Clifton Laboratories 7236 Clifton Road  Clifton VA 20124 tel: (703) 830 0368 fax: (703) 830 0711

E-mail: Jack.Smith@cliftonlaboratories.com

[01 September 2006 - I've extensively revised this page to reflect the version 2.0 buffer amplifier and to add a discussion on Elecraft K2 BFO leakage. I've deleted older material no longer relevant.]

I've found a lot of interest in the SpectraScan panadapter from Elecraft K2 owners. The K2 has a nominal IF frequency of 4914 KHz, and the SpectraScan works well at that frequency. (My early Z90 design used an internal IF of 9830 KHz, and I built several experimental filters with 9830 KHz crystals. However, you may notice a relationship here -- 9830 KHz is almost exactly double the K2's 4914 KHz IF. Internal birdies generated by this combination made me rethink the SpectraScan's IF frequency and I changed to 8.000 MHz for K2 compatibility.)

Unlike some receivers and transceivers, the K2 does not have a IF output jack. Hence, it's necessary to add an IF connection.  I've designed a general purpose IF buffer amplifier that fits within a fully-loaded K2.

The buffer amplifier can be built in two versions, by choice of components:

• A version optimized for the K2, with a bandpass response, centered around 5 MHz for the K2's 4915 KHz IF; or
• A broadband version that will work with any IF from near DC to 100 MHz and is accordingly not limited to just the K2.

Both versions share a common PCB, approximately 1.375" x 1.25" (35mm x 32mm). I will ship the board with the parts necessary to build either the K2 version or the broadband version.

I also have designed a BNC-connector version of the broadband amplifier, but this seems to be of little interest and it likely not become a finished product. If someone wants one or two, I can make available a version without silk screening or solder mask.

How the buffer amplifier is connected to the K2 depends on whether the K2 has the optional noise blanker board.

Stan Rife, W5EWA, has been very helpful in working through an elegant mounting and connection arrangement for the K2 buffer amplifier board.

The buffer board's input wiring consists of a short length of Teflon coaxial cable, and a single piece of hookup wire for power. These wires terminate in a standard 3-pin 0.1" spaced male header plug, insulated with heat shrink tubing.

The 3-pin plug from the buffer amplifier plugs into the J12 socket at pins 1, 2 and 3.

The 3-pin lead from the buffer board plugs into the newly added 3-pin socket. 

Stan, W5EWA, reported seeing a steady signal displayed on the Z90, even with no antenna connected to his K2, but it was not clear if this signal was an aberration in his particular K2, or if it will be generally seen in all K2 transceivers.

I'm seeing the same signal, and it appears that it is BFO leakage through the reverse gain of the IF system. The BFO leakage level is approximately -80 dBm as seen at  the IF pick-off point.

Whether the BFO leakage signal will be a problem depends on your perception. Here's what I've found with my K2, and Stan's results are similar.

It seems that the BFO signal is passed back to the buffer amplifier's input via one of two possible paths, as illustrated in the marked-up K2 block diagram below.

It's not clear which of the two possible paths is the culprit. Considering that the BFO signal is quite strong (I measured nearly 2 volts peak-to-peak at U11), even if it is attenuated 90 dB, it still shows up as a clearly visible signal at the IF pick-off point.

I've tried alternative buffer board locations, and the BFO leakage signal remains the same, which indicates the problem is unlikely to be stray coupling into the buffer amplifier. Rather, the BFO leakage is directly in the signal chain. (I've had E-mail exchanges with Wayne Burdick at Elecraft confirming this is the likely source of the signal.)

Several possible reasons:

• The K2 is a single-conversion design, and hence the BFO is at the same frequency as the point where we sample the IF frequency. A multiple conversion receiver will not have the BFO on the same frequency as the high IF.
• Relatively simple IF chain. There are only two active stages between the IF pick-off point and the BFO. Thus, the BFO suppression is critically dependent upon (a) U11 (NE602) balance and the reverse gain of U12 (MC1350) IF amplifier. If both U11 and U12 have -40 dB reverse gain (how strong the signal at the amplifier's input is when fed into the output), the BFO will be attenuated only 80 dB at the IF sample pickup point. Other receivers have more IF amplification stages, which improves the overall BFO leakage proportionally.
• Single board construction. At one extreme, commercial and military grade receivers have each major module constructed in a separate, shielded compartment. The K2 has all its RF components on a single PCB. This provides a significant cost benefit, but may contribute to the BFO leakage.
• If you tap off an IF stage operating at the same frequency the BFO, you may well see BFO leakage.

Because it is at the same frequency as the BFO and because it is much weaker than the direct signal. The leakage that shows up on the Z90 panadapter (and on my HP8558B spectrum analyzer) will have no effect upon normal K2 operation.

The most promising method is to move the buffer amplifier connection point from the recommended output of Q22 to Q22's input, i.e., to the mixer output, at the junction of C159, R80 and R81. This will decrease the BFO leakage signal by Q22's reverse gain.

The problems with connecting at the mixer output are:

• Less gain, as Q22 will not be in the circuit. The buffer amplifier gain can be increased to offset this loss to some extent.
• Much less mechanically convenient, as the suggested connection point is not brought out to a convenient jack.
• Since Q22 has a lot of negative feedback, its reverse gain may not be as high as desired, and the degree of BFO leakage suppression correspondingly modest.