Copyright © 2006 Penton Media, Inc., All rights reserved.
Printing of this document is for personal use only.
Reprints

If you've ever designed an RF oscillator, you've probably encountered squegging. Sometimes called "motor boating," squegging causes oscillators to start and stop at frequencies much lower than the frequency of interest. Viewed on an oscilloscope, squegging looks like bursts of oscillations. On a spectrum analyzer, it looks like a Christmas tree. In some designs, such as super-regenerative receivers or wildlife radio tags, this might be a desirable side effect. In most cases, though, it's a nuisance.

Squegging is inherently a nonlinear effect. As such, it's difficult to model mathematically, although it can be effectively simulated in your favorite flavor of Spice. The cause of squegging is usually a shift in the bias of a transistor or other active element responsible for providing positive feedback.

As oscillation starts, nonlinearities cause bias voltages to shift extensively, to the point where oscillation cuts off. The bias voltage then returns to its quiescent value and oscillation starts again. The cycle repeats at a frequency related to time constants within the bias network. Poor circuit layout or inadequate power-supply decoupling can also cause squegging. These causes are readily addressed, but the solution to the shifting bias problem is sometimes elusive. Changing bias points can help, but it's not always clear whether or not squegging is effectively squashed.

One way to avoid squegging is to use a current-mirror bias arrangement rather than the resistors found in traditional designs. As shown in the figure, the circuit is essentially an overtone Colpitts oscillator. Assuming transistors Q1 and Q2 are a matched pair, the current through Q2 is nearly identical to that of Q1, which is essentially the current through R1. As long as the supply voltage is constant, the bias current will not shift, since the base-emitter voltage is essentially constant.

Choke L1 isolates Q1 from Q2 at RF; at dc, it's a short circuit. L2 serves two purposes: It keeps Q2's emitter at ground and, along with C2, determines the crystal overtone frequency. Q2's base presents a negative resistance to the crystal at frequencies above the resonance of C2-L2. Reactance values of 100 Ω or so work well. Larger reactances increase negative resistance and reduce bias current requirements. Choose the resonance of C2-L2 to be between the desired overtone and the next lower crystal resonance. Bias currents on the order of a milliamp or less are sufficient, given the reactance values above.

The only critical components are the matched transistors. Transistor arrays, such as the LM3046, are adequate for VHF. Transistor pairs like the Panasonic XN6537 or XN6543 are usable up to several gigahertz. The crystal can be replaced by an inductor, SAW resonator, strip line, or other resonator. Bias current depends only on the current through R1, so it's easy to control the transistor operating point.

Send us your Ideas for Design. We'll pay you $150 for every Idea for Design that we publish. In addition, this year's top design as selected by our readers will earn an additional $500, with two runners up each receiving $250. You can submit your Ideas For Design via:

• E-mail: dbs@penton.com

Or by

• Postal mail to:
Ideas For Design
Electronic Design
45 Eisenhower Dr., Suite 550
Paramus, NJ 07652
Go to www.electronicdesign.com for our submission guidelines.