4.5W one transistor transmitter for 80m CW

I received invitation to the 10th QRP Minimal Art Session from Lutz DL1RNN. QRPCC -club (http://www.qrpcc.de) is the organizer

of the contest, where minimalistic QRP equipments shall be used. Max component count for a transmitter is 50. If there are less,

bonus is given. The contest is on 80m CW.

This transmitter was built for that happening. It meets the < 5W criteria and has 16 components. 16 components is

34 less than 50, 68% less than maximum. Therefore it should add 68% bonus points on top of normal QSO-points.

Capacitor C1 is not really needed if your antenna accepts power supply DC voltage to be present. This would mean 15 components.

One MOS-FET IRF510 is used as a power oscillator. Input capacitance of the FET is 180pF.

If feedback through XTAL would go directly to gate, XTAL-current would become quite high and burn the XTAL. In order to lower

the current transformer K1 is used. The XTAL now sees only 1/4 of 180pF as a load and the current is 1/4 too. As power dissipation

in the XTAL is I^2 * Rs, the transformer lowers power dissipation to 1/16 compared to situation without it.

Transformer K1 and capacitor C5 act also as a high pass filter, which prevents the transistor to oscillate on lower frequencies,

where its gain is higher.

Keying is by activating the gate bias. +13.6V supply voltage is connected to the key, which shall be isolated from the ground.

Relay or other isolating means shall be used, if keying against ground is preferred. I used level translator with one transistor

and two resistors to fit the TX to my 1volt keyer.

As the oscillator works on quite high power level, it do not allow much frequency tuning. Few 100 Hz would be possible by

adding 100pF capacitor from transformer center-tap to ground. Better way is to change XTAL.

I tried also ceramic resonator but keying got too "chirpy".

Measured output power to 50 ohm resistive load is 4.5W. Keying tone is very clean and clicks are small.

DC input power is 450mA * 13.6V = 6.12W. This leads to 73% efficiency. The explanation for good efficiency is

that the transistor operates close to class E. See Drain voltage and current relations on waveforms page.

The circuit was designed and optimized with help of modeling tool. Real life results were quite close to those of the model.

I made some modeling for other bands too. 160m and 40m are possible with IRF510 with similar results.

On 30m output power decreases even XTAL current has been significantly increased (R3 lowered).

Inductances L1-L3 were wound on Amidon iron powder toroids T-94-2. Fine tuning to specification was done by adjusting

the length of the coil (or pitch). Large size of toroids is not because of power but high inductance values.

The other reason is that I had those in my junk box. Air coils would be even better but need more space.

Very high QRN prevented me to hear in the contest. I got no contacts. The day before I worked Pete DL7UEO with

this transmitter and half sloper vertical antenna.

Pekka OH1TV, OH1WX on QRP

28.5.2009