still ran PSupp on the anode.

--Shane
KD6VXI
Logged
kc2we
Member

Offline Offline

Posts: 90


« Reply #30 on: March 12, 2020, 09:23:32 PM »

Use a dc blocking cap between 50 ohm load and bias feed. Feed bias through RF choke. See the TMC PAL-500 amplifier schematic. They use a driver tube, but adapting the input circuit for swamped input dictates the DC block. 
Logged

Seth Taylor
W2BX
Member

Offline Offline

Posts: 23



« Reply #31 on: March 13, 2020, 01:50:50 PM »

Thanks to all who replied. Some great info here.... and a special hello to Master Vu on the channel, cawman!  Grin. Some of you may not know that Master Vu only likes "fresh sushi". Hello Tom, it's been a long while.. Almost 15 years since I moved to NC from NY.
Hope you're doing well. I'm hoping to get back on AM soon.

I'm pretty sure I can make -35 IMD by running a stiff, reduced voltage plate supply and a stiff well regulated screen supply as mentioned.

Does this look right? (attachment) ... and one more thing... I'm using SK630 sockets which directly DC grounds the cathode on pins 2,4,6 & 8.





* grid schematic.jpg (508.85 KB, 1300x731 - viewed 57 times.)
Logged
kc2we
Member

Offline Offline

Posts: 90


« Reply #32 on: March 13, 2020, 04:28:55 PM »

Yes, like the sketch. OK. You get fancy and feed bias separately to each tube to balance idle current. That would require a dc block for each grid. Since there's no grid current, series resistance value in the bias supply isn't important. The individual adjust is closer to what TMC did. ST
Logged

Seth Taylor
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 7469



WWW
« Reply #33 on: March 16, 2020, 12:00:56 AM »

The need for parasitic suppressors on the grids can be pretty much removed by grounding the screen at DC

You hook the positive side of the supply to the ground and the negative return to the cathode.

This does nothing for imd, you'll still need a regulated supply for that.  However, it sure does give a high gain tube a hell of a lot more stability.

I can't claim credit for this method, but it sure cured a partially stable 4cx5000 I built years ago.

That used a 3 turn 9:1 input.  140 in would net 14 large into the dummy load.

Still ran PSupp on the anode.

--Shane
KD6VXI

YES!! double plus good on grounded screen for DC (and RF IMHO), with regular element-to-element operating voltages, doing wonders for stability! Let that screen be a screen!
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
W2BX
Member

Offline Offline

Posts: 23



« Reply #34 on: March 16, 2020, 07:12:32 AM »

The need for parasitic suppressors on the grids can be pretty much removed by grounding the screen at DC

You hook the positive side of the supply to the ground and the negative return to the cathode.

This does nothing for imd, you'll still need a regulated supply for that.  However, it sure does give a high gain tube a hell of a lot more stability.

I can't claim credit for this method, but it sure cured a partially stable 4cx5000 I built years ago.

That used a 3 turn 9:1 input.  140 in would net 14 large into the dummy load.

Still ran PSupp on the anode.

--Shane
KD6VXI

YES!! double plus good on grounded screen for DC (and RF IMHO), with regular element-to-element operating voltages, doing wonders for stability! Let that screen be a screen!

Ok, I'm missing something... how do you dc ground the screen if you need to supply +350vdc to the screen??. RF ground yes, through the built-in bypass caps in the socket. Just to clarify, this is a grounded cathode grid driven design.
 
Logged
W1ITT
Member

Offline Offline

Posts: 347


« Reply #35 on: March 16, 2020, 08:04:54 AM »

Interelectrode voltages are all relative.  It just depends on where you define your reference point.  If you set the screen at DC ground (chassis), and then you set the cathode at -350v, you have a DC ground at the screen, but it's still 350 volts above the cathode and the tube runs like a normal tetrode.  And you similarly have to adjust the reference point of the control grid and the anode.
If you have one of the old W6SAI "West Coast Handbooks" he does a good job of showing and explaining the technique.  I ran a 4-1000 tube in this configuration for many years and it was stable as a church.  The screen bypass capacitor can be a source of problems as it doesn't always look like a pure capacitor at all frequencies.  Bolting the screen right to terra firma solves that.  You have to pay attention to metering but it all becomes quite simple to implement.. 
By the way, the Bill Orr W6SAI handbook is definitely worthwhile for the amplifier builder and they show up often at hamfests, assuming we ever get back to having hamfests again.
73 de Norm W1ITT
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2133


Making AM GREAT Again!


« Reply #36 on: March 16, 2020, 08:23:47 AM »

I found my Billy Orr book at Barnes and Noble.

Was somewhere in the neighborhood of 30 bucks.

If you are going to have the screens floating, it really takes a ton of bypass caps to do the jib right, and then it's still unstable at times.

Grounded screen is the hot ticket.

And make sure you ground with nice wide strap.  None of this #22 hookuo wire crap.

--Shane
KD6VXI
Logged
W1ITT
Member

Offline Offline

Posts: 347


« Reply #37 on: March 16, 2020, 09:43:32 AM »

There's a bunch of the William Orr Radio Handbook on Ebay.  I quit counting after ten.  I have the brown one and the gray one and I'm keeping them both!
73 de Norm W1ITT
Logged
W2BX
Member

Offline Offline

Posts: 23



« Reply #38 on: March 16, 2020, 12:35:01 PM »

I'm going to stick with the tried and proven design of grounded cathode and +v on the screen. I have the Bill Orr Radio Handbook 23rd edition right in front of me. On page 17-2 is "A universal Grid-driven tetrode amplifier". This design, like every other 4CX250, 350, amp I've seen, is a +V on the screen, grounded cathode design. I'm not doubting a grounded screen may be a better approach but..... considering the arrangement of the Eimac SK630 sockets I'm using, a low inductance path from screen to ground is not possible. If there was a socket that connected the finger stock screen contacts directly to ground and floated the cathode connections... then yes. Considering I'm already well into metal work and socket wiring, I'll stick with the original plan.

Shane, the screen won't be "floating" since it will be connected to a very stiff well regulated supply.
Logged
kc2we
Member

Offline Offline

Posts: 90


« Reply #39 on: March 16, 2020, 12:37:52 PM »

Check http://www.nj7p.org/Manuals/PDFs/Books/orr_radio.pdf

1959 Editor's Engineers 1959 Radio Handbook, pages 629-631. There's a 4-400A amplifier with a grounded screen application. It explains how to do this. I scanned the pages on to a three page word doc to show here. Better quality by downloading the pdf from NJ7P. Although this design is a grounded grid cathode driven, the screen circuit can be replicated by itself while still driving the grid.




* EE HB 629_631.docx (482.46 KB - downloaded 8 times.)

* EE 59.jpg (547.48 KB, 1876x2604 - viewed 46 times.)
Logged

Seth Taylor
Opcom
Patrick J. / KD5OEI
Contributing
Member
*
Offline Offline

Posts: 7469



WWW
« Reply #40 on: March 16, 2020, 01:48:50 PM »

and
https://archive.org/details/books?and%5B%5D=radio+handbook&sin=
Logged

Radio Candelstein - Flagship Station of the NRK Radio Network.
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2133


Making AM GREAT Again!


« Reply #41 on: March 16, 2020, 01:51:28 PM »

Your screen is still floating for rf depending on the resonances of your screen ground caps with a screen floating from the chassis.

The DC supply doesnt effect the rf stability, other than changing the gain characteristics of the tube.

By grounding the screen you end up with a very stable amplifier.

By utilizing the capacitors that run from screen to ground you may or may not have a stable amp.  Those caps will have resonances, and if your layout exacerbates the problem, the tube may take off.

With a screen at ground potential, the grid and cathode are nearly isolated from the anode.  Hard to take off like that.

Either way works, but as I said.  I had a 4x5 that was squirrelly.  I changed it to grounded screen.  This involved tying the screen directly to ground, having the positive side of the screen supply tied to the ground and the negative tied to the cathode of the tube.  The screen then isolated at all frequencies.

Both methods work.  However, the grounded screen works mo better with high gain tetrodes, especially multiband.

AG6K, Rich Measures (SK) is the one who turned me on to this method.  He published it in his website.

--Shane
KD6VXI
Logged
W2BX
Member

Offline Offline

Posts: 23



« Reply #42 on: March 16, 2020, 02:53:42 PM »

I appreciate the info, and I understand the concept for the sake of discussion but a grounded screen configuration is just not practical with 4CX350's (this project) and the sockets they are designed to be used with. A 4-400a is a tetrode but a different animal. (besides the anode, all the connections are on the bottom ...big fat heavy thick pins)

You mentioned you're working on a 4CX250 amp, you must be aware that the Eimac sockets don't lend themselves to a grounded screen config. You would need a custom socket with fingerstock that grounds the screen ring on the tube.
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2133


Making AM GREAT Again!


« Reply #43 on: March 16, 2020, 05:54:31 PM »

You can make clips, but depending on the socket that becomes a problem.

Or, ground pin 1.

The Russian sockets that will fit a 250 make it easy ttto modify the socket with spring clips.  They are an erector set.

--Shane
KD6VXI
Logged
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 0.069 seconds with 18 queries.