RF Column 29 - March 1994 Copyright (c) 1994,1995 H. Douglas Lung ALL RIGHTS RESERVED TOPICS: Build a simple standard frequency receiver & calibrator List of standard time and frequency stations -------------------------------------------------------------- Last month I mentioned I'd lost the parts for my WWV receiver / frequency standard project. Well, fortunately I found them in Los Angeles before the earthquake rearranged my apartment. I hauled them back to Miami and am pleased to report I've come up with a working design that uses two IC's and one FET. This month I'll show you the details so you can build and perhaps improve on the design. Although I've only tested this circuit with the U.S. standard frequency station WWV operating at 10 MHz., there is no reason why the circuit can't be modified to work on other frequencies. Most time and frequency standard stations operate on 5 or 10 MHz. Table 1 is a list of major world wide time and frequency standard stations operating full time. I have not listed all stations or all frequencies. Shortwave guides usually list these stations. The most complete list I've seen is in the "World Radio TV Handbook" published by Billboard Books. It should be available at most amateur radio stores. Why build this receiver / frequency standard? You may remember this F.C.C. Rule from my column several months ago when I promised the receiver design. Paragraph 73.1540(c) of the F.C.C. Rules states: The primary standard of frequency for radio frequency measurements is the standard frequency maintained by the National Bureau of Standards or the standard signals of Stations WWV, WWVB, WWVH and WWVL of the National Bureau of Standards. While it is possible to use a calibration laboratory with traceable standards, the calibration they provide usually isn't valid for more than six months. Also, with less expensive frequency counters, the long term stability of the time base in the counter may not be sufficient to provide accurate enough readings even over six months. For most F.C.C. required frequency measurements, a timebase accuracy of 0.2 parts per million (PPM) is needed. The over the air frequency standard stations provide the easiest way to check the calibration. I've found I can adjust the oscillator on the receiver to within 0.1 PPM listening to the audio output. Using an oscilloscope, I was able to maintain the frequency within 0.02 PPM, equivalent to 0.2 Hz. at 10 MHz. I describe how to do this next month. First, lets look at the circuit. Figure 1 shows the receiver / standard. The heart of the circuit is the NE602 double balanced mixer / oscillator integrated circuit. The SA602 is an equivalent chip. I used one made by Signetics. The NE602 is designed for use in cellular radios as a second mixer, but does an excellent job at HF frequencies. The specifications show a typical noise figure of 5 dB at 45 MHz. with a third-order intercept point of -15 dBm. The NE602 amplifies the off the air standard frequency signal applied to L1 and mixes it with the signal from its oscillator. The output of the NE602 is the difference between the oscillator's crystal frequency and the standard frequency station. A capacitor across the crystal (X1) permits it to be adjusted to same frequency as the standard (zero-beat) or to a precise offset (more on this next month). The output of the NE602 also includes the sum of the two frequencies - 20 MHz. I filter this out with an RC network (R5, R6, C10, C11) before amplifying the audio with an LM386 amplifier. Capacitor C14 increases the gain of the LM386 to 200, which is enough to drive an efficient speaker. I had some reservations about the circuit while building it. First, I wasn't sure I'd have enough RF or audio amplification. The first circuit I tried didn't. I increased the audio output by using the balanced output of the NE602 and the balanced inputs of the LM386. I had to eliminate the volume control when I did this. If you need a volume control, use a dual 10K audio taper potentiometer after C12 and C13. Connect the wipers to the LM386 - a coupling cap isn't required here. I also found it was necessary to use a tuned input circuit to increase RF coupling efficiency. While it worked without one, performance was horrible. I was also concerned the internal oscillator circuitry in the NE602 would not be stable enough for use as a frequency standard. This wasn't a problem. I was easily able to keep the frequency within 0.02 PPM long enough to calibrate my frequency counter. I did find that connecting the counter to the oscillator slightly changed the frequency. This should only be a problem if you are attempting to use the calibrator without receiving the standard frequency station. Reducing the value of C3 and/or inserting a resistor in series with it should help isolate the counter output. If the coupling is reduced too much an additional amplifier stage may be necessary to get enough voltage to operate the counter. Other than the IC's and the crystal, the parts aren't particularly critical. The quality of the capacitors used in the oscillator circuit will directly affect its stability. Use ceramic capacitors with "NPO" temperature drift characteristics for C3, C4, C5 and C6. I found mine at Radio Shack and Digikey. The coarse (C7) and fine (C8) frequency adjust capacitors need to be high quality variable capacitors. Do not use compression type mica trimmers here. I used a Sprague-Goodman "Filmtrim" capacitor from DigiKey for C7 and an air variable 1-3 pf. capacitor I found at a hamfest for C8. NPO ceramic trimmers should work fine for C7, but for C8 an air variable or piston trimmer is better, since it they have less of a tendency to change capacitance when the tuning tool (or your hand if you use a knob) is removed. Whatever type of variable you use, be sure to ground the side the tuning tool will touch. The 0.1 uf bypass caps aren't critical, however, I'd recommend using a film type capacitor rather than a ceramic capacitor for C11 and C10. All capacitors with polarities shown are electrolytic. I used tantalum caps for C12, C13 and C14 and a plain electrolytic for C15. None are critical, but C12 and C13 should be matched. A film type capacitor would be best here, but big! The inductors L1 and L2 will probably be the most difficult to locate. I wound my own using iron power toroid cores. L1 has 30 turns of #26 wire on a T68-6 core with a 5 turn coupling loop. L2 ended up being 2.6 microhenries - 23 turns of #22 wire on a T50-2 core. I like toroid cores because they are self shielding, however, conventional coils wound on a form like a plastic 35mm film canister ought to work. L1 needs to be 4.2 microhenries. The coupling loop should match the 1:6 turns ratio. None of this is critical as long as the tuned circuit L1 and C1 resonate at the same frequency as the crystal and the standard frequency station. I used another "Filmtrim" variable capacitor to tune L1. Don't be afraid to remove turns or add a parallel capacitor if needed to resonate the circuit. The value of L2 will depend on the crystal used. It may not even be necessary! I used a 10.000 MHz. CTS microprocessor crystal for X1. While not the best choice for long term stability, it was fine for this application. My original circuit did not include L2 and I found I could not get the crystal to oscillate below 10.002 MHz. I added inductance (L2) until I could tune the crystal using C7 and C8 through 10.000 MHz. Unless you can tune the crystal through zero beat it is tough to be sure the frequency is correct. I'd recommend building the circuit without L2. If, after constructing it, you find the oscillator will not go low enough in frequency, add L2. On the other hand, if the frequency is too low, even with C7 and C8 at minimum capacitance, put them in series with the crystal (capacitors go to ground) instead of in parallel. If you intend to use this circuit as a standard without simultaneous reception of a standard frequency station, I recommend a temperature compensated crystal oscillator (TCXO) or a crystal in an oven instead of the internal NE602 oscillator. In this case, remove C4, C5, C7, C8 and X1 and inject at least 200 mV of RF from the oscillator into pin 6 of the NE602 through C6. You should move C3 from pin 7 to pin 6 to couple the oscillator output to the counter amplifier (Q1). Be sure the frequency of the oscillator you choose can be easily adjusted plus or minus a minimum of a few hundred Hertz. My first receiver had problems with feedback. Loud feedback. Apparently, a small amount of RF was getting through the LM386 and back into the NE602. I solved the problem by adding an RF choke ( I used a 100 microhenry choke from Radio Shack) in series with the power supply line to the LM386. A choke wasn't required for the NE602, since it draws very little current (under 3 mA.). A series resistor (R1) working with the 0.1 uf bypass capacitor on pin 8 provided sufficient decoupling. If the bypass caps you use don't have good characteristics at radio frequencies, add a 0.005 to 0.01 uf ceramic capacitor in parallel with them. The power supply for the frequency standard receiver is simple - 4 AA cells. I measured the current draw at 8 mA. when using earphones at audio output. The batteries will last a long time and provide isolation from the power line. This type of receiver is known as "direct conversion" and they have a history of being very sensitive to picking up hum. Since we'll use the power line frequency next month to get extremely accurate frequency readings from the circuit, there can't be any hum on the signal. Direct conversion receivers are also known to pick up nearby AM radio stations. The inductively coupled input and resonant input circuit should eliminate this problem. I haven't seen a problem with "bleed-through" testing it an urban area (Miami). Construction isn't difficult. I built the first version in one evening using the "ugly" copper clad board method. The ugly method made it easy to change components and optimize the design. With a little care, it might not even look too ugly. If you're not familiar with this style, here's how to do it. Take a copper clad printed circuit board blank, about 4 x 6 inches. (10 x 15 cm) in size and position the IC's, capacitors and connectors around the board to keep the lead lengths as short as possible. The circuit board serves as a ground plane. Once the items are positioned, secure the IC's dead bug style (legs up) on the board by using a short piece of buss wire or component lead to tie the ground connections to the board. Next, solder the IC bypass capacitors to the IC leads and ground, keeping the capacitor lead length to a minimum - 1/4 inch - 6 mm or less. Finally, connect the rest of the components using their ground connections or other components to support them. Double check all the connections and the position of the IC's before applying power. The pin connections for the IC's are shown in figure 1 as they appear from the bottom of the IC. If all this construction sounds too difficult, I found an alternative. Ramsey Electronics, Inc. (793 Caning Parkway, Victor, NY 14564 Phone 716-924-4560 or Fax 716-924-4555) sells a kit with the NE602, LM386 and all the other parts needed to make a tunable 10 MHz. receiver. (Batteries not included) I paid $30 (US) for the kit. I haven't had a chance to test it yet, but I don't see any reason why it couldn't be modified for this application. It doesn't include a crystal, so you would have to modify the board to use my crystal circuit. Also, only one of the NE602 outputs and one of LM386 inputs is used, without much filtering. It doesn't appear that it would be too difficult to modify it to match my circuit. You'd lose the volume control, but since the kit includes an RF gain control it shouldn't be missed. Ramsey uses a 9 Volt battery to power their kit. The NE602 can't handle 9 volts, so Ramsey includes a 6.2 volt zener to reduce the voltage to the chip. I'd recommend using 4 AA cells and eliminating the zener. That's all the space for this month. Next month, I'll describe what kind of antenna you'll need for the receiver along with instructions on how to align and operate it. If you have any questions or comments, fax them to me at 305-884-9661 or call me at 305-884-9664 after 6 PM Eastern. The fastest way to reach me is via CompuServe or Internet. My CompuServe ID is 70255,460. On the Internet, use 70255.460@compuserve.com. You can also write me at 2265 Westwood Blvd., Suite 553, Los Angeles, CA 90064. -------------------------------------------------------- Table 1 - Time and Frequency Standard Stations ATA - India (New Delhi) 10 MHz. BPM - China (Xian) 5, 10 MHz. CHU - Canada (Ottawa) 3.330, 7.335, 14.670 MHz. JJY - Japan (Tokyo) 5, 10 MHz. LOL - Argentina 5, 10 MHz. VNG - Australia (NSW) 2.5, 5, 10 MHz. WWV - U.S.A. (Colorado) 2.5, 5, 10, 15, 20 MHz. WWVH - U.S.A. (Hawaii) 2.5, 5, 10, 15 MHz. -------------------------------------------------------- Copyright (c) 1994,1995 H. Douglas Lung ALL RIGHTS RESERVED