Posts by PA3FYM

SV1BDS Old technology (IC-490!) combined with new technology (e.g. ADF4351).

I have a FT-780, could be an idea to build a super simple transverter around it ; -)

True. It is more than precise enough for our purposes.

So we leave the discussion concerning absolute frequency accuracy behind us and wait for the outcome of my incentive of this thread : -)

SV1BDS So, your complete transverter setup is now mounted near the dish?

DH2VA Yes, I know about the NDA ... (when I reside under an ND(C)A I always say: "I don't know" <-- more polite ; -)

With the risk getting off topic in my own thread ... ah well ... ; -)

I'm not sure about your connotation of 'non zero offset', what do you mean with that? What offset, and what is allowed to be 'non-zero' ? (e.g. 8089.5 ?)

Besides TV Ku/X band transponders, this satellite carries more transponders (like e.g. S/X band ; -) but also Ka band tpx and anti jamming facilities as well as TDOA capabilities (like almost every modern geosat has to pointify jammers/intruders in co-operation with other satellites). In order to do this a precise on board clock is very handy ; -)

DH2VA I would suspect this LO generation is locked to the internal bus which has it's own (absolute i.e. Cs) reference combined with Rb ?

The PSK beacon is used as reference. With a Costasloop the (surpressed) carrier is recovered.

As a Dr. I am always into helping university students.

From what I understand from the above you want to make a precise as possible link budget tool, founded on practical data?

DB2OS Almost ... I will look up the IARU (R1) standard for you. Keep you posted.
(I was VHF-manager of VERON Netherlands, so I have the Handbook somewhere)
(albeit that it's on the net now .. it was a long time ago ... ; -)

Edit: Found it. See excerpt picture below (from VHF-mgr handbook IARU R1)

(assuming R1 is 'close enough for the footprint of Es'hail-2 / QO-100)

The standard is ... the carrier is on the nominal frequency (10489.550000 MHz). If it transmits its message, first the carrier goes to 'space' (400 Hz lower, so 10489.549600 MHz), remains there for a while and then sends the 'mark' (read: information) at the nominal frequency (10489.550000 MHz), then a period of 'space' (10489.549600), and then the carrier back at 10489.550000 MHz.

Having deployed a 50 MHz beacon for years (and being involved in the recommendation) the argument is that above 9 MHz USB is the amateur standard, so a listener (tuned zero beat in USB at the 'space') will/can/may hear the beacon as A1A (see text below).

So, deltaF = 400 Hz (or 500 Hz)

Hope this helps you.

PS. If the freq boundary of .550000 is a hard limit, it can be reversed, so that in LSB the beacon can be received in 'A1A', but I think 400 Hz 'violence' is acceptable ; -)

The proven existence of gravity waves which have impact on the size of the universe, therefore the wavelength changes, therefor the frequency ; -)

(no, just kidding)

On board LO converter (in LNB) may wobble a bit? (according to Paul)

Still overthinking the process Mike and I went through when 'fiddling' with the design in relation to Rasto's analysis about the somewhat squinted lobe (not that is may be of much importance, but it's nice to think about it)

If I remember correctly, the first try G0MJW did was to model the K3TZ patch and the subsequently added the protruding waveguide ('copper pipe'). Of course problems with the match occured. After that I suggested to insert a short in the middle of the patch.

Using an M3 bolt as short (the middle of the patch is Z = 0 + j0 Ohm), no matching problems were encountered. Increasing the diameter of the 'short' revealed worsening of the RL. We could match the real part (R) (Z = R + jX Ohm) to 50 Ohm but there was a significant part of X which could not be eliminated. The approach was then to find another feedpoint where the RL was enough.

This led to the first prototype made by M0EYT. Indeed, it had a good RL, but was it also circular (enough) ? From the Smith Chart it appeared that the 1st prototype was circular, but not at 2400 MHz. The 'balloon' (my words) was at the wrong place and the challenge now was to iteratively find a feed point where the center of the Smith Chart (Z = 50 + j0) lies in the middle of the balloon. This appeared with feed point coordinates X,Y = 8,28 mm and I built the 2nd prototype accordingly.

Several simulations followed, but it still converged to 8,28 (could have been we found a local optimum?) so that was the end value of the feed point which was published subsequently.

The patch can be seen as two antennas. One somewhat too short (-j) , the other somewhat too large (+j) to such an extent that their individual phase offset amounts 45 degrees relative to the design frequency, so that the total phase shift is 45 + 45 = 90 degrees, necessary for circular polarization.

Having a 'short' in the middle the matching method can be illustrated as a 'gamma match'. The center-feedpoint distance (so along the Y-axis, with X = 0) defines the real part (R) of the impedance and the X-axis (Y = 0) the imaginary (j) part.

With the M3-bolt in the middle (with neglectable influence) there is one variable, being the Y-value (hence X = 0). With the copper pipe there are two variables.

Because the feed point is now 'off centered', hence introducing a frequency dependent component by itself -in this case (iirc) some inductance (+j)- the whole thing gets some 'imbalance' and -apparently- some small squint.

DL1GNM Yep, I read that article with @OM6AA's comprehensive analysis of the current patch. Despite the design looks utterly simple you can see the whole thing is sensitive towards a variety of combinatorical factors (factors influencing each other, which are influencing each other, which are etc).

The POTY was NOT designed to squeeze every pB (picobell) out of the performance but after intensive modelling deliberately as a reproducible compromise without needing (complex) metalworking equipment other than metal scissors.

DH2VA Thanks for the explanation : -)

DB2OS: Thanks for your consideration.

DL1QC: That is also a possibility, just like (uWave) beacons. A carrier for several minutes and then identification/message/whatever.

Concerning frequency accuracy (which is not the main topic, but I can rename the thread so that it falls into ; -) : bear in mind that the satellite still has some Doppler in the order of a few 10 Hz due to the fact it's not totally rock stable in its 'geostationary cube'. Furthermore the beacon signals aren't generated in the satellite itself, so the reference needs to propagate an additional ~37000 km where perhaps other (propagation?) issues may arise.

So from my perspective I can't confirm if your values are (absolute) right but I also noticed some 'offset' and changes in the offset, but I didn't test it in such a way to determine if there is a 'day' or 'night' offset.

Perhaps you can measure the 11199.8 MHz (H) beacon from Es'hail-1 ? It's not exactly on 11199.800 MHz (but a few kHz off), to determine if you've the same offset pattern (day vs night).

That is actually done by somebody (don't know his/her call).

When the POTY was simulated/designed (by G0MJW) we discovered that a horn seriously degrades the performance of the S-band part of the POTY (primarily gain (<-- read: illumination of the dish and somewhat axial ratio). Mike then tried another design, a round patch with 'notches'. With this design the patch surface (iirc) could be increased so that the influence of a horn was relatively less, maintaining the necessary S-band gain (<-- read opening angle = illumination of the dish).

So there is a trade off, which is also dependent on the usage. For NB usage the degradation of the S-band part can be (more easily) compensated with increasing power and more gain/SNR with RX (with a (cut off) horn (from a LNB). However, for DATV the uplink link budget is more critical, there every 0.5 dB counts (as I have been told)

dg0opk Nah, I don't think so. That pirate will use an injection locked microwave oven in the focus of a dish, like we used to do 20 years ago ; -)

(but we were scared then . . . )

DD0KP Heiner, didn't Philips/NXP secretly store a cookie on your computer with the result the amplifier will be delivered at your home and you are U$ 10000 poorer? ; -)

FSK (F1A) is a normal narrow band mode and not referred to as 'FM' (e.g. F2A in this case). As long as the spectral impact is narrow enough, there is no problem.

The upper beacon also fluctuates

I would like to ask AMSAT-DL to consider operating the CW beacon in FSK, i.e. with constant carrier.

There are IARU-R1 recommendations concerning (terrestrial) uWave beacons using FSK, but that's an administrative issue.

My main argument is that it's relatively difficult to precisely fine tune the RX feed with a pulsating signal. Also during the two (too short imho) carriers QSB is visible (depending on my local situation and the situation above Bochum (clouds etc), which can be around +/- 3 dB and then it's too short to align within the precision needed after averaging the SNR.

I am doing measurements with several prototypes of dielectric lenses and have an accuracy of around 1 dB and experience slight differences in results when e.g. optimizing on the 11199.8 MHz beacon compared to the QO-100 10489.550 MHz beacon.

Is AMSAT-DL willing to reflect on this?

SV1BDS I see '-5 dBm' (LO) at your TX mixer. Which/what mixer do you use?

OK, understood. Here (The Netherlands) I am allowed to let an unlicensed person use my station (and therefore callsign) as long as I am present.

So we don't have 'training callsigns' etc, but i know some German laws are (very) different compared to Dutch ones ; -)