Tuned Q2 Voltage Probe Antennas
80 Meter Band Version
Some experiments with tuning the input and output of the Q2 stage were undertaken. These designs attenuate out of band RF and intermodulation products from the Q1 and Q2 stages. If mixing occurs within Q1 or Q2 results in products that are within the passband of the filters, they will sail right through the amplifier and into the receiver.
On the 80 meter band, I normally operate/listen at 3.697 to 3.745 MHz. A VPA was designed for a CF of 3.7 MHz to facilitate casual listening in and around this band area. The schematic is shown in Figure 1.
The chosen inductance was 5.1 uH, which has an inductive reactance of 118 ohms at 3.7 MHz. To resonate the tank, a capacitive reactance of 118 ohms would be around 366 pF according to my calculations. As it turned out, the required value was much lower and ended up being around 300 pF. The actual measured total capacitance value for each tank is shown in the schematic. This may vary in your designs.
I found in all of my experiments on this page, that the input tank of Q2 required a little bit more capacitance to resonate the inductor when compared to the output tank. With the fixed capacitor values shown, it was possible to lower the center frequency to around 3.64 MHz by adjusting each of the trimmer caps.
To set the center frequency lower than this, add another 10 pF or so to both of the tanks. Another solution would to substitute a variable trimmer capacitor that went higher in maximal capacitance such as the 12 - 60 pF units.
The performance of this VPA in terms of noise and IMD was superior to any of the VPAs found on the broadband VPA page according to my listening tests. The Q1 source 0.1 uF capacitor in the VPAs on this page may be omitted as explained earlier.
An RF generator was coupled to the Q1 input and a sweep from 3.0 to 4.25 MHz was performed while measuring the output peak to peak voltage in my oscilloscope. The results are shown in Figure 2. Figure 3 shows a graph of decibels versus frequency made using Microsoft Excel.
Advice on doing the math was given to me by W7ZOI. Decibels is 20Log(voltage). The -3 dB bandwidth point can be determined experimentally by using a RF generator and a oscilloscope.
First, find the peak voltage when the tuned circuit is at resonance. Multiply this peak voltage by 0.707 and record that value.
Second, decrease the frequency of your RF generator until the scope shows this value and record this as the Frequency Low value.
Third, increase the RF generator frequency above the peak until you once again get 0.707 of your peak measured voltage. Record this as Frequency High.
Frequency High - Frequency Low = the filters -3 dB bandwidth. In my case, the Frequency High value was ~3.72 MHz and Frequency Low value was 3.62 MHz, so my 3 dB down bandwidth is just under 100 KHz. The -6 dB bandwidth can be found using the same method except that the peak voltage is divided by 2 to get the 6 dB down voltage.
The Q1 and Q2 tanks tune sharply and have a narrow bandwidth. For home built QRP receivers that tune 10-60 KHz of any particular band, they are ideal. If a builder wants to cover the whole 80M band, then filter redesigning is in order. Front panel mounted tuning is one possible solution.
Listening tests confirmed this VPA has no gain outside of the 75/80 meter band, as shown in Figure 3.
In the 2 stage VPAs shown so far, gain is around 10 dB. Some builders may want this to be higher and so the Figure 4 VPA was built. Unfortunately, noise increased in tandem with signal strength, so I wonder if there was actually an increase in sensitivity?
A 3 stage VPA would likely improve sensitivity on the 20 M band or higher, but in the 80-40 meter band versions I built, the extra amp magnified local interference and QRN significantly. Placing a 6 dB, 50 ohm impedance pad between the VPA output and the receiver gave good results.
Perhaps a 3 and 6 dB pad could be built and either one switched in seperately or together to give the builder a range of attenuation while keeping the output impedance at 50 ohms.
Above. The experimental board used for the 3 stage VPA. When I am experimenting to find component values, capacitor and certain other component leads are left long, so the parts can be reused later. This helps reduce the cost of experimentation.
In final projects, leads are kept short. The gate and other leads on the MPF102 should be as short as possible to prevent oscillations. The 15 ohm resistors on the JFET drains are to prevent oscillations at VHF and UHF. Any value from 10-22 ohms is probably okay to use. Ferrite beads do the same thing.
40 Meter Band Version
The 40 meter band version is shown in Figure 5. The chosen center frequency of this VPA was 7.040 MHz, my main QRP operating frequency.
The whip on this VPA is 6 feet long and it has to be kept on a low table in order to fully extend the telescopic antenna. A supplemental page showing photographs of the 40 M version being built was created.
Construction/Testing of the 40M Version
Above. The 40 meter band VPA prior to the LED "power on" indicator being added. It has a 6 foot whip for maximal gain.
Experiments With a Double Tuned Filter on the Output of Q2
I recently re-read Wes Hayward's 1991 QST article entitled "The Double-Tuned Circuit: An Experimenter's Tutorial", and wondered how using a DTC on the Q2 output would play. If you do not have this article, please obtain it.
The article basically states that double tuned circuits (DTC) may exhibit a double-humped response to a wide sweep of RF below, at and above resonance because of the size of the coupling capacitor. The coupling capacitor which connects the 2 resonant circuits together should allow a nice flat response throughout the passband without being so small the filter has excessive insertion loss. The capacitance that fulfills both of these design goals is called the "critical coupling" for the DTC.
As my RF generator will not provide output at 6-8 MHz, I could not sweep the filter. I managed to resonate each tank using a coupling capacitor of 5 pF all the way up to 20 pF. My basic experiment is shown in Figure 6.
I sent the results to W7ZOI and he used computer software to evaluate the critical coupling. Surprisingly, even the 5 pF coupling capacitor exhibited a double-humped response. This is shown (solid red line) in Figure 7.
Some of the variables displayed in Figure 7 and 8 are mathematically derived and are beyond the scope of this topic. Note in Figure 8 how increasing the coupling capacitor to 10 pF worsened the response of the filter even further.
From these 2 plots, I concluded that the critical coupling was below 5 pF. The smallest value capacitor I had in stock was a 3.3 pF unit. When I built a 7 MHz VPA using this DTC, the extra insertion loss was very noticeable and rendered the gain too low for practical purposes. Therefore, using a DTC on the output of Q2 was abandoned.
The whole topic of designing and using DTCs is one I would like to learn more about.
Above. An experimental VPA using a double tuned circuit on the Q2 output. The whip is not yet hooked up. At this stage the -6 dB pad was disconnected and a piece of RG-174 coax is connecting the output to a receiver. The long leaded, coupling cap can easily be seen in the photo. It was a 3.3 pF NP0 ceramic type.
