Best method of switching control on an AM TX

[W4RFM]

When desiging my transmitter (which will be AM only) should I plan to switch the RF plate and Modulator plate on an off on the transformer primary side with a relay arrangement and leave the screen  B+ on the 4-125 RF final, or is their a better way? (Triode modulators). I have a supply for the modulator B+, separate from the RF B+ and the screen supply for the RF final is also a 3rd separate supply, so I can have a lot of flexibility in control.  I just need to know which is best.

(I have no idea what I was thinking when I said leave the screens on.. )

[KA2DZT]

You CAN NOT leave the screen voltage on without plate voltage.  You'll wipe out the screens in a few seconds.  I always think it's best to key the HV supply, keying the primary.  You'll need a step-start relay setup.  You can also key the screen supply at the same time.  Same for the mod supply with triodes.

Fred

[w4bfs]

Bob ... be sure to look in QSO about the long thread about K1JJ's Fabio II saga ... I think this is the most difficult conceptual part of am transmittiing ...73 ...John

[W2VW]

The screen can be left on but the tube must be cut  off. No drive and safety bias. Probably not a good idea  unless a screen current overload is also used.

[W4RFM]

I don't know what I was thinking earlier, I knew that would take of the screens in a flash (literally),but it came out anyway!! I do better thinking at night - I think..  Think I will work on a step start design kinda like my old FM transmitters from Quincy had. Thanks.

[KA2DZT]

I don't know what I was thinking earlier, I knew that would take of the screens in a flash (literally),but it came out anyway!! I do better thinking at night - I think.. Thanks.

Bob,

It's OK, we'll let you slide this time, but don't let that happen again.

Fred

[WD5JKO]

When desiging my transmitter (which will be AM only) should I plan to switch the RF plate and Modulator plate on an off on the transformer primary side with a relay arrangement and leave the screen  B+ on the 4-125 RF final, or is their a better way? (Triode modulators). I have a supply for the modulator B+, separate from the RF B+ and the screen supply for the RF final is also a 3rd separate supply, so I can have a lot of flexibility in control.  I just need to know which is best.

   Using triodes eliminates the screen supply control issue, and more importantly eliminates the need to properly modulate the screen at the proper percentage, and in phase with the plate. If your going to use a 4-125 in the RF final, one option is to get the screen supply via the plate with a series dropping resistor. This method gets increasingly impractical as the amplifier power level rises since the screen dropping resistor dumps a lot of heat. At the power level of 250 watts or so with a 4-125, the screen dropping method is still doable. Then you get into the need for a clamp tube, or other means to limit the screen voltage if the plate  voltage is on when there is no RF excitation.

  There are many ways to do this, and I just brought up one as an alternative to multiple switched supplies.

Jim
WD5JKO

[KA2DZT]

Jim,

Very good point, I use a screen dropping resistor in my 813 rig, works perfect.  I also use a 6Y6 clamp tube that senses the neg grid voltage.  No grid drive, the 6Y6 conducts and pulls down the screen voltage to a safe level, it also works perfect.

Fred

[W4RFM]

These are good thought guys, thanks. I have looked at the clamp tube in a pattern similar to my Apaches operation..  I will visit that and the Handbooks theory section again.

[wa3dsp]

Bob,

 For reference also check out the Johnson transmitter line. They used clamp tubes for screen protection. There are also solid-state devices that would work today if you care to mix technologies.

[WD5JKO]

Bob, Fred, etc...

Look at the Gonset G-76 method of clamp tube operation with a 6DQ5 RF final. They address two things often not done well. The first is the clamp tube when saturated (no RF drive) cannot pull the screen low enough to cut off the tube plate current. The other is that the series dropping resistor method results in an over modulated screen grid. So what Gonset does is to put a resistor at the screen to ground (22K I think), and then insert a VR tube in series with the screen supply. So when the clamp tube kicks in, the VR tube goes out, and the screen voltage goes to zero. This way the plate current of the RF final is just above zero...great for CW key UP. The screen to ground resistor during AM modulation also acts as a voltage divider to in effect lower the screen modulation percentage.

Doing it the Gonset way at higher power will be a juggle of the various variables to get a workable solution. If it can be pulled off, it will be worth the effort. A VR-75 should handle 40ma...is that enough?

The last issue with the screen dropping G2 supply is to keep the screen modulation in phase with the plate modulation, and to do so up to at least 5 Khz. Some R-C compensation is needed...think about how a 10X scope probe is compensated.

Jim
WD5JKO

[KA2DZT]

Bob,

The 6Y6 clamp tube circuit is simple to set up.  The cathode is grounded, the screen can be tied to the plate or maybe through a 100ohm resistor.  The grid is connected to the neg grid voltage line somewhere between the low side of the grid coil tuning circuit and the grid leak bias resistors. You'll probably have a grid RFC at the low end of the grid tuning circuit and then grid bias resistors.  Connect the 6Y6 grid to point of the PA grid resistors and the grid RFC, also use a .001ufd bypass cap at that point and one on the grid of the 6Y6.

The 6Y6 plate gets connected to the screen of the PA tube.  But, I use a number of screen dropping resistors in series.  At the screen pin I have about 1-2Kohm resistor with a .001ufd bypass cap at the screen pin and another .001ufd cap on the other side of the 1-2K resistor.  Connect the clamp tube plate to the 1-2k resistor not at the screen pin.  The remaining part of the screen resistance can be mounted anywhere in the rig.  This part will need to be bigger wattage resistors 50-100watt.  Put them somewhere were they will get more open air for cooling.  Mine are mounted on the side of the PA enclosure, outside of the enclosure.

Fred

[W2VW]

   Using triodes eliminates the screen supply control issue, and more importantly eliminates the need to properly modulate the screen at the proper percentage, .

  There are many ways to do this, and I just brought up one as an alternative to multiple switched supplies.

Jim
WD5JKO

A lot of people will mention extra drive power needed for triodes and simply write this solution off. By the time all the tetrode support circuitry is built it becomes more complex than the larger driver circuit.

Triodes also can be made to neutralize down to almost perfection for the OCD builder.

[KA2DZT]

Dave,

I'm working on (early stages) a single 304TL rig by 813s.  I heard 304s are hard to neutralize, any experience with 304?? I have 50 watts of drive if needed.

Fred

[W2VW]

Dave,

I'm working on (early stages) a single 304TL rig by 813s.  I heard 304s are hard to neutralize, any experience with 304?? I have 50 watts of drive if needed.

Fred

Hi Fred,

     I have no experience with that bottle. I've worked with the 250TH, 3-500Z and a single 833 plate modulated. Not sure why the 304TL would be any worse than an 833.

Also had a 4-1000 plate modulated and built up a bridge neutralized bandswitched circuit using mostly MF marine transceiver tank parts. It worked but not as well as the old BC-610 when it came to max out, plate dip and grid I peak all concurrent.

You have enough transformers for the new rig?

Dave

[Steve - K4HX]

Talk to Chuck, W3FJJ. He built a 304 rig and ran it successfully for many years.

[W4RFM]

Here is the RF diagram of what I intend to build. From the 1956 Orr handbook.

[KA2DZT]

Bob,

Looked at the schematic,  You're using fixed grid bias, so you don't need the clamp tube.   It will not work with the separate screen supply and choke circuit.

The circuit looks good. Although, one thing I see, you have too much screen bypassing.  Two .01ufd caps will bypass all the audio.  Use .001ufd caps. A max total bypassing on the screen shouldn't be more than about .002 ufd.

Fred.

[W8ACR]

Bob,

Triode RF amplifier circuits are simple to build and easy to troubleshoot. The only drawback is the need for relatively higher grid drive power. But at the power level you are planning to use, 10 to 15 watts of drive would be adequate for many tubes (for example an RCA 8000 can be run at 1600VDC and 250mA and requires only 8 watts of grid drive.

I know that you already have plans for the 4-125 circuit, but if you are planning to use a commercially built exciter such as a Johnson Ranger or DX-60, then a triode RF amp would be feasible at the 350 watt input level, and would greatly simplify your construction. Although I am a relative novice at transmitter construction, I have built four different medium power RF amp decks using triodes and all worked well. One of them had a parasitic that was easily eliminated by shortening various leads, all the others worked the first time power was applied.

My most recent project uses a Taylor 822 triode at 400W input level. Plate B+ voltage is continuously ON for both RF final and modulators. I use cutoff bias of about -125VDC for the RF final. Applying grid drive allows grid leak bias to override the cutoff bias (it rises to about -300VDC, that is, it becomes more negative). Also, there is -2.4 volts (battery) grid bias for the 811A modulators which have an idling current of about 40mA. When in standby, the plate voltage is disconnected from the speech amp tubes (you don't want audio getting to the modulators when there is no load on the secondary of the mod transformer). Keying the exciter applies drive to the grid of the 822 and also throws a relay that applies plate voltage to the speech amp, enabling the audio input. I use a pi network final tank with balanced grid tank and grid neutralization. Using a swinging link final tank would probably be even simpler.

Attached is a schematic of one of the RF decks I have built in the past. I wanted you to see the bias circuit for the RF amp. Very simple, and works well. I can get you some large triode tubes if you want to give this a try.

Ron