A tuned speaker for CW

It’s probably my aging ears, but I’m always on the lookup for a way to reduce background noise and enhance CW signals. Also, I can use any help I can get for accurately zero beating.

I’ve built several resonant speakers using 2” PVC pipe similar to the one described in QST, Feb. 2017[i] and they worked OK, though tuning wasn’t very sharp and band noise was still there. When I purchased a Xiegu G90 transceiver last year my incentive to improve on this increased: the internal speaker in the G90 seems poorly suited to reproducing CW frequencies.

A note by W9ZMV about resonant speakers in the Jan. 1989 QST Hints and Kinks got my attention. Jim points out that short, fat pipes are low-Q, and thin, long pipes are high-Q. A smaller, longer, pipe should have a sharper resonance. This should help reduce background noise, assist in zero beating, and narrow the bandwidth. And so it does.

At the time I wrote this, I had my sidetone set to 700Hz, so my goal was to build my hi-Q resonant speaker for that frequency. I used a speaker unit taken from an old iMac.[ii] The speaker was about 1” in diameter so matched nicely with 1” Schedule 40 PVC that I had in the barn. I started with a longer piece than I thought I would need and tested by sweeping through the frequencies with an audio generator, listening for the peak. The pipe needs to be air-tight against the speaker, and I used rope caulk to seal it. Gradually I cut off short pieces of pipe until it was clearly peaking at 700Hz, with a nice drop-off on each side. Because the directions for the “fat pipe” speaker had sealed the rear of the speaker, I closed the port in the speaker enclosure and inserted some construction-style insulation.

This resulted in a 14” tube plus the speaker unit. It may look weird (the bird on the end of the pipe in the picture isn't functional!), but it works great. Background noise is much reduced compared to other speakers, including my “fat pipe” tuned speaker, and there is a clear peak as I tune to frequency. It is a pleasure to listen to.

I was curious whether I could make this shorter and found a 1” speaker in my junk box – very similar to the $1.95 one at Adafruit.[iii] I glued this to a 1” PVC coupling (it needs to be air tight), then tried various configurations. First I tried putting several 90 degree bends in, but this didn’t seem very promising. Also, I needed more pipe but didn’t want to go to the hardware store during this time of isolation. I decided I was stuck for now.

Then I picked up one piece of PVC about 8 ½” long that I had been using for the above experiment, stuck it in the coupling, and swept through frequencies around 700Hz. To my surprise, there was a very nice peak at 700Hz. The true resonance seemed to be higher – maybe around 1200Hz, but the peak at 700 seemed quite functional, and the overall speaker is much smaller than my earlier one.

The sound from this smaller one seems “brighter” than my larger one, maybe because the rear of the speaker is in open air. Background noise isn’t reduced as much, probably because of the higher resonance peak or peaks, and tuning may not be quite as sharp. However, it is a good compromise for anyone who doesn’t want something looking like a potato gun aimed at them in their shack.

If you want a very cheap and easy solution, here is one inspired by N3HEE’s YouTube video.[iv] Instead of a speaker, this uses a tiny Dayton Audio exciter or transducer (DAEX13CT-4).[v] These are designed to be fastened to a surface, using the surface as a sounding board instead of a speaker cone. I found that a Panera Bread 20oz plastic cup was about right for 700Hz. Other sizes give different resonant points. It is a simple matter of peeling the protective paper off the exciter and sticking it on the bottom of the cup. Tuning isn’t as sharp as the first one above but it reduces background noise and has a nice bright sound. Note, though, that the exciter’s impedance is 4 ohms and may not be compatible with some audio outputs. There’s also the possibility that the cup logo may make you hungry for a sandwich.

My final version – the one I am now using with my Xiegu G90 – uses a 28oz tomato can instead of a cup and has better sound. It seems to have it’s peak between about 600 and 700 Hz and works well with the 600 Hz sidetone I now use. The can sits on four rubber feet to avoid dampening the sound.

Notes: [i] The QST version is similar to the one described here: https://www.qsl.net/n5iw/RESONANT.htm. These plans include a disc with a small center hole over the speaker. That apparently causes the restricted sound that I have experienced with these. I removed that disc recently and the results from these speakers are better. I haven’t used anything like that in the tuned speakers described above. [ii] These are often available on eBay. [iii] https://www.adafruit.com/product/1890. Be careful not to blow these speaker with too much gain as they are rated at .5W. [iv] https://www.youtube.com/watch?v=c985at5sR8o [v] Available from Parts Express https://www.parts-express.com/dayton-audio-daex13ct-4-coin-type-13mm-exciter-3w-4-ohm--295-214for $2.66 plus postage. However, they were backordered so I got mine through Amazon for around $10.00, shipping included.