4-1000 grid driven design

[Terry N4RQ]

  Anyone have a good design for a 4-1000 grid driven amplifier?? have searched the internet all I can find is grounded grid.  Any help would be great
Thanks 73's Terry N4RQ

[k7yoo]

I believe you will find one in the 15th edition of the Editors & Engineers (Wm Orr) handbook. It is a very short article with cryptic disclaimer about it being an engineering exercise and too high powered for ham use! Ya gotta love it!!
The 431D1 FRT 24 Collins transmitter was a single grid driven 4-1000. Schematic available.
1200 watts of carrier with no sweat.
Skip

[AMLOVER]

Terry,
Check this   http://207.45.187.74/~wa4kcy/page7b.htm

Stefano

[K1JJ]

Anyone have a good design for a 4-1000 grid driven amplifier?? have searched the internet all I can find is grounded grid.  Any help would be great
Thanks 73's Terry N4RQ

Terry,

Is this amp for RF LINEAR service?   If so, I would recommend placing it in GG service simply for the cleanliness and spectral purity advantage.  GG gives an inherent negative feedback, thus is cleaner but harder to drive. (and more stable too)

Keep both grids grounded for RF, but put 500 volts regulated on the screen. This will generate less grid current (more screen current) and improve IMD.

From actual building experience and hearing others on the air, in grid driven service, the tube needs to run very conservatively with no grid currrent (and regulated to be sure) and recommended REGULATED screen current to approach the -30db 3rd order IMD figure.  (AB1 service) In this mode, the tube will not put out nearly the power as in GG.  I've heard a number of these configurations where the tube was pushed into grid current to get a "normal" power out for a 1KW tube and it was dirty. Splatter.  Unfortunately, we must draw grid current to bring a 4X1 to life and put out serious power.

One solution is to put RF negative feedback around the grid driven stage. This helps a grid driven RF amp immensely.

In contrast, a 4-1000A in GG can be pushed to to its plate dissipation and still remain reasonably clean. (AB2 service) I've owned a few 4X1's in GG and never had a bad splatter report. I've had a few grid driven amps and always had problems to the point of tearing them down for GG config. (Because I tried to get some reasonable soup out of them)

One thing to remember - A grid driven tube (tetrode)  requires recommended screen current to be clean. (Or whatever screen curent is drawn for a given power output)  I used to think that no grid current AND no screen current were best. But recommended screen current runs the most linear.


None of this applies for a grid driven class C amplifier in plate modulated service. Just run recommnded grid and screen current and it will be clean if set up properly..

T

[KE6DF]

If you want to use the tube as a grid driven AB1 linear amp them take a look at this Eimac application note as it has a lot of info on this use:

http://w5jgv.com/downloads/Eimac-appl-9.PDF

Also the older glass tube care and feeding might be useful:

http://w5jgv.com/downloads/Eimac-appl-8.PDF

See the last page of the first one (note 9) for comments on using Eimac tubes in GG.

[KM1H]

I used to build swamped grid driven one and two hole AB2 4X1 amps to order when local BC pulls were available. Bias was set so just a bit of Ig was visible on voice peaks so this wasnt deep into AB2 but more than AB1 where efficiency sucked at a really zero Ig.

With layout care they were stable and RF clean with high 20's IMD, about the same as most rice boxes and sweep tube exciters.

Ive built them GG but they are drive hogs, especially with pulls, that require lots of HV to be worth the effort and can still be hard to tame on the higher bands.

No special circuitry required. Start with any single 4-400 amp as a guideline, calculate the tank using GM3SEK's spread sheet. http://www.ifwtech.co.uk/g3sek/

Calculate the driving voltage available, use suitable swamping values to get the drive required to the grids and let it rip. SS rigs with an ATU can usually handle a 200 Ohm input or get one of those LDG tuners. They used to make just a PC board version that fit right into an amp, dont know why they dropped it....you might ask them.

The biggest problem with 4x1 homebrew is sloppy layout and wiring. The tube loves to become an oscillator.

Carl

[W2PFY]

Terry,
Check this   http://207.45.187.74/~wa4kcy/page7b.htm

Stefano

My browser reports this link as a problem? It says you might pick up crud if you visit. This is not an attach on the poster or on WA4KCY. His stuff is great.

[K1JJ]

Kind of off-subject, and not talked about much, but something to keep in mind when running QRO... (ssb or AM)

For sake of example, let's say that the average LSB station is running 100 watts and they all have an average of -30db 3rd order IMD.  So, when we tune across their signals when very loud, the average station has, say, side crud products that are S9 when 3.0 kc away from center.  

Now lets say a  station gets on running 1KW and has the same amplifier purity, -30db IMD.  That 1KW station's side crud products are going to be at S9+10db over in the same situation. (everything is 10db louder)  Over time, conditions average out. Sometimes we are stronger and sometime weaker as a result - with the average being normal.  Now, anyone operating within 3kc of this 1KW station over time is going to say he interferes, or has worse side crud than the "average" station running 100w.  No one knows how much power is being run by anyone and most don't bother to see how far down the crud is in DB from the center signal.  They simply see and hear splatter.  I've been there - on both ends.

So what I'm saying is when we build a QRO amplifier, we have a big responsibilty to do everything possible to make it CLEANER than the average bear....  If not, then take the consequences when the band gets crowded and our signal propagates BIG into an area of whiners.  If we're running 5db more power than the average station, then it better be 5db CLEANER than the average station JUST to blend in.

There's many ways to make an RF linear amplifier run cleaner - good supply(s) regulation, heavy plate loading, high quiesent idle current, the driver is very clean, the pi-network is set up for the best load, RF negative feedback, picking a good tube designed for linear service, and most importantly, the overall system (driver and amplifier) has been tuned/optimized  using a spectrum analyzer for their IMD sweet spots based on the aformentioned parameters.

T

[KE6DF]

Terry,
Check this   http://207.45.187.74/~wa4kcy/page7b.htm

Stefano

My browser reports this link as a problem? It says you might pick up crud if you visit. This is not an attach on the poster or on WA4KCY. His stuff is great.

My browser didn't complain and the page came up with no problems.

However, I'm running Linux with the Firefox browser and hardly any viruses will cause me problems so I tend not to worry about them.

[W2VW]

Great advice. In addition someone using AM should make an extra effort to keep the exciter from being a trash generator.

Kind of off-subject, and not talked about much, but something to keep in mind when running QRO... (ssb or AM)

For sake of example, let's say that the average LSB station is running 100 watts and they all have an average of -30db 3rd order IMD.  So, when we tune across their signals when very loud, the average station has, say, side crud products that are S9 when 3.0 kc away from center.  

Now lets say a  station gets on running 1KW and has the same amplifier purity, -30db IMD.  That 1KW station's side crud products are going to be at S9+10db over in the same situation. (everything is 10db louder)  Over time, conditions average out. Sometimes we are stronger and sometime weaker as a result - with the average being normal.  Now, anyone operating within 3kc of this 1KW station over time is going to say he interferes, or has worse side crud than the "average" station running 100w.  No one knows how much power is being run by anyone and most don't bother to see how far down the crud is in DB from the center signal.  They simply see and hear splatter.  I've been there - on both ends.

So what I'm saying is when we build a QRO amplifier, we have a big responsibilty to do everything possible to make it CLEANER than the average bear....  If not, then take the consequences when the band gets crowded and our signal propagates BIG into an area of whiners.  If we're running 5db more power than the average station, then it better be 5db CLEANER than the average station JUST to blend in.

There's many ways to make an RF linear amplifier run cleaner - good supply(s) regulation, heavy plate loading, high quiesent idle current, the driver is very clean, the pi-network is set up for the best load, RF negative feedback, picking a good tube designed for linear service, and most importantly, the overall system (driver and amplifier) has been tuned/optimized  using a spectrum analyzer for their IMD sweet spots based on the aformentioned parameters.

T

[K1JJ]

Dave-Made said:
"
Great advice. In addition someone using AM should make an extra effort to keep the exciter from being a trash generator."

Good point Dave. This is overlooked by many of us.  Once upon a time back in the 80's I was using a pair of 4-1000A's in GG. It was a reasonably clean amp as most GG amps can be.  It needed a little more drive to see its full potential, so I added in a GG 4-400A IPA.  (intermediate amplifier)

I remember getting on with Chuck to do some pecker matching into Eu. After a few minutes of cracking the whips, a guy broke in and yelled at us for chewing up the band. Of course I figgered it was Chuck doing it and he figgered it was me, so we continued on.

I figgered the 4X1 amp was OK and didn't even consider that maybe my drivers were dirty. Anyway, later on out of curiosity I put a scope probe on the 4-400A IPA and put through a tone. When pushed hard for pecker matching, the IPA was square-waving!  I had the IPA amp loaded way too lightly. I should have noticed it on the final output, but didn't in the heat of battle.

So, bottom line is alway check each stage of amplification under pecker matching conditions to be sure. We sometimes assume our drivers are clean, but they can become dirty real fast if the have variable pi-networks loaded too lightly (tubes) or even riceboxes that are run with ALC on or are matched poorly.  

It's all a system, everything from the mic to the ant. Heck, even the antenna can arc slightly and make the rig a spark gap transmitter. It happened to me once.


The other night I finally hooked up a SoftRock to a computer in spectral analyzer mode and optimized my FT-1000D and Dr. Love linear. (using the pan-waterfall screen with a 3kc mask is the best method)  First thing I saw was my opposite sideband suppression was poor! The standard scope will not show this. After fixing that, I then put through some programmed voice material into the dummyload and retuned everything until the IMD improved further. There IS a set of sweet spots that a system will like to suck in that last bit of crud from splattering.

This lashup has been reasonably clean for 2 years now, but using the spec analyzer, for the first time in years I can get on and not WORRY about some guy telling me there's crud 4kc away. I can SEE my tightly defined bands of 2.8kc audio (LSB) brickwalling whether I am on the exciter or when using the linear amp. I never thought it was possible to have a QRO system as clean as the FT-1000D barefoot, but it is...   The amplifier is CLEANER than the FT-1000D since the 1000D is about -34db 3rd and the amp is somewhat better than that. The result is a signal that is maybe 1db less clean than a barefoot driver. Not too shabby.

It's easy to build up a big mo-fo linear amp. The HARD part is working to set everything up afterwards to make it clean. That was an impossible dream I pursued for many years with poor results until recently. The key was selecting a tube designed for linear amplifier  service and using a driver that was well overrated for the job.  Running our drivers and final amplifiers at ˝ power or less of rated specs does wonders for cleanliness. (Yes, I run my amps at about 1/2 power max)

Tuning the band with the softrock spec analyzer in pan-waterfall mode using a 3kc mask is quite a trip once the eye gets acclimated to what  a “normal” signal should look like. (All of this discussion applies to both ssb and AM -  just add a few kc width for AM)

T

[KM1H]

And all this from a guy who keeps having intermittent audio, beats on the rig and it moves up the band about 3kc every smash and meanwhile merrily and blissfully in locked mike mode yapping away as he is heading into slopbucket land

I wish I had it on tape!

Couldnt resist Tom

If you want a clean linear use a linear tube; 3CX800A7, 8877, 3CX3000A7. YC-156, even 3-500Z's at 2000V. Old design tubes will be hard pressed to get beyond the low 30's 3rd order on a good day and the 5th is often almost as bad.

BTW, your rig sounded great when it was working and the nice steady signal was a big help in checking the performance of the new Beverages Ive been cranking out. I was laughing so hard I didnt dare try and break in on the TS-950SDX. And then there was Frank with his 27MHz echo box....

Watch it, Im on a roll....rolling right up to bed soon.

Carl

[WA1GFZ]

That was priceless Carl. Every time Tom smacked the rig it moved up a couple KHz. I was watching him on the spectrum display. The echo is the HPSDR audio delay. I was playing with my KAZ antennas. and forgot to mute it.

[K1JJ]

heheheheh.... HIHIHIHIHIHI - FB OM. How's the XYL, by golly?  (Just trying to divert attention and cover up) 

Yep, I forgot that the homebrew VFO sits on top of the banged rig. Ended up around 3900 talking to myself.  I'll bet the hornet nest above us just loved it....

This has happened a few times. The VFO dial has a roller stub sticking out that catches on my shirt when I walk by. I've jerked myself up 100kc away at times. Gotta fix that. Gotta love it.

What you said on the tube choices. I realize that many of us use what we have in the junk box. Many of these older tubes were designed for class C service years ago. The newer linear design power grid triodes are expensive and harder to find, so we use what we have.  But I wish I had paid up years ago and bought one of the tubes on your list. It wud have saved me mucho time and effort in the long run.... sigh.

T

[W2VW]

ahem.

[W2VW]

I was thinking of stupid Ranger tricks. Underload the final for less power or use the Junkston attenuator into a linear amp. Neither one should be on the air at legal limit.

[W2PFY]

I was thinking of stupid Ranger tricks.

I once unloaded my BC-610C and was able to make it run at 5 watts output. It may be a record low output from a BC-610

[W2VW]

That's an accomplishment.

[KM1H]

What color was the 250TH?

[W2PFY]

What color was the 250TH?

Blue, the tube was running so cool that ice was forming at the plate cap

I actually set it in the tune up position and decoupled it to the max. I could still modulate it. Was in there with a bunch of guys and they said it sounded horrible