When I made my first attempts at receiving on the 137kHz band, I used my 7m vertical antenna (fishing rod) on the roof, a tuned circuit and an HP selective voltmeter as a receiver. I was very happy of the performance of the antenna once resonated with an high-Q coil, but disappointed of the receiver, because two very strong signals completely obscured reception: they were two time stations, one at 135.425kHz (-42dBm) and the other at 138.828kHz (-52dBm). Too strong, too near the ham band and too near to each other to be able to avoid them entering the IF of my receiver. By using Argo on the PC, I was only able to see disturbances from those strong stations all over the band.

So came the idea of building a low-IF receiver using a LO at 135.42kHz and an high dynamics mixer. After the mixer, one of the disturbances goes near to DC, the other near 3.4kHz. Both can be filtered with low and high-pass audio filters, leaving the ham band on the middle. Image rejection is not the best you can get, but fortunately there aren't other strong interferers just below 135.42 and the Q of the resonant circuit connected to the antenna is sufficiently high, so background noise from the image frequency is attenuated.

The mixer is based on a CMOS switch, the Maxim MAX4653 in full-bridge configuration, followed from a low-noise differential amplifier based on a Linear Technology LT1028 (see 2nd photo). The differential amplifier includes the first stages of hi- and low-pass filtering. After that we have other 4 op-amp stages (2xNE5532), giving amplification and filtering. On the end we have a 5th-order Butterworth hi-pass at 300Hz and a 5th-order Chebychev low-pass at 2kHz.

I haven't completed measurements, but results appear quite good. Receiver noise is well below background noise from the antenna. Time stations are still audible, but no more disturb the ham band. The only disturbances are a couple carriers from broadcast stations on higher frequencies, which are not sufficiently attenuated by the resonator and pass the mixer in harmonic mode: but they are not very strong, fortunately.

In order to have a very low phase noise, allowing long averages in QRSS, I decided to use an XTAL-based LO. Since I wanted to experiment with PLD devices, I made it with a small Altera PLD driven by a 26MHz commercial VCXO (4th photo). The Altera implements a divider by 192 (actually /96 and then /2 to have a 50% duty cycle) to get 135.4167kHz, quite near to my needs. It also implements a divider by 189, to get 137.57kHz useful to transmit.

And here comes the second part: the transmitter. This is quite simple: an IRFR110 MOS is driven by the Altera clock through an EL7104, and drives the coil that resonates the antenna. This gives a small power of 1-2W, but enough to have more than 1kV peak voltage on the antenna, a safe limit to avoid sparks on the cable and connectors. With my short antenna radiated power is of the order of microwatts... ok, very small hopes of being heard!


Go back to my home page