Analog drive for push-pull FETs

[ka1tdq]

This project keeps morphing, but I'm thinking about another analog driven transmitter.  I'm wondering if my analog drive cores will work, and they can be seen in the right side of the picture.  

RF drive is floating above ground and the primaries will be fed out of phase.  They are six turns each, both wound identical with no overlapping turns.  The secondaries are a single turn and will feed the gates in phase.  Twisted pair wire of good size will go to each gate of the push-pull.

I have the output transformer wound and it's on the left side of the picture.

Jon
KA1TDQ

[WBear2GCR]

what is this project?

Linear?
Class D/E??

Class C "plate" modulated??

[ka1tdq]

Ha!  It's a 40 meter, push-pull, class C AM transmitter.  I did a little more construction work so that it might be a little clearer.

I'm wondering about the gate drive using a rice box as an exciter.  I do it with my current 40 meter transmitter, but that's a single-ended design.  I want to build one push-pull.  I just have to get opposite drive signals for the gates.

Jon

[WBear2GCR]

transformer with a center tap is the usual way to generate out of phase signals without an active circuit.

[ka1tdq]

Yeah, that didn't seem quite right. Here's a picture of what I had in mind all soldered up. I'll change it to just one transformer.

Jon

[ka1tdq]

Ok, I changed to one transformer with 6 turns on the primary and 1 turn each for each secondary, out of phase. 

The output transformer has 1 turn on the secondary and one turn on the primary?  Am I better off going with one or two FETs per side?  I plan on using FQA11N90's since they're cheaper and reliable.  This is also class C.

Jon

[WBear2GCR]

I'm no expert on these ferrite core transformers, but more turns (in ratio) might make for a better transformer.

Do you have a schematic sketch of how you are making the windings out of phase?

                    _-_-

[ka1tdq]

Here's a sketch of the input transformer.

The output core is the same as Steve's from his website, just half the number of cores.

Jon

[VE3ELQ]

Its fun to experiment and try new ideas.  But is also prudent to first define your design objectives and then do some research before getting too far into a project. I dont want to be overly critical here but what you seem to be trying to do is fraught with problems.  
Class C by definition means the FETs will be biased just below the point of conduction and you plan to drive them with an analog sign wave. That means each FET will be operating in the linear part of its curve even if highly overdriven.  Very bad news for switching FETs which is what you are planning to use. That will result in very poor efficiency and certain FET destruction at anything more that very modest power levels.There are lots of articles on-line discussing this, have attached but one. You want to drive those FETs as fast as you possibly can unless you actually want to run then linear.

Also suggest you read up on "Boost Converters".  That's essentially what our digital drive class D and E rigs are, two boost converters operating out of phase only instead of a diode charging a capacitor for DC they are transformer coupled to the load providing AC at the desired freq. The inductors are critical since they store and release 100% of the RF power, the FETS just charge them. If the primary inductance matches the deck impedance and is made resonate with FET and any external capacitance, a near sign wave OP will result. It is very difficult to achieve the correct impedance/turns ratio with binocular core transformers, you are limited to 1 or 4 or 9 or 16 etc ratios. But a tunable OP network can help with that to get the correct match.
If you could better describe your design objectives perhaps you would get more constructive advice.
I wish you all the best with your project, sounds like fun.

73s  Nigel

[ka1tdq]

I guess saying class C isn't correct with what I was planning. I planned on getting enough analog drive to turn the FET on and off, which is at least 24 volts pp. then, I planned on putting 48 volts to the drain to see what I got for RF. then, modulate it.

My plan for this rig is to be a backup for my current one, or a mobile rig.

Jon

[ka1tdq]

...and, now that I think about it, what class is the transmitter that I'm currently using? I always just assumed class C since I just had a 5 element filter on the output, but the FET runs in saturation.  A picture of it is on my QRZ profile.

Jon

** My cell phone spell corrected "FET" to "midget"... sorry for the confusion.

[ka1tdq]

With 10 watts applied through an LDG auto tuner I'm getting about 6 volts pp at the gates, not nearly enough for saturation. My ricebox doesn't have any power setting between 10 watts and 100 watts... It's either all or nothing. The waves are out of phase though.

This experiment may only be academic, but it's still neat.

Jon

[ka1tdq]

Although, on 10 meters I'm getting about 25 volts pp with 10 watts input.  I'm using the MFJ-16010 tuner for the input, and the SWR is 1:1.  I had to put the tuner in reverse though by switching the antenna and transmitter jacks.

Jon

[KD6VXI]

Don't know what kind of rice box you have,  but almost any can have a variable power installed.

10k pot,  9 volt battery and whatever plug the radio needs for alc input.

Search 'alc potentiometer 9 volt'.   Depending on your rig, you may need a positive or negative alc voltage.

Set the radio at full power.   Use the pot to dial the carrier level to whatever you want.  Works best in cw or fm.

This will also help kill am distortion.

--Shane
KD6VXI