28 November 2012

Keyer "in crosta di sale"

CQWW CW contest is one of those events that can be tackled with a simple 2-memory keyer: callign in one, "CFM 599 TU" in the other. It also overcomes my "safe" sending speed limit of 20 WPM: in QRP the slower the better and send it right the first time!
After a couple of contacts I dug my trust keyer based on PIF16F84, opened it to insert the second AA cell required and I discovered that one battery had leaked on the circuit board covering the IC with salt.

Ouch! What I loved of this keyer was that it lasted forever on a couple of AA batteries. Unfortunately I am not able to program a 16F84 anymore, so I will have to switch to something else. Fortunately there is a fairly complete (and complex!) keyer based on Arduino/AVR boards by K3NG, so I am probably going for it.

27 November 2012

Troubleshooting the biquad

First thing I checked on the biquad was the continuity of the coax+antenna setup. Electrically the quad is a short circuit, so measuring resistance across the antenna plug in the shack should be as close to 0 ohm as possible.
Instead my DVM was reading ~10 ohm "round-trip", but measuring at the bare coax cable end returned a fraction of ohm...
The antenna plug has a screw for ensuring electrical contact, which I promptly secured with a solder joint. Electrical resistance dropped to a more meaningful value and incoming signal from the local beacon increased. The biquad is still beaming away from it, I need to move the antenna to ther side of the house so that it will be facing the beacon, but currently my only properly terminated coax cable is running through a wall :-)

Next weekend there will be a 23cm contest, so I will have other on-the-air signals coming through.

21 November 2012

Nothing beats the good ol' dipole(?)

Last night I monitored 1296 MHz with the RTLSDR receiver and the homebuilt biquad antenna, fixed in one direction.

Apart from a strong station, I could hear nobody else. Then the usual HB9 placed a CQ and he was weaker than expected, so I switched back to the indoor 23cm dipole and his signal was at least 5 dB higher: something is wrong with the biquad! Last night's test confirmed my suspicions.

The outdoor antenna should gain about 10 dB over the dipole. The cumulative coax and connectors loss could be 3 dB, leaving a 7 dB boost in favor of the biquad. But according to my late night observations the difference is 7+5 = 12 dB in favor of the (indoor) dipole!

Either the biquad I have built has a weird radiation pattern or something is really wrong. Troubleshooting begins.

16 November 2012

23cm biquad simulation

Before assembling my 23 cm biquad antenna I wanted to know what happens when loop-to-reflector spacing is reduced. In theory, at least.

The standard spacing is 1/8th of lambda, while I want to halve it to 1.5 cm which is about 1/16th lambda.
Why? Because it simplifies building the antenna since the double loop can then be soldered directly to the back of a BNC female head. :-)

3D model of biquad and wire mesh reflector
I use MMANA to simulate antennas. So I drew a biquad with a reflector spaced 1/8th lambda and simulated it (the .maa file). The result was encouraging; both Z and gain were within expected values.

Each simulation run takes 36 seconds in "free space" and 48 seconds if the antenna is said to be on "real ground", so I chose not to increase reflector density.

After saving to file the "far field" data, I reduced spacing to 15 mm (0.015 m) and re-run the simulation. In both cases the antenna was set to be over real ground.

Comparison of 1/8th vs 1/16th lambda in free space.
The comparison of far fields for both antennas (1/8th is in black, 1/16th in green) show a reduced gain and F/B ratio for the shorter antenna. Gain difference is 0.5dB (theoretical), which is negligible for my final, RX-only, application. F/B looses 3dB and doesn't worry me too much, even if I will mainly use the front lobe beaming out of the balcony... I am going to build the biquad right off a BNC female head screwed to the reflector panel (a copper clad board about 20x30 cm).

Next, since I had some spare CPU cycles, I ran an MMANA overview over +/-40 MHz from the center frequency (1295 MHz). Z/Gain/F-B were computed in 5 points, SWR is interpolated. Apart from an impedance discontinuity at 1315 MHz, all other relatively flat values give hope for a normally performing antenna ... especially taking into account all my mechanical bulding errors.

Gain and other parameters comparison at various frequencies.

SWR interpolation over 80 MHz span.

Someone may notice in screenshots that SWR is computed for 75 ohm and not 50. The fact is that the RTLSDR dongle is meant for TV reception, whose impedance is 75 ohm. Also I will use a SAT-TV coax to reduce losses, so why not reason in 75 ohm terms?

14 November 2012

23cm biquad spacing

As 1296 MHz receive only antenna I have chosen to try a biquad (or dual-quad). It is composed of two full-size quadriangular loops (23cm perimeter each), set electrically in parallel and "touching" at one corner, the feedpoint.

Mechanically speaking, a 23cm biquad is simple to build, requiring a thick copper wire (1mm diameter or so) and a large reflector, say, 20x30cm, like a copper clad board. Loops-to-reflector distance should be 1/8th wavelength, which is about 2.9cm.

2.9cm ... that's the tricky part for someone like me that is not equipped to do mechanical work. I prefer what can be done with the soldering iron. So I wanted to understand what would happen to my biquad if the element separation would be of ~1.5cm, that is the height offered by a BNC female head screwed on the PCB/reflector element.

Online resources suggest that a change in spacing modifies forward gain and F/B ratio, but impedance doesn't vary too much. Anyway I have no means to measure SWR at 1200 MHz, so all I am interested in is a confirmation of the gain, amount and direction.

An MMANA simulation is on the way...

12 November 2012

10m and 30m are not harmonically related

Since a week I have switched my antenna from 10m to 30m, to be able to use some after-sunset propagation. On 30m it is composed of a 3.8m long radiator with a loading coil at the base, held horizontal out of the balcony with a telescopic fishing rod. Previous posts: [1] [2] [3] (all on this blog, open in a new window)

Then I remembered that the two HAM bands 10/30m are not harmonically related, which is a prerequisite for building a multiband "parallel" dipole. In my case it would be a parallel dualband monopole. Or three-band, since the 10m radiator could be used on 12m with acceptable SWR.

Since this is an experiment I have never tried before, I will check it out soon.

05 November 2012

RTLSDR as a (cheap) lab spectrum analyzer

While listening to a local repeater with RTLSDR (sdr# software) I spotted a strange oscillation in the output carrier.

I didn't measure the frequency oscillation width, but it was clearly visible at every silence. I suspected a sampling artifact, maybe in conjunction with an output subtone, because I had never seen the same effect on other NBFM carriers. As a cross-check I tuned another UHF carrier, which looked clean:

This unmodulated carrier, receiver over the air (see the signal disappear in the waterfall as I moved around the antenna), was clear and stable, which excluded an instability within the RTLSDR hardware.

Few days later I met the repeater technician (on the repeater itself) and he confirmed the PLL instability in the Motorola transmitter. So, it is not a sampling artifact but I was really looking at an unstable signal.

This oscillation is about 1 Hz, which is hard to spot on the equipment I have access to. But with the RTLSDR + waterfall software it was easy to see. Not bad for a 20 USD device that covers between 60 and ~1400 MHz. Good to know for my 70 MHz building activities.