2009-06-20

Summertime RF Propagation and the CMMR-6P-60

Quite a few people have problems with getting the CMMR-6P-60 WWVB Receiver to work. Usually the breakdown in communications is atmospheric or manmade noise.

The 60 kHz signal we're attempting to pull out of the air is an amplitude shift keyed signal (really slow pwm). Our receiver depends on being able to hear a 17db drop in the power to receive a 1 bps binary encoded decimal data stream.

The slow data transmission rate of our received signal allows us to visually monitor the signal quality quite easily. An LED can be driven by the TCON pin on the CMMR-6P board. I've also included an TCO RX indicator LED driven by Arduino Pin 4.

Using either of these LEDs, you should see a once per second flash with the ON/OFF duration depending on whether a zero, one or mark symbol is being received.

If your TCO indicator LED is sitting there flickering like LED 13 does while you're uploading your Arduino sketches, you have a local man-made RFI or weak signal problem. Something is raising the noise threshold up to the point that the receiver is unable to hear WWVB properly or else you are in a location where the signal is being attenuated (metal building, thick concrete walls, basement). The antenna needs to be at least 3-6 feet (even more is better) from noise sources like computer monitors, displays, unshielded microprocessor circuits, fluorescent lights, etc. 60 kHz theoretically should penetrate most walls, but placing it near a window can't hurt. NIST Special Publication 960-14, Page 42-43 has a listing of things to consider that will cause interference or reduced signal strength.

Another thing to consider with placement is that the loopstick antenna used on this receiver is directional. Signal nulls occur off the ends of the rod, maximum signal is received with the antenna positioned broadside to Fort Collins, Colorado. Try to get the noise source in the null in addition to distance from your noise source.



The other receive problem manifests itself with the TCO LED intermittently giving two or more flickers per second instead of the one per second you are expecting (pulse, pulse, pulse, flicker, pulse, pulse, etc). This can be man-made impulse noise such as having a neighbor drive past with a bad ignition harness, motor starts, etc. More likely is atmospheric impulse noise (distant lightning, daytime solar radiation). During the night, the 60 kHz longwave signal propagates further so the signal tends to be strongest then. In summertime, not only do the number of daylight hours increase causing reduced signal propagation (Page 12, table 2), but atmospheric noise increases as well.

We have one trick up our sleeve to deal with impulse noise. We're using a loop antenna which acts as a magnetic sensor probe to pick up our WWVB 60 kHz signal. Our impulse noise tends to propagate as a mostly electric component. This means that if we construct a partial electrostatic shield, we can block a portion of this noise which can allow our receiver to get a few more decodes than would be possible during the noisy part of the day.

Here's our prototype CMMR-6P-60 WWVB receiver with 100mm antenna.


It needs to be spaced at least 3/4 inch away from the shield.


The electrostatic shield is a sheet of aluminum foil curved into a u-shaped trough.


The shield in place. Notice the ground wire that bonds it to our receiver ground.

No comments:

Post a Comment