by Andy Nemec, KB9ALN
Happy April Fools! This year we seemed fall victim to an early prank of sorts in the form of a heavy March snowstorm. It seems that my words were somewhat prophetic in this column last month. For that, I apologize. But do hope you can "keep on smilin'" through it all.
I received a nice note via packet from Rich, AA9L near Milwaukee. One of the inquiries he made of me concerned the WAPR web page. He wondered when it was due to be updated. Yes, by the time you read this, you should be able to see the updated Web Page. Point your browser to http://www.netnet.net/~ke9lz/ and look for the index. It should be fully updated with all of the "Using the Wisconsin Network" parts in it, along with links to the new TAPR digital frequency listings.
Rich also expressed interest in my mention of data decode error prevention and the results of my experiments. I will tell you the basics - that the concept of TNC/Radio isolation does indeed seem to make a measurable performance improvement. It seems that getting rid of ground loops is of benefit, and I have seen perhaps a 50% improvement in my situation (maybe even better than that). Unfortunately, a blow-by-blow account of what was done and how would take more space than I can take here. But I will fill you in on the basics, should you decide to try this out yourself.
The isolation network includes a 1:1 audio coupling transformer that is available from Radio shack. If you have an old junk telephone (dead cordless phones are great for this), you can rob the coupling transformer out of it. The important thing is to have a transformer with a 500 ohm primary and secondary windings. This couples the microphone audio from the TNC to the radio, while separating the grounds.
The second transformer required for this is one with one 8-ohm and one 500 to 1000 ohm winding. This separates the radio audio path from the TNC ground. Radio Shack also sells a transformer for this purpose, but it is extremely low-power. It is simply mandatory to have a resistive voltage divider between the radio and the 8 ohm winding of the transformer. This will keep the transformer from getting the "Unintentional Power Conditioning" that will surely ruin it. I do not have the resistor values calculated yet as I am using a different, higher-power transformer. That transformer uses a different attenuator than what is required for the Radio Shack transformer. If you are handy with Ohm's law and the power law, this should be no problem to calculate, though. In addition, the final plans call for a pair of limiting diodes placed across the transformer primary to prevent destructive overdrive.
Next, isolate the PTT line with a simple reed relay, also available at Radio Shack (no, I don't work for them!). The relay coil is powered from the TNC power source, and is keyed by the TNC's PTT terminal. A diode is placed across the relay coil (cathode to the supply voltage) and the contacts go to the radio PTT wiring.
Naturally, the ground path to the radio must not come into contact with the ground path to the TNC inside of the interface (that is the whole reason for the interface). If you house the isolation network in a metal box, do not connect either ground system (the TNC or the Radio) to the metal case. Instead, connect the box to your station ground bus. Some other pointers:
- Use a shielded serial cable. Do not remove grounds from this cable in an effort to reduce ground loops. It will generally make little or no difference, and removing one or more of these grounds may well make a great deal of computer-generated RFI that you don' t need.
- If at all possible, dedicate a power supply to the radio used for packet, don't share it with another radio, and certainly not the TNC.
- Use a master station ground bus, and make certain that all connections to it are truly good RF grounds (no skinny wires here).
Once I constructed the isolation network, I found out two things. First, the receive decode sensitivity seems better. So the volume level on the radio will have to be readjusted. Second, the transformer has a certain amount of loss involved, so the AFSK level coming out of the TNC will have to be increased to compensate for this.
I have noticed a great drop in the number of mangled packets, and have found it to work rather well. If you do plan on building it, please let me know what your results are (good and bad).
And that is about all I have space for. It is my hope to make this a real step-by-step project. If it is published, I will certainly let readers know where to find it, and when.
Keep looking for the announcement concerning the WAPR Spring meeting on your local BBS. We will be meeting soon, and you are invited!
That is all I have for this month. As always, I invite your comments, suggestions, and input. My BBS address is: KB9ALN@KB9ALN.#GRB.Wl.USA.NOAM
Until next time, 73 from Andy.
Back to the WAPR Home Page and Index