Here's some measurments of DC voltage, measured with an average-reading DVM.  Note that the Pixie in question is running off a somewhat tired 9V battery (sags to 7.9V on transmit)  A 50 ohm dummy load is connected to the antenna terminal.  AC measurements were taken with a digital scope, using a 10:1, 1 Megohm 9pF probe.  Click on the AC measurement to see the scope trace.

Note that the DC voltages on the bases and emitters of the transistors seem a little odd.  This is due to the fact that the transistors are operating with large signals on them, and these signals look somewhat un-symmetrical. Normally one expects to see the base at 0.7V higher than the emitter for a transistor that is biased on. However, in the case of Q1, the distortions in the large waveform on base and emitter cause the average value of the voltage on each node to appear about the same.

  One additional note about the pixie I have now: I finished this one after writing the article, and made another change in the design in the interest of a little more power: I used a 2N2222A for Q1 instead of a 2N3904.  It seemed to have a little more gain at the higher collector current we're running at here (versus the original 40/80m Pixie)

  Another change from the article:  You might want to add a 0.1 uF to ground from the collector of Q1.  Especially if your battery leads are very long.  This could help prevent any instability problems
 

Low Power Output (*NEW* 3/24/00)

You can see that with less than about 4 dBm of drive, the stage has no gain at all!  You really want 7 to 8 dBm from the oscillator stage to see full gain.  At that point you'll probably get about 100 mW out.  To get the 300mW or so the rig is capable of, you need about 10dBm of drive, which looks like about 1.6Vpp at the emitter of Q1, due to the non-linear nature of the load presented by Q2's base.

  So, if the output from Q1 is too low, what could be the problem?  Here are some things to check:
 

• Is Q1 biased correctly?  Check the values of R1 and R2.
• Is the beta of Q1 too low?  Try substituting a different 2N3904, or a 2N2222, to find one with more gain.
• Is the VCXO, or VXO circuit limiting the oscillator's output?  Try shorting the junction of X1 and L4 to ground, thus bypassing the VXO. If this makes a dramatic difference in the transmitter output, you're probably running right around 4 mW of oscillator output.
• Is your crystal's Q too low?  If the ESR (Equivalent Series Resistance) of the crystal is too high, this can limit the oscillator's output.  You are using a fundamental mode crystal, right?  An overtone crystal won't work very well. Also, really small crystals (the "low profile" HC-49/US types and similar) have more ESR in general. Full sized HC-49U types seem to be ok. (About 1/2 inch tall)