The interface to the receiver has gone through three different designs:

Magnetic Coupling

The first interface was magnetic coupling with relay coils. Relays were taken apart and all the parts were removed. The coil was mounted on top the VFO toroid and driven by the output of the stabilizer. See Frequency Stabilization using Magnetic Coupling for information on this technique.

If you would like to try this technique, I still have the relays in stock. Let me know when you buy the stabilizer or email your interest and I will send you a relay and the output transistor used to drive the relay. Experimentation is necessary to get this technique to work.

A couple disadvantages of the magnetic technique are a very limited range of stabilization and possible overheating of the transistor driver. The VFO must have a very low rate of drift and the relay current needs to be monitored to avoid overheating the transistor driver. A very large transistor should be used to prevent temperature increases. When a transistor heats up, the collector current changes which raises the stabilizer voltage to the the relay.

A couple advantages of the magnetic coupling are no modifications to the VFO and no additional phase noise. This technique could be used in commercial equipment and kits where mods to the VFO are very difficult. And this coupling will not add phase noise as opposed to using varicaps which have been known to add phase noise to VFO outputs. The phase noise has not been a problem in the ELR and would only be important in very critical applications.

The second interface replaced the relay with MV2104 varicaps mounted on the main tuning capacitor. The MV2104's have worked very well. The construction technique is explained at Stabilization using Varicaps on the Main Tuning Capacitor

Two varicaps and two 100k resistors are mounted on the main tuning capacitor. The mounting is done on the tab side with round holes. The 100K resistors are connected to the cathodes of the varicaps and tied together on the other end to connect to the frequency stabilizer output.

Note in the picture above that the cathode of the varicaps is attached to the stator of the capacitor.

The two end sections of the main tuning cap are tied together for more capacitance. A jumper wire is used to tie the sections together. They are only tied together on this side. The connections to the stator on the other side of the capacitor are used to connect to the VFO.

The output of the stabilizer (labeled relay on the stabilizer board) is connected to the two 100K resistors.

One disadvantage of this technique is that the range of the stabilizer varies according to the placement of the main tuning capacitor. When the main tuning capacitor is at full capacitance, the range of the stabilizer is only 1-4 kHz, while at minimum capacitance is 50-60 kHz.

Another disadvantage were the modifications to the VFO and tuning capacitor. A higher value coupling cap in the 14 MHz VFO was needed and sections of the tuning capacitor were tied together. These modifications were easy, but it changed all the stock settings and ranges.

The effect of the MV2104's would be very little at the maximum capacitance of the capacitor but would be considerable at the minimum capacitance setting. The result is too little stabilization at the maximum setting and too much at the minimum setting.

Be aware that the capacitor value used in this kit is rather large (200pf). In other VFOs where the tuning capacitor is smaller, the varicaps tied to the tuning capacitor may work fine over the entire range.

Varicaps in Series with Trimmer/Adjustment Capacitors

After more experimentation to locate the varicaps for aa even stabilization range, it was found that placing the varicaps in series with a trimmer or a frequency setting capacitor gave an equal range over all the settings of the main tuning capacitor.

This technique is recommended and the construction and pictures are on the Frequency Stabilizer Construction page.

The picture above shows the modification made to the Salmon Pink boards for this circuit. Stabilization is far superior to the varicaps mounted on the main tuning capacitor.

For immediate stabilization of the VFO and a long time between resets, a range value of 10 to 15 kHz for the stabilizer worked very well. In most of the early articles on stabilizers, a range of 4-5 kHz was presented as optimum. But my experimentation has found 10-15 kHz, or even higher, give excellent results.

The VFO coil needs 19 turns instead of the 18 recommended for the receiver. Other than this modification, everything else is left the same. In existing designs with 18 turns on the toroid, the turns can be squeezed together to get 14 MHz within range if the 8pf ceramic trimmer.

The schematics below show the varicaps added to the VFO. The first one shows how the MV2104 is implemented and the second shows how an H66 (a 50pf varicap) is implemented.

Note that C83 (6.8 pf NPO) at the 40/20 tune circuitry has been added in series with the MV2104. The reason is the MV2104 only has about a 10-12pf capacitance. Putting the MV2104 on the 47pf cap would raise the frequency of the 40/20 VFO out of range.

With the 6.8pf and the MV2104 only a minor readjustment of frequency is needed. For retrofitting early kits, there should be an extra 6.8pf or 10pf capacitor after building. Either the 10pf or 6.8pf will work. I would expect that any value from 6.8pf to 15pf would work at C83.

At the 30/17 tune circuitry, the MV2104 is larger than the 8pf trimmer capacitor and does not move the 14 MHz VFO out of range of the trimmer. The VFO coil windings should be raised to 19 turns, but by squeezing the windings together, an 18 turn VFO coil can be made to work.

Here 50pf varicaps are used. No changes or additions in the circuitry is needed other than lifting the 8pf ceramic trimmer and 47pf capacitor from ground and inserting the H66 varicaps.

Then 100K resistors are soldered to the cathodes of the varicaps (the junction between the trimmer/cap and the varicap) and the other ends of the 100K resistors tied together and wired to the output of the stabilizer.

This technique should work well in other circuits. Be sure that the varicap value is larger that the trimmer or capacitor that is lifted from ground. A rule of thumb, to effect the circuitry the least, would be to use a varicap twice the value of the coupling/series capacitor.

The picture above shows the parts on the new PCB boards currently shipping.

The only connection needed when using the stabilizer is to run a hookup wire or small coax between the output of the stabilizer and to the "Stabilizer In" connection shown by the yellow arrow above.

The empty footprints for the 2/8 trimmer and 47pf NPO capacitor are used when the stabilizer is not used with the receiver. It is a simple matter if the stabilizer is not initially built with the receiver to add it later by moving those two capacitors and adding the resistors and varicaps.

The varicap on the right is a MV2104; the one on the left is a 50pf varicap, either a H66 or H65 (supplied in the stabilizer kit). Any similar varicaps will work if you are adapting another type of stabilizer.

The Reset Button is used to set the stabilizer back to the middle of its operating range. The stabilizer should be reset after the first five minutes of turning on the receiver and then about every couple hours afterward.

The input buffer/amplifier (upper left hand corner) is the same MOSFET amplifier that is used in the receiver. This gives a very high impedance input so that the output of the VFO is not affected.

The crystal oscillator (bottom left) runs at 32 MHz. A CA3140 op amp (bottom right) is used as the output device for this stabilizer.

This stabilizer can also be used in different receivers with different frequency VFOs. The only change that will have to be made is at the "Q" connection at the first 74HC4020. For lower frequency VFOs (under 10MHz), use the "Q1" connection at pin 15 of the 4020. For frequencies higher than 14MHz, use the "Q3" connection at pin 2 of the 4020.

This rule is not cast in stone. Experiment with different division outputs of the 4020 until you find the best division pin. See the schematic for which pins to use.

This stabilizer circuit evolved from the stabilizer circuit in the February 1996 issue of QEX, "Frequency Stabilization of L-C Oscillators", by Klaas Spaargaren, PA0KSB. For understanding circuit operation check out this article and the others listed below at Hans Summer's web site.

The only change I made in the output to work with the relay was to change the output resistor on pin 8 of the CA3140 from 10K to 330 ohms. (Reference diagram is from Klaas' article above.)

The other changes were using 4020's instead of the 4060's, and an IC crystal oscillator (32 MHz) replaced the discrete transistor oscillator. Connection changes were made between the first and second 4020's (pin 1 to 11 instead of 6 to 11) to adapt to the kit's higher frequencies.

I also added a very high impedance MOSFET input amp. The output of this amplifier drives the stabilizer. The "Freq Cntr" box is not used. It was intended to drive a Frequency Counter but the output was too high and overloaded the input to the counter.

For further reading and understanding of the circuit, please go to Hans Summer's web site and look up the following articles:

"'Stay Put' The Improved Huff & Puff VFO", by Chas F. Fletcher, G3DXZ, Radio Communication, December 1997 (file name rcdec97as.gif). This article gives all the troubleshooting, formulas, and theory on the operation of his stabilizer. The stabilizer used in the kit is a variation of his design, and all his formulas and troubleshooting apply. A must read if you want to understand this circuit.

Technical Topics, Rad Com, July, 1996 (file name ttjul96as.gif). The circuit in this article is very close to the one used in the kit. The differences are that the kit uses a crystal oscillator can at 32 MHz, rather than a discrete oscillator running at 48MHz, and uses 4020's instead of 4060's. This article also gives excellent information on theory and troubleshooting. The second page of this article is extremely useful at helping to pinpoint a problem with the stabilizer. The circuit diagram contains errors.

"Frequency Stabilization of L-C Oscillators", by Klaas Spaargaren, QEX, February 1996, pp 19-23. This circuit diagram was noted for its lack of errors, and the inventor describes his improved version of the stabilizer. Another complete article that explains circuit theory and troubleshooting. This same circuit is explored in the Technical Topics article above. The kit's stabilizer evolved from this circuit.

"Waveform Conversion, Part I - Sine to Square", Alternate input circuits for the stabilizer. Circuits to convert RF sine wave inputs to square wave for use in the stabilizer.

Be sure to check out Hans' magnetic coupled stabilizer.