Introduction
While looking through a 1992 edition of "Hints and Kinks" (published by the ARRL) I found the simple code practice oscillator set described by Art Zavarella, W1KK. on page 9-13. It has a very low parts count and looked like a good circuit to use for a simple MCW set-up. I built it as Art shows in the book and found it to be surprisingly loud for it's simplicity and low power supply voltage. To make it suitable for on the air use, I added a few additional components to produce a controlled keying characteristic and a more sinusoidal output waveform. The results are acceptable for MCW use.

Click here for a parts list for this circuit with Radio Shack part numbers

Circuit notes

1. The complete circuit as modified is shown in Figure 101.
2. The value of the capacitor across the primary of the output/feedback transformer has much to do with determining the operating frequency. That frequency would normally be between 500 and 1000 cps.
3. Almost any transistor can be made to work. If a PNP, rather than an NPN, is available it will work, too. Just remember to reverse the power supply connections.
4. The 50K feedback/bias pot is adjusted for best waveform, not maximum loudness. It's best to look at the keying on a scope, for final trimming of that value, but a good ear might do well enough.
5. Adding the 10 ohm resistor in series with the speaker produces a voltage drop that can be used to provide a signal for MCW. It's addition also made the oscillator more predictable in it's operating frequency and feedback requirements when different speakers were tried. Power consumption dropped, as well. As shown, the circuit works well for any speaker load. If more sidetone volume is required, try a lower value. Connecting the speaker directly to the secondary winding results in the loudest sidetone, but the feedback adjustment point will hold for only that particular speaker assembly. This connection could be used if the set were to be used only for code practice, as there will be no MCW tap, of course.
6. When using this system for MCW generation, be careful about setting the audio level into the transmitter. Too much level can result in a terrible sounding signal and may create adjacent channel interference. Not enough level will result in a signal that is hard to copy.
7. R5 is selected according to the requirements of the transmitter. Some radios require a low impedance load at the mic jack as a signal to go into the transmit mode. A value of 3900 ohms works with my HTX-202. It can be much higher in value for some radios, and may be left out for others. The switch would connect to the PTT line in those cases. C2 is not needed for my radio, and R5 is connected directly to the wiper of R4.
8. I soldered the D cell to the wires. At 5ma key down, it will last for it's shelf life of about three or four years.
9. Due to the low keyed voltage, this setup probably will not work with electronic keyers that don't have an actual relay closure in the keying line. Using a 9V. battery and a voltage divider ahead of R2 might be a solution.