19 June 2009

BPSK splatters

I had not seen such a bad signal yet in my digimode life. 30m, BPSK63, 2009-06-18. I did everything on my FT-817, IPO, ATT, volume control but to no avail: I could still see the signal twice and with many distortion products.

I worked him with 1W, so his signal was indeed very strong. I mentioned the bad IMD (-15dB) but nothing changed. I gave RSQ 555 but should have been 591.

A couple of hints:
Make sure the ALC is not active when transmitting. With ALC=0, just below its ALC>0 point you are already transmitting at full power. After that ALC reduces your audio drive, creates non-linearities in the TX chain and sends out splatters. Moreover with ALC>0 your measured power is distributed amongst all unwanted signals, thus you actually reduce your main signal power (see below)!

Transmit above 1500 Hz in the waterfall. This helps not to send the signal twice. If you sit at around 1000 Hz, your TX SSB bandpass filter will let through 2000 and 3000 Hz too, so you'd transmit any harmonic generated within your computer of first TX audio stages. Better stay between 1500 and 2000 waterfall Hz, and retune your RTX main dial. This is a clever hint that I have not seen mentioned anywhere.

"below". A proof that his distorted 50W did no more than my 2.5W comes from pskmap: we both reached the same stations, including one in the US East Coast.

18 June 2009

Console switch

Have you ever come across a keyboard+mouse+VGA mechanical switch? I have found a 4-way model (4 computers, 1 actual device) and couldn't resist to look inside it.

It is routing about 25 lines. The solderwork look done by hand and with just one color of wire! Hats off!

Any use in the shack?

16 June 2009

2 x Balcony QSO

Funny enough to report it here. Last night I was working 30m PSK31 when a Mr. Enzo from Paris came back to my CQs... after the usual data exchange it turned out we were both using an antenna on the balcony. Him 80W, me 5W, 100% copy on both sides. Moreover he is Italian, according to his qrz(dot)com profile.

I think that was my first BPSK31 2xBalcony QSO. HI.

It is almost time to open a club and give awards, HI.

12 June 2009

EFHWA tuner quirks

For the upcoming Summer holidays I decided to try something new: an End Fed Half Wave Antenna (EFHWA). Since all operating conditions will be known (same place of last year), I should be able to compare two different antennas.

I chose the EFHWA because it is reported to have a decent radiation angle even when low above the ground. This antenna has a high impedance between 2500 and 5000 ohm depending on the way it is installed, so it requires a matching device.

Given I will be using QRP, I built a transformer on T50-2 code with 3:27 turns ratio. This 1:81 step-down impedance transformer matches 50 ohm to ca 4000 ohm. With a polyvaricon capacitor it tuned between 6.2 and 22 MHz.

Then I reduced the turns ratio to 3:25, so that it would match a 3300 ohm antenna, in between the EFHWA expected Z ranges. This increased the resonance to 9-26 MHz since the secondary inductance had reduced. Good 30 to 12m coverage, I wouldn't be able to throw a 20m of wire for a 40m EFHWA anyway.

Ready for boxing the tuner... surprise! The tuning range has now shrinked to 11-26 MHz! I have lost 30m when fitting it into a plastic box. This means that the maximum capacitance or inductance available has reduced (I did shorten wires of 2-3 cm if that matters), or both. Need to get back to it and try something different.

For those interested: I found out the tuning range using an antenna analyzer (MFJ259B) and connecting a 3k3 resistor in place of the antenna. Gave full capacitance and looked for the SWR dip with the 259. Same at minimum capacitance.

11 June 2009

CW vs PSK vs FeldHell

You might have read of my recent interest in machine generated modes (or "digital modes"). Lately I have filled my logbook with tens of PSK31/PSK63 contacts and one FeldHell. All on 30m with 2.5W power output into the horizontal base-loaded monopole antenna.

One night I could not get any contact in PSK31, so I got bored and moved to the CW part of 30m, where I collected 3 QSOs. All their sigs were 599 on my meter, including that UK HAM that was running 5W.

So, CW wins over MGMs? I did an Internet search and found references to a work of Tony K2MO (search it yourself). Besides his measurements, he published a minimum Signal-To-Noise ratio for error-free copy, where CW 20WPM scores -13dB, PSK31 -10dB, PSK63 and FeldHell -7dB.

The lower SNR, the more noise is accepted for error-free copy. So, CW@20wpm is more robust than PSK31. Given the 3dB difference, I need to double the power when doing PSK31 to obtain the same CW SNR. One more double for FeldHell.

So: CW 2.5W, PSK31 5W, FeldHell 10W.

In theory of course.