Keyboard Shortcuts
Likes
Search
Re: Shielded /Coax connectors for wiring in BitX40smd
|
This business of not connecting the ground braid on one end of a piece of coax came up in regard to tuning clicks from the Raduino. ?Some here felt that since multiple grounds from Bitx40 to Raduino created "ground loops", this could be a reason for the clicks. ?But the consensus here seems to be that the clicks are adequately suppressed by using the correct Si5351 library and with proper power supply filtering on the Raduino. ?I don't recall anybody reporting that they solved their tuning clicks by disconnecting the signal ground associated with the DDS signal. ?But you are welcome to try it, unlike motorcycle oil you don't need to wait 100000 miles to test this particular hypothesis. ? From post 25386: ?There are two causes for the clicks: ?power supply to the raduino, and new firmware. ?The firmware as shipped should be fine with regard to clicks, on mine I have to disconnect the antenna in order to hear the clicks. ?If that's not true for you, I'd suggest a big electrolytic cap (47uF or bigger) across the 12v supply into the Raduino. ? ?Better yet, add a 50 ohm 1/2 Watt resistor between your +12v supply and the 12v input into the Raduino, then add that 47uF cap (or bigger) from the Raduino 12v to Raduino ground. From post 25393: ?I think some have been having success taming the tuning clicks with large caps, but here's a post that says a parallel combination of 0.1uf plus 0.001uf work fine: ??/g/BITX20/message/25081? Seems right, likely injecting RF noise, not audio. With the braid disconnected on one end, the coax still functions as an electrostatic shield of the signal. ?The braid must be connected on both ends if you expect the coax to be a 50 ohm transmission line at RF, you definitely want a transmission line for the antenna connection and perhaps for the DDS cable. ?(The Si5351 is actually a PLL based part, not a DDS. ?But the Bitx40 schematic says DDS so I'll go with that.) There are cases where having a single grounding point or bus bar in the system reduces noise issues, all ground wires go to that one point. ?That way the voltage drop due to heavy currents in some wire (perhaps into the IRF510), doesn't inject noise into other parts of the design (such as the Raduino). ?Also, loops of wire can act as loop antennas and pick up stray RF that might be floating about due to lack of shielding and lack of coax somewhere else (like that antenna port). ?The Bitx40 circuit board has a solid copper ground plane, that's the best defense against either of these cases. ?The wires to a separate assembly like the Raduino are to be looked at closely, but power supply filtering seems to solve that particular case. ?Keeping the wires short helps on ohmic voltage drops, twisting them together or using coax keeps them from picking up stray RF fields. In general, I figure the more grounds the better, ?a grid of short ground wires performs almost as well as a solid copper ground plane. ? But what works best in any particular situation is not always obvious, it can come down to just trying stuff. ?Where I really get worried about grounds is when I have two pieces of equipment plugged into two different circuits from the main household fuse box, a refrigerator kicking on somewhere can cause a spike of several volts between those two chassis grounds if the house safety ground wires are absent or are carrying current for some reason. Jerry, KE7ER On Sat, May 13, 2017 at 10:22 pm, Thomas Noel wrote:
? |