Crystalline resonators add properties to photonic devices

Vladimir S. Ilchenko, Anatoliy A. Savchenkov, Andrey B. Matsko, David Seidel, and Lute Maleki

Crystalline whispering-gallery-mode resonators are becoming practical optical tools in radio frequency photonics, enabling novel components.

17 February 2010, SPIE Newsroom. DOI: 10.1117/2.1201002.002536

Mainstream broadband applications greater than 1GHz currently dominate radio frequency (RF) photonics, driven by expanding digital communications needs. However, narrowband applications represent important opportunities in analog communications, high-spectral purity RF and microwave signals, photonic front ends, and narrow passband (1–100MHz) RF signal processing. Our work is focused on using the advantages of ultra-high quality factor (Q) crystalline whispering-gallery-mode (WGM) resonators in such applications (see Figure 1), and we have made several recent advances in functional device integration.

Figure 1. A typical crystalline whispering-gallery-resonator on a fabrication mount.

True single-sideband modulators

Crystalline WGM resonators fabricated with electro-optic (EO) materials are particularly useful as narrowband modulators (see Figure 2). The extremely high intrinsic optical Q and small mode volume of WGMs provide for both narrow bandwidth and unprecedented effective interaction lengths of optical and RF fields. Modulators operating with saturation powers as low as −26dBm (corresponding to effective V_π≃18mV) have been demonstrated in X- and Ka-band modulator prototypes at optoelectronic (OE) waves. Resonator-based modulators can also efficiently operate at any desired wavelength within the transparency window of the EO material. These devices play key roles in micro-optic implementation of microwave photonics functions.

Figure 2. Mechanical model of an early optoelectronic wave prototype of an injection-locked, distributed feedback laser-based narrowband photonic receiver. The prototype is equipped with gradient index lens fiber coupling ports for external laser diagnostics and baseband signal retrieval using an external photodetector. WGM: Whispering-gallery-mode.

We recently introduced a new class of WGM-based single sideband (SSB) modulators that results in a high RF return. They also exhibit center frequency tunability over a very wide range (exceeding an octave). As such, they represent a new class of highly efficient devices that simultaneously provide a narrow modulation bandwidth over a very wide RF range. These properties make them easier to use in current applications and enable new applications such as widely tunable RF photonic receivers and oscillators.

Optoelectronic oscillators

Optoelectronic oscillators (OEOs) are used to generate spectrally pure RF signals using photonics.¹ A generic OEO includes a laser, an amplitude EO modulator (EOM), one or multiple optical delay lines, a fast photodiode, a narrowband RF filter, and an RF amplifier. The EOM modulates the continuous wave light emitted by the laser. The modulated signal passes the optical delay line(s) and is transformed into an RF signal with the photodiode. The RF signal is subsequently filtered, amplified, and fed back to the modulator. As a result, a closed active RF circuit is produced that will generate self-sustained oscillation when the RF amplification exceeds the circuit's integral loss. Some circuit elements can be merged or replaced. For instance, a directly modulatable laser and an optical bandpass filter can be used instead of the RF filter, but the oscillation principles remain intact. High-Q crystalline WGM resonators can be used to replace the EOM or for buffering the RF signal and cleaning the OEO's supermode spectrum.² The tunable resonant SSB modulators enable tunable, compact OEOs.³ We have developed various types of ultra-compact OEOs based on high-Q WGM resonators. These oscillators operate in the X-, Ku-, and Ka-bands and are able to generate spectrally pure RF signals characterized with less than −120dBc/Hz phase noise at 100kHz from the carrier (see Figure 3).⁴ The phase noise floor (<−140dBc/Hz) is limited by the signal's shot noise received at the photodiode. We have demonstrated both tunable and fixed frequency oscillators.

Figure 3. Typical phase noise of a tunable opto-electronic oscillator based on a WGM tunable electro-optical modulator.

RF photonic receivers

Photonic RF receivers have large dynamic range and high sensitivity at high frequencies. These compact devices also use relatively little power for operation. Direct processing of high-frequency RF signals with conventional electronic approaches is hindered by the absence of efficient RF amplifiers, detectors, and methods for up and down conversion of the received signals. RF photonic receivers enable conversion of the RF signals to an intermediate frequency (IF) or to baseband, where subsequent efficient detection is feasible. We recently developed a K_a−band photonic receiver based on all-resonant interaction of light and RF radiation in solid-state WGM resonators^{5, 6} (see Figure 4). The core of the receiver, a mixer based on material nonlinearity of the EO WGM resonator, operates well for frequencies ranging from several GHz to 100GHz. The coherent photonic receiver can have a spurious free dynamic range exceeding 55dB and sensitivity of better than −100dBm in the 10MHz reception band. (The sensitivity does not degrade with increasing RF frequency.) It also enables separation of detecting and processing RF signals in space when combined with high-quality optical links used to transmit the up-converted RF signals.

Figure 4. Schematic diagram (top) of coherent photonic receiver and packaged optoelectronic wave, Ka-band photonic receiver prototype on an RF test board. IF: Intermediate frequency. LO: Local oscillator. The entire optical bench (laser, resonator, photodiode, and coupling optics) is incorporated into the surface-mount RF package (bottom) inside the interposer (center, lid removed).

RF photonic notch filters

Notch filters are used to reflect electromagnetic radiation within a selected spectral region while allowing high transmission outside of it. Tunable RF photonic filters benefit phase array and other kinds of radar. Our work led to a novel, highly efficient photonic WGM notch filter that is based on a dual-polarization interferometer with coincident optical paths in the arms. We developed a 10MHz filter with 5.5dB insertion loss and 45.5dB of rejection.⁷ The measured rejection value is limited by the finite (3kHz) line width of our laser.

Our work suggests that RF photonics can strongly benefit from the use of crystalline WGM resonators. Our research has already resulted in several novel WGM-based photonic devices with advanced functions. Future work in the field will focus on improved packaging of the resonators to produce devices that withstand severe environmental conditions. We expect that maturing fabrication, handling, and coupling techniques for resonators, together with adaptation of assembly methods and components from broadband photonics, will result in successful deployment of unique narrowband, high-sensitivity, and tunable functions of WGM resonators in compact manufacturable devices.

Highly Oblate Microspheroid as an Optical Resonator

Large values of resonance quality factor and finesse have been observed.

NASA's Jet Propulsion Laboratory, Pasadena, California

Experiments have shown that a highly oblate microspheroid made of low-dielectric-loss silica glass can function as a high-performance optical resonator. The shape of this resonator (see figure) is intermediate between that of (1) microdisk or microring resonators and (2) microsphere resonators, which have been described in a number of previous NASA Tech Briefs articles. As described below, the oblate spheroidal shape results in large values of both resonance quality factor (Q) and finesse. Large values of these parameters are favorable for single-mode operation of a laser or an optoelectronic oscillator.
A microsphere resonator exploits the circulation of light by total internal reflection, in "whispering-gallery" (WG) modes characterized by large values of Q. In contrast, the Q values of microring and microdisk resonators are limited because of significant scattering losses on their flat surfaces.
The preferred WG modes of a microsphere resonator are those in which light circulates by propagating along the equator. As a consequence of spherical symmetry, a microsphere resonator is characterized by a large spectral density of modes because, along with the equatorial modes, some modes with small propagation-vector components transverse to the desired equatorial circulation are also coupled to an input/output device. A large spectral density of modes is not favorable for single-mode operation.
The highly oblate microspheroid resonator is not subject to the disadvantages of microsphere, microdisk, or microring resonators. In the highly oblate microspheroid resonator, the greater curvature of the surface in the direction transverse to the desired equatorial circulation effectively decouples the partly transverse modes from the input/output device. As a result, the resonator can be operated in a regime similar to that of single-longitudinal mode Fabry-Perot etalons. The free spectral range (FSR) [the difference in frequency between successive modes] is defined by successive integer numbers of wavelengths packed along the equatorial round-trip light path. For a highly oblate spheroid with an equatorial diameter (corresponding to D in the figure) of the order of hundreds of microns and a typical wavelength of 1.55 μm, an FSR as large as 1 THz is expected; in contrast, for a microsphere of approximately equal parameters, the FSR can be expected to be much smaller (typically between 2 and 10 GHz).
At the same time that it affords a much greater FSR, the highly oblate microspheroid resonator retains the high Q (up to about 10⁸) typical of microspheres. This high Q corresponds to a resonance bandwidth of a few megahertz. Consequently, the resonator is characterized by very high finesse (finesse � FSR/resonance bandwidth): typical values of finesse range from 10⁴ to 10⁵. Heretofore, such high values of finesse were available only in relatively large Fabry-Perot resonators.
If resonators like this one were utilized in simple diode-laser frequency-locking schemes, robust single-mode operation should be possible because the FSRs of the WG modes would be compatible with the gain�bandwidth of typical diode lasers. For spectral-analysis applications, resonators like this one offer a highly attractive combination of miniaturization and unprecedented spectral resolution. For optoelectronic oscillators, resonators of this type could provide convenient sideband frequency references in the terahertz range, provided that appropriate detectors and modulators for this frequency range were also developed.

The Highly Oblate Spheroidal portion protruding from the cylindrical portion of this object acts as a high-finesse optical resonator. This object was fabricated by heating a sphere of low-loss silica glass to the softening point and squeezing it between flat cleaved tips of an optical fiber.

This work was done by Vladimir Iltchenko, X. Steve Yao, and Lute Maleki of Caltech for NASA's Jet Propulsion Laboratory. For further information, access the Technical Support Package (TSP) free on-line at www.nasatech.com under the Physical Sciences category.
In accordance with Public Law 96-517, the contractor has elected to retain title to this invention. Inquiries concerning rights for its commercial use should be addressed to
Technology Reporting Office
JPL Mail Stop 249-103 4800 Oak Grove Drive Pasadena, CA 91109 (818) 354-2240 Refer to NPO-20951, volume and number of this NASA Tech Briefs issue, and the page number.

Operation of opto electronic oscillator

Most OEOs utilize the transmission characteristics of a modulator together with a fiber-optic delay line to convert light energy into stable, spectrally pure RF/microwave reference signals. Light from a laser is introduced into an E/O modulator, the output of which is passed through a long optical fiber and detected with a photodetector. The output of the photodetector is amplified and filtered and fed back to the electric port of the modulator. This configuration supports self-sustained oscillations, at a frequency determined by the fiber delay length, the bias setting of the modulator, and the band pass characteristics of the filter. It also provides for both electric and optical outputs. The conditions for self-sustained oscillations include coherent addition of partial waves each way around the loop and a loop gain exceeding losses for the circulating waves in the loop. The first condition implies that all signals that differ in phase by some multiple of 2π from the fundamental signal may be sustained. Thus the oscillation frequency is limited only by the characteristic frequency response of the modulator and the setting of the filter, which eliminates all other sustainable oscillations. The second condition implies that, with adequate light input power, self-sustained oscillations may be obtained without the need for the RF/microwave amplifier in the loop.

Opto-electronic oscillator

An opto-electronic oscillator (OEO) is an optoelectronic circuit that produces repetitive electronic sine wave and/or modulated optical continuous wave signals.
An opto-electronic oscillator is based on converting the continuous light energy from a pump laser to radio frequency (RF) and microwave signals. The OEO is characterized by having very high quality factor (Q) and stability, as well as other functional characteristics that are not readily achieved with electronic oscillators. Its unique behavior results from the use of electro-optical (E/O) and photonic components, which are generally characterized with high efficiency, high speed, and low dispersion in the microwave frequency regime.

Analysis of Wien bridge oscillator

If a voltage source is applied directly to the input of an ideal amplifier with feedback, the input current will be:

Where v_in is the input voltage, v_out is the output voltage, and Z_f is the feedback impedance. If the voltage gain of the amplifier is defined as:

And the input admittance is defined as:

Input admittance can be rewritten as:

For the Wien bridge, Z_f is given by:


If A_v is greater than 1, the input admittance is a negative resistance in parallel with an inductance. The inductance is:

If a capacitor with the same value of C is placed in parallel with the input, the circuit has a natural resonance at:

Substituting and solving for inductance yields:

If A_v is chosen to be 3:
L_in = R²C
Substituting this value yields:

Or:

Similarly, the input resistance at the frequency above is:

For A_v = 3:
R_in = − R
If a resistor is placed in parallel with the amplifier input, it will cancel some of the negative resistance. If the net resistance is negative, amplitude will grow until clipping occurs. Similarly, if the net resistance is positive, oscillation amplitude will decay. If a resistance is added in parallel with exactly the value of R, the net resistance will be infinite and the circuit can sustain stable oscillation at any amplitude allowed by the amplifier.
Notice that increasing the gain makes the net resistance more negative, which increases amplitude. If gain is reduced to exactly 3 when a suitable amplitude is reached, stable, low distortion oscillations will result. Amplitude stabilization circuits typically increase gain until a suitable output amplitude is reached. As long as R, C, and the amplifier are linear, distortion will be minimal.
An alternative approach, with particular reference to frequency stability and selectivity, will be found in Strauss (1970, p. 671) and Hamilton (2003, p. 449).

Amplitude stabilization of wein bridge oscillator

The key to Hewlett's low distortion oscillator is effective amplitude stabilization. The amplitude of electronic oscillators tends to increase until clipping or other gain limitation is reached. This leads to high harmonic distortion, which is often undesirable.
Hewlett used an incandescent bulb as a positive temperature coefficient (PTC) thermistor in the oscillator feedback path to limit the gain. The resistance of light bulbs and similar heating elements increases as their temperature increases. If the oscillation frequency is significantly higher than the thermal time constant of the heating element, the radiated power is proportional to the oscillator power. Since heating elements are close to black body radiators, they follow the Stefan-Boltzmann law. The radiated power is proportional to T⁴, so resistance increases at a greater rate than amplitude. If the gain is inversely proportional to the oscillation amplitude, the oscillator gain stage reaches a steady state and operates as a near ideal class A amplifier, achieving very low distortion at the frequency of interest. At lower frequencies the time period of the oscillator approaches the thermal time constant of the thermistor element and the output distortion starts to rise significantly.
Light bulbs have their disadvantages when used as gain control elements in Wien bridge oscillators, most notably a very high sensitivity to vibration due to the bulb's microphonic nature amplitude modulating the oscillator output, and a limitation in high frequency response due to the inductive nature of the coiled filament. Modern Wien bridge oscillators have used other nonlinear elements, such as diodes, thermistors, field effect transistors, or photocells for amplitude stabilization in place of light bulbs. Distortion as low as 0.0003% (3 ppm) can be achieved with modern components unavailable to Hewlett.^[1]
Wien bridge oscillators that use thermistors also exhibit "amplitude bounce" when the oscillator frequency is changed. This is due to the low damping factor and long time constant of the crude control loop, and disturbances cause the output amplitude to exhibit a decaying sinusoidal response. This can be used as a rough figure of merit, as the greater the amplitude bounce after a disturbance, the lower the output distortion under steady state conditions.

Wien bridge oscillator

A Wien bridge oscillator is a type of electronic oscillator that generates sine waves. It can generate a large range of frequencies. The circuit is based on an electrical network originally developed by Max Wien in 1891. The bridge comprises four resistors and two capacitors. It can also be viewed as a positive feedback system combined with a bandpass filter. Wien did not have a means of developing electronic gain so a workable oscillator could not be realized.
The modern circuit is derived from William Hewlett's 1939 Stanford University master's degree thesis. Hewlett, along with David Packard co-founded Hewlett-Packard. Their first product was the HP200A, a precision sine wave oscillator based on the Wien bridge. The 200A was one of the first instruments to produce such low distortion.
The frequency of oscillation is given by:
 

The Vackar VFO oscillator

Jirí Vackár invented his VFO oscillator during late 40s. It is probably the most stable VFO oscillator known. Vackar configuration is rarely used because of known reason. (The NIH, not-invented-here syndrome).

The frequency tuning range is above 2.5, not observable in any other type of oscillator. The Coupling ratio is fixed; typical range is 1:4 up to 1:9. The tuning is provided independently of coupling. Transistor's parametric variables are isolated from the resonator. The transistor input is not overloaded as Clapp or other circuits. The collector output is at low impedance providing low gain just to maintain the oscillation. The feedback division ratio is fixed. Even if the VFO is tuned, the impedance divider is fixed. The stability is close to XO. Jiri Vackar published his work in a book, providing theory and analysis of each type of his oscillator. What was the last model, V66? Who knows?
A while ago, I thought about subtracting oscillator for PLL VHF synthesizer. Few designs failed high expectations. I used Colpitts, Clapp, Hartley, Pierce, and Seiler. Junk. The Butler is better than single active component oscillator, but it is not good enough. Commonly used oscillator configuration does not guarantee good performance. The signal clipping by diodes guarantees additional phase noise and thermal frequency drift. The best oscillators very commonly use two or three active devices. This is valid for VCOs, TCXOs and OCXOs. The second device acts as an impedance converter, isolation amplifier, AGC circuit, and a phase shifter. The articles in QST very often copy old mistakes, and limiter diodes. U.L.Rohde from Univ. of Washington wrote few articles about the poison of limiter diodes in oscillators. TS-950 use diodes in oscillators..
Ordinary oscillator has poor tuning range, the output voltage swing is unstable, and the frequency stability is poor as well. The industry tries hard to make its sale pitch, to replace single oscillator with 50 ICs, digital dividers, approximation registers, thermostats, and other junk. Now what?
I checked the Vackar. First measurements with frequency counter were quite positive. The Vackar VFO was running in freezer at -30C. That is about -22F. Not bad for Yukon Territory. Stable. Can it run more stable? The Ferrite and iron-powder tuning slugs went out, down the pipes. There is a strong public belief the iron powder cores are good. It's not. In thermal stability they are better than heavy ferrites. Amidon and Micrometals cores use iron powder technology from the 50's, nobody use it any more. They are quite lossy, the grains are not properly bound together. You will have "brown fingers" from handling these cores. That's better than brown nose. Generally, ferrite is based on NiZn or MnZn alloy.
Quite good are Ferroxcube-Philips ferrite cores, presently manufactured in Spain. The Spanish sampling service was good. The reps for US are fine. The Ferroxcube's sales reps in Poland are asking $230 for delivery of $1USD stuff. Sort of robbers the Poles, is that right?
Iskra-Feriti from Slovenia manufactures good ferrite products, toroidal cores and double hole cores, excellent for transformers. I recommend this company. On the other hand, the experience with Epcos ferrites is weird. Epcos is not able to deliver anything in time, screwing around "-we forget". However, this is about VFO.

There are few critical components - good caps of known properties, inductors, voltage regulator, and the transistor. It will run with dual gate MOSFET as well. The LM317 voltage regulator is not the right choice. LM317 is good for car battery chargers. Low noise MAC01 voltage regulator will do the job, even LM78L06 with 50uV of noise is fine. The best regulator of all times is maybe the LM723. It has low noise, high input voltage, and the flexibility. For the oscillator, expect 80dB spur free spectrum. The TFT caps are very good. The mechanical design has to be stable. The coupling with buffer is loose, and at low impedance. The varactor is coil tapped, or cap divider tapped. The fine-tuning with varactor will slightly change the frequency-temperature coefficient. Direct tuning with varactor will ruin temperature characteristics and phase noise of every oscillator. Watch how many designs with single varactor and a 500kHz tuning range ruined all advantages of the oscillator. Varactor behaves as a non-linear resistor with variable capacitor. All parameters change with temperature, DC voltage, and RF voltage. Varactor is good for designs where Kvco = 150MHz/V and above. Direct varactor tuning of low noise oscillator is nonesense. Different attitude is improving the output amplitude variation with AGC. The AGC circuits likely add phase noise. Modeling the circuit phase noise helps a lot. The obvious catch 22 is, in reality the circuit works little bit different way. That's not problem of models or software. You have to cover the whole system. And have fun for the winter weekend. That's the whole story.
I used ceramic coil form with diamagnetic tuning slug for better temperature stability (brass, aluminum). The brass slug works as a single short turn. A single short turn makes no difference. Teflon coil form is good, or Plexiglas will work as well. Keep the Q high, and shield the whole box. Stability of 2Hz at 7MHz was measured. Under 1ppm? Here I stopped. "The VFO can create stable beat with crystal oscillator, and it will stay like that for hours".
The concern was why to use another PLL loop? The oscillator phase noise is lower than any synthesizer use to have. The uController used for synthesizers burn power, radiate heat, and generates broadband spur spectrum. The DDS requires backpack of lead batteries to power the device. The DDS phase noise and residual noise floor is so bad, only morons will use it for receiver application.
The VFO is better solution than the AD9850,AD9891 DDS - direct digital synthesizer chip. The DDS chip has fine-tuning of fractions of Hz. That's right. With X-Tal tolerance of 25ppm. Major performance limitations are clock tree radiation, crystal multiplication (more power and more spurs), DAC aliasing, discrete spurs, discrete dynamic spurs, rich grass type of spectrum, and finite level of residual flat background phase noise, the output frequency is never a round number (5,000.000 kHz). The phase noise parameter is the worst performance issue. If you divide the output signal by ten, the background noise is still there, but lower by 10dB or less. If you want phase noise of -80dBc/Hz @10kHz and worse, use the DDS.
Careful observation of the oscillator performance reveals well-known issue, the limited dynamic range and limited S/N ratio of "improved" HP Spectrum Analyzers. Use rather Rohde & Schwarz, or your own spectrum analyzer. The oscillator circuit works well with J-FETs (square law). Components with cubic transfer characteristic (dual gate, tunnel diode) have excellent frequency stability and performance.

Links related (UK):

Click to zoom

The genuine Vackar oscillator circuit by G3PDM.
With C1/(C4+C6) and C3/C2 = 6. Use a high-quality variable capacitor with ball bearings, two-wheel transmission. Adjust feedback control C2, an air-dielectric trimmer, so the circuit just oscillates. Use a strong box from solid metal. C1, C3 and C6 are silver-mica or ceramic types glued to a solid support to reduce sensitivity to mechanical shock. The buffer amplifier is essential. Circuits using external gate-to-ground diode suffer from high phase noise and instability. The diode loads the circuit. The signal is rectified by the diode, and dynamically shifts the operating point of the J-FET. Single-point grounding is important. The inductor used a ceramic form. Use thick solid wires (#16 to #18 gauges) for mechanical stability. The coil is always mechanically fixed by paint. My choice is transparent nail polish with lacquer thinner. Mechanical 20:1 reduction gearing with anti-backlash may do the trick. Take out the Zener and replace it with voltage regulator. Clean all components and the box in ultrasound cleaner.
Set-up of the bias and feedback caps ratio changes the close-in phase noise. A wonderful nice sine wave output doesn't mean it is low phase noise oscillator. Calculate an T or PI type, five element low pass filter with 1dB ripple. Place is on the output. Cebyshev or Eliptic structure is fine. Don't forget the termination resistances. Think about the frequency plan and frequency dependency, and different types of pulling. The buffers are essential.

What is the TEMP COMP? The ceramic capacitors are manufactured with different temperature coefficients of capacitance. It means, choosing the right combination of capacitors will get you zero thermal frequency shift. The cap combination you have to find out the hard way, by thermal measurement. Standart TEMPCO values for capacitors are Kc = +44(white), +33, -47, -125, -470, -750(purple), -1250 and whatever more. The resulting cap value is C = Co *(1 + deltaTemp*K*1e-6). The units are [pF, C or Kelvin]. The tuning capacitors have mostly very negative Kc. Coils have "positive" Kc influence on the system. Decent present SMD manufacturers of capacitors make caps with different Kc. (AVX, Murata,..) Do you want to spend time calling around and dealing with ten weeks lead time? An old cap from an old TV tuner can help to solve your problem. The treasures are around you.

Look at the microwave systems or WDCT VCOs, Bluetooth, etc., how the frequency planning is done. Atmel got some chips right. Systems with indirect frequency generation have excellent inherent stability. How you can get 0.1ppm stability? It means, you will not amplify 10MHz vco, and pass it to 1kW amplifier. You will not use directly 10MHz.. Do you think it is correct when the VCO output buffer drives directly the mixer? no. The input signal modulates the VCO via the mixer.. what is reciprocal mixing? Some folks asked me how you make low noise VCOs. I told them, you never know in advance.

Good luck! va3diw


DDS ? That's a big question. Welcome to the aliasing world and phase noise background.

The Vackar VFO Circuit.

The Vackar variable-frequency oscillator appears to have some advantages over the usual Clapp (1) circuit. In the latter, the output amplitude varies greatly with frequency. In the Vackar circuit, the output varies only a little with frequency. The useful frequency range of the Clapp circuit is about 1.2 to 1; in the Vackar it is about 2.5 to 1. The first of these advantages should be of interest to amateurs. My friend and colleague, Mr. James B. Ricks, W9TO, has pointed out that the 6AG7 is not the best tube to use for a series-tuned v.f.o.; indeed the several papers originally describing these circuits invariably show triodes. The best tube is that one which has the lowest ratio of change of input capacitance to its mutual conductance. The operating mutual conductance for the cathode, control grid, and screen grid of a 6AG7 (as typically used as an oscillator) is low, despite its high value for the normal grid-to-plate circuitry. Also, it has a high input capacitance and high heater and plate power inputs. In consequence, this tube is not ideal for the purpose. A small dual triode, the 12AT7, offers higher oscillator gm in one triode section, lower input capacitance, and about one third the heater and plate power inputs required by the 6AG7. In consequence, it is a superior tube for series-tuned oscillators. The output voltage will be lower for the 12AT7, naturally, but a tube should not be evaluated for v.f.o. use on the basis of power output. W9TO has adapted the Vackar circuit to an amateur v.f.o. with output on 80 meters using the 12AT7 in the circuit of figure below. The first triode unit and its associated components form the oscillator proper; the other triode unit is a cathode follower which reduces loading effects on the oscillator frequency Two of these v.f.o. units have been made and tested; their frequency stability is excellent, and they key well. The output r.f. was measured as 1.2 volts r.m.s. using a General Radio v.t.v.m. The total current from the 255-volt regulated B supply was 16 ma., key down.

In series-tuned oscillators of the Clapp or Vackar type the characteristics of the series capacitor Cx are critical if the oscillator is to be keyed. An annoying chirp, slight but detectable, was finally traced to imperfection of this capacitor, even though it was a low temperature coefficient silvered mica one. Several silvered micas of good make were tried; they all produced slight chirp, some less than others. A so-called zero temperature coefficient (NPO) ceramic capacitor gave less chirp (very little, in fact), but the chirp was eliminated by using an APC air trimmer for Cx. Apparently, there is enough r.f. current through Cx to cause dielectric heating and a small resulting change in capacity even in these high-grade capacitors. This, was confirmed indirectly by using for C1 a negative temperature coefficient (N750) ceramic capacitor. The chirp was tremendous! Of course, the series capacitor is not the only possible cause of chirp; poor plate voltage regulation or a long time constant in the keying circuit might also contribute. To avoid this, the plate supply should be regulated, and series resistances and shunt capacitances in the keying circuit should be kept to a minimum.(2) The circuit shown will key cleanly without chirp; with the constants shown it will be somewhat clicky, due to turning on and off rapidly; this makes it very desirable for use in a differential keying system in which the oscillator is turned on before the amplifier, and the amplifier is turned off before the oscillator. W9JK ___________________________________ (1) Clapp, J. K., "Frequency. Stable LC Oscillators," Proc. of the LR.E., Aug., 1954, Vol. 42, No. 8, page 1295. (2) The chirp discussed in the preceding paragraph evidently is a slow one attributable to temperature effects. A chirp of the "dynamic" type often manifests itself as a click when the time constant of the keying circuit is very short, becoming observable as a chirp when key-thump elimination methods are used. Ed. This material originally appeared in QST for November, 1955. Ed

The Vackar series-tuned v.f.o. circuit at W9TO. The tube is a 12AT7 dual triode. R.f. output from the cathode-follower second section is 1 .2 volts r.m.s. C2 Silver mica.
C4, C5 Mica.
Cx APC air variable.
Other capacitors are ceramic.