Moonbounce on 1296

[VK2JDS]

Seeing there were no posts in this topic I thought it might be good to start one.
There are several local stations active on 23cm E.M.E. so it would be good to discuss setups and schedule skeds here.
Myself I have been actually getting contacts only since the end of march, after many years of setting up the right gear . The dish went up at the start of 2003, and the steady plodding along getting all the gear, feedpoint designs, weak signal generator, power supplies, heliax, coaxial relays etc.
Then noise testing all the cables, homebrew/kit preamps etc, loss testing and isolation of the coaxial relays and making feeds to come up with something that gave good sun noise response.
i damaged preamps from inconsistent relay switching, cooked the input of the vk5eme transverter resistors with a tad too much drive etc , then got a ic910h. with most of the testing done using the transverter the ic910 fitted in nicely and dropped the noise floor that little bit more.
if you get one , put in the high stability option.
i have had some good contacts into europe on the stations with dishes of 4.5metre and above level. i run a 4.6metre. but the really big stations can work people with 2.1 metre dishes and 10 watts out of a ic910!
Now the europeans are big signals on jt65c but the best i have seen is our very own VK3UM. now thats a massive signal Doug puts out. Clearly audible CW and a absolutely white trace on spectran. With limited equipment and antenna budget you could hear/see Dougs signal.
73's

[VK2JDS]

just a bit of an update, yesterday i successfully worked SM5LE, he runs a 2.2 metre dish and 200 watts. i was decoding him at -20 quite comfortably but he wasnt getting that good a signal from me. we made it a 2 way qso, but i think i need another one of Alans amplifiers to be able to easily work stations with dishes that small.
but we have proved it can be done
73's

[VK2JDS]

some more observations:
i have been using a variety of dish feeds but had settled on a doublequad with reflector plate which proved very successfull for EME.

after destroying a couple of gasfets in the vk5eme kit preamp with the circular feed due to insufficient port to port isolation in the coax relay i was reluctant to go back to the plate feed, but a partial rewind of a weather damaged coax relay coil brought the good relay back to life!.

plate feeds require the preamp to be disconnected from the plate and put into a shorted N connector to protect them when you shove oodles of power into the plate they were connected to.

so, off we go , i checked the through loss using my ultraweak signal source from the shed, then transmitted through it and measured bugger-all loss, so it went back into service after a contact clean.

I tried pings off the moon, i tuned around for several days but nothing turned up, the noise floor rose significantly when i plugged it in, and it reduced a lot as i moved my hand over the feed standing on a huge stepladder to get to the feed with the dish as low as it goes but still no pings back from the moon. i did get some significant sky noise floor changes so i know it works but who knows if its as good as the canadian circ feed designs.

So I lowered the dish and changed back to the doublequad feed. This time i set it vertically polarised as there was some interest from some VK horizontally polarised stations in Adams chat logger(great resource) .
With vertical reflecting off the moon i might be able to get the horizontal pol stations. Straight away i was getting weak visible pings on spectran.

Tonight at full elevation i am still getting consistent visible pings at the end of each jt65c transmission.
This still amazes me. wow. only 65 watts.
work in progress.
any comments?

[VK2OMD]

... This time i set it vertically polarised as there was some interest from some VK horizontally polarised stations in Adams chat logger(great resource) .
With vertical reflecting off the moon i might be able to get the horizontal pol stations.

Why?

[VK2JDS]

Faraday rotation.

http://www.nitehawk.com/sm5bsz/polarity/emepol.htm
SM 5 BSZ - Selecting Polarisation for EME

heres a quick clip from it "If faraday rotation is 45 degrees, a horizontally polarised antenna will produce a wave that is twisted by 45 degrees when it passed the ionosphere. After reflection at the lunar surface it will come back to the same point above the ionosphere, still twisted 45 degrees. After passing the ionosphere a second time, the signal is twisted 45 degrees again in the same direction, so it will be vertically polarised and not audible by a horizontal antenna."

the paper goes on into greater detail about it and other reasons for polarisation change, i just roughly simplified it by saying i would tx on vertical to work horizontal stations... its a lot more complicated, but its a bit of a start.
73's

[VK2OMD]

Faraday rotation.

http://www.nitehawk.com/sm5bsz/polarity/emepol.htm
SM 5 BSZ - Selecting Polarisation for EME

heres a quick clip from it "If faraday rotation is 45 degrees, a horizontally polarised antenna will produce a wave that is twisted by 45 degrees when it passed the ionosphere. After reflection at the lunar surface it will come back to the same point above the ionosphere, still twisted 45 degrees. After passing the ionosphere a second time, the signal is twisted 45 degrees again in the same direction, so it will be vertically polarised and not audible by a horizontal antenna."

the paper goes on into greater detail about it and other reasons for polarisation change, i just roughly simplified it by saying i would tx on vertical to work horizontal stations... its a lot more complicated, but its a bit of a start.
73's

Dave.

I don't think that the reference actually states that Faraday rotation is 45°, but it does say "If Faraday rotation is 45°,..." (my emphasis).

It does suggest that Faraday rotation is proportional to f^-2, and that it is dependent on elevation. He gives examples for 144MHz at 180° at the zenith and four times that at 10° elevation.

My interpretation of his article is that Faraday rotation is not generally 45°, that it is variable depending on frequency and geometry, and combined with other polarisation shifts in the path, that optimal tx polarisation to a horizontal station might be anything, and that the difference between best as worst angle of linear polarisation will get less as frequency increases.

Have you tried to rotate your feed for maximum received signal strength or S/N (different things)?

Owen

[VK2JDS]

Yes the paper indicates its quite variable, i just took a guess.
As for moving the feed its a bit tricky to do 'on the fly' due to the size of the dish and getting access to the feedpoint so a switchable/rotatable feed is a project for the future.
A better circularly polarised feed needs to be built. Looks like the link to the pics of the plate feed i tried are down but they were popular for satellite work and you could stack bands on them as well as being good for illuminating tvro dishes.
not so good on my dish though.
tonights transmissions show weak but clear spectran pings again, vertically polarised doublequad.
73's

[VK3HZ]

Faraday rotation varies with time, and the rate of variation is inversely proportional to freq squared. You can't predict what it will be, and on the higher bands, if you're out of luck, you can end up being "locked out" for lengthy periods. So, if you're using linear polarisation, it doesn't really matter if you're H or V - sometimes it'll work and other times it won't:
http://www.ifwtech.co.uk/g3sek/eme/pol4.htm

If you're serious about EME, you should go for a circularly-polarised feed, and the best choice at the moment seems to be the Septum feed. A square Septum with a circular choke ring is probably the best bang for your efforts:
http://www.ok1dfc.com/eme/emeweb.htm

BTW, if you want to evaluate the performance of your system, get a copy of the VK3UM EME Calculator to work out what your system should be doing:
http://www.vk3bez.org/vk3um_software.htm
Then get the VK1OD Noise Figure Meter software and do some Y Factor noise measurements to see if it's up to scratch.

Regards,
Dave.

[VK4CDI]

Slowly trying to be QRV on 1296 EME. At the moment, 5EME txverter, MGF 1302 preamp and 50W SSPA mounted on mast to a 45 El loop yagi. Will get serious about a dish when I get back from Hols. 3 - 3.6Mx dish and a OK1DFC Septum feed are planned.

Cheers
Phil VK4CDI

PS
I have a 3.7 mx dish on offer in Sydney, any body coming this way with a Ute or a truck? Couldn't find any courier or removalist interested!

[VK4GHZ]

"Hands up all those active on 23cm EME!!"

Would like to, but it became cost prohibitive... by the time you add something like an IC-910 and a decent PA into the mix.

I don't want to go down the track of kits, etc, because they have almost ZERO resale value.
Been there done that with other transverters for other bands.

[VK2JDS]

yeh, the 910h's with the high stability option really makes it happen, transverters are great for terrestrial and i have used mine for EME, but its the stability that counts for jt65 and internet time sync on the laptop too.
set your afc tab on jt65c to ON.
last nights experiments in cloud were dissappointing, i could still visibly make out where the moon was and had it aimed up ok , but no spectran visible return pings. so its likely a lossy path due to all that moisture/clouds and some help from faraday rotation again.
as a bit of a comparison the sun noise though thick clouds today was well down on what i get with clear sky.
usually the noise floor is near-dead quiet until you push the dish up at the sun , then the noise jumps up from almost clear to fully white on spectran , and the 910 goes from s0 to s4-s5

so in clouds its still a big jump up from rx noise S0 with a nice waterfall, to the sun noise of still s0 but nearly white spectran display!

its raining here tonight so i am staying inside by the fire. its just above zero outside and the big dish is clamped down
in the shed i am cutting up some sheet aluminium for the septum feed, work in progress
73's

[VK7XX]

VK2DAG on Thu Jul 17, 2008 2:27 pm wrote :-

"Hands up all those active on 23cm EME!!"

I have just taken delivery of a 910h with 23cm module so, at some time in the future, hope to become 23cm eme active.

[VK4XA]

Keep an ear out for John, VK4TL, he's active in FNQ with a 3.8M dish and 300 watts. (or thereabouts)

John is just starting to get active on EME again after an amplifier and EME station rebuild.

I am pretty sure he's a member on here, so check his profile and give him a shout via email for a sked.
He doesnt have or use jt65, he prefers CW.

Cheers,

Jeff C.
VK4BOF

[VK2JDS]

Yes , today as 6 metres was open I spoke to Trevor vk4zsc who is working on some gear for John vk4tl and he will pass on a hello and hope I can work him soon to help test his gear out.

Septum feed. Its a work in progress. i will have to work out how to post a picture but its at the stage of fitting the end cap.

Making the Septum in a home garage setup.
refer to the link in the above posts then go from there to here:
firstly the materials: I like recycling so the metal recyclers is my choice. A scrapped large aluminium roadsign was a good start for the metalwork. choose one with the fewest bullet-holes.. lol. its 1.5mm tensile aluminium and a gas powered blowtorch of some sort and paint scraper will be required to remove the reflective coatings.
(you could heat it over a campfire) .make sure you set your measurements to make the uncoated side to be on the inside of the feed assembly. (you want the clean side inside)
using a vice with scrounged angle iron extending a foot either side you can bend the aluminium with a lump hammer, or utilise a sheet metal bender if you can. I cut mine to size with a cheap angle grinder.
the angle iron and lump hammer will be necessary anyway even if you can use a bender to sharpen the inside corners of the bent metal. The septum piece is a sheet angleground to the right shape as shown.
Once you have cut the sheet aluminium and bent it to size its time to try and fit the 2 halves and the septum piece together. add the inserts made from 1.5mm strap or roadsign offcuts and temporarily drill and clamp it together at either end and this will allow you to assess how close its going to go together. if its like mine it will require some bending and adjustment to straighten it up , dont worry, it can be fixed. by clamping and hammering you can get the sides straight and even. Tidy up the edges where they clamp to the septum piece and spacer by hammering in the vice.
Thank goodness its not stainless steel!.
Once you have the ends together drill the clamping holes down either side through the septum piece , actually i setup one half with the septum over it and drilled it-clamped it then added the other side and then measured and drilled using the trusty makita cordless drill.
using hardware store machine nuts and bolts join the 2 halves together. use small washers either side of each nut and bolt to make joining easy.
Now to add the transmit and receive connectors I used N panel mount sockets and 60mm of 6mm copper tube ( refrigeration line) scrounged from the recyclers soldered to the centre pins.
using 1/8"countersunk brass screws and countersinking into the sides i mounted the N bulkhead sockets.
now its time to add end the endpiece, thats tomorrows job
73's

[VK2JDS]

today i drilled and mounted the end plate with the brass screw threaded adjusters. they were easy to make , drill a hole smaller than the brass 50mm long screw, then scrounge a steel screw and mount it in your vise. then make a thread in the aluminium sheetmetal by turning it back and forth till its all the way through. if its too tight run the drill through again and move it about a bit to enlarge the hole a tad.
For the heads of the adjusters i cut the heads off 2 spent .22magnum casings , cut down the side of the cartridge side and flatten them out with a lump hammer. Cut them into a circle and they give a good tensile 15mm brass plate easy to solder to the brass screw adjusters.
Use a gas soldering iron setup from dick smiths and attach the blowtorch head piece, blue flame is what you want.
now to mount the end plate find the smallest self tapping screws available or cut some long/thin ones off. thats what i did, it makes a good seal.
so, how did it go? aweful actually.
VSWR was very poor. measuring the isolation / power by using a detector on the opposite port showed a dip on each port with the adjusters. seems it works at a lower frequency ok.
tuning the frequency of the IC910 as low as would go really improved the VSWR, so i decided to cut the dipoles down to 5.7cm each and see how that goes. they are cut and should be close to resonance, testing? well thats tomorrows job. another hour of reassembly and testing. i will post the correct lengths and details once its tuned.
edit : its now shimmed with a 2mm thick piece of 12mm strap on the transmit side which was bent a bit short of the required 74mm, i added extra screws along the joints either side to reduce the spacing to 3 cm, and it tuned up just nicely with 5.7cm dipoles. Vswr dipped perfectly on both ports with the tuning screws in the same spot for each side , reversing the input and output and its dead-on. so , a good design and now its time to mount it to the 4.6 metre dish feed arm
73's

[VK2JDS]

Update:
the septum is now mounted on the dish, i extended the feed mount arm a metre using U section aluminium and tek-screwed a alum plate to it from the bottom screwmounts of the septum join. the heliaxes run along the arm and up into either side of the feed. the lna is isolated by the coax relay (which shorts the input on tx) at the same time connecting the rx port of the feed into a 25 watt 50 ohm dummy load.
testing:
VSWR looks good, sun noise looked a bit double humped so i moved it back out from the dish an inch. Tests on the moon showed it to be a tad broad but pretty close
To ping the moon I used cw with split mode on the 910h, typically its over +3khz with the moon rising upwards from the east.
I could see my own pings on spectran most of the time as vk2xso (who had come out to stay a night) fractionally moved the dish as i transmitted in an attempt to pattern the feed.
we did actually Hear some pings but no one to work off the moon as it was only open into America .
Europe is the best place for EME contacts on 23cm.
Tonight I moved it back another inch along the aluminium rail and its sharpened a bit more, now i am hearing my moon return signal tone slightly stronger than last night and the spectran trace is quite white when i have the aiming right.
Comparing the feed position with that of a Cband tv receve LNB, i would have thought the septum might have been best an inch closer to the dish, but its fine where it is now. the design of the septum with its port isolation means you only need a modest coax relay for rx switching like the vk5eme kit relay. its cheap and effective.
its a work in progress
73's

[VK3HZ]

If I can make a few comments based on my investigations:

The lna is isolated by the coax relay (which shorts the input on tx) at the same time connecting the rx port of the feed into a 25 watt 50 ohm dummy load.

You'd be better having the coax relay switch the input of the preamp to the 50 ohm load. Then the load also becomes a (roughly) 300K noise source that you can use to check if your preamp is working. The Rx port of the Septum doesn't need to be terminated - it can be left open.

... its sharpened a bit more, now i am hearing my moon return signal tone slightly stronger than last night ...

You should try to adjust the feed position for best Sun noise and try to do it in one session as both Sun Noise and Moon conditions can change significantly from day to day.

... the design of the septum with its port isolation means you only need a modest coax relay for rx switching like the vk5eme kit relay. its cheap and effective ...

and relatively lossy meaning that you'll significantly degrade your receive performance. You should use the lowest-loss coax relay you can find such as one of the SMA microwave relays.

BTW, what isolation did you measure between the Septum Tx and Rx ports? Did you measure it again after it was mounted on the dish?

Did you download the VK3UM EME Tools? These will allow you to model your system, see what effect changes (such as the 5EME relay) will have on performance, plus estimate how your system should be performing to know how many more dB you can achieve.

Regards,
Dave.

[VK2OMD]

In assessing the antenna performance, it serves well to remember the fundamentals of Directivity and Efficiency (/Loss/Gain).

In decibels, we can express gain as Gain=Directivity-Loss.

To achieve high Gain, you need high Directivity and low Loss.

Antenna pattern is evidence of Directivivity. If you cannot achieve the expected Directivity, you will never achieve the expected Gain.

So, calculate the expected beamwidth (evidence of Directivity for this type of antenna) and adjust the feed / measure beamwidth to achieve expected beamwidth. (ROT: half power beamwidth of a parabolic dish~=70/D*λ°. I think you previously said you are using a 4.6m dish so half power beamwidth should be close to 3.5º.)

Then perform some Hot/Cold measurements on known sources (eg ground if you can point at it, Sun etc) to determine G/T and compare it to expectations.

The article Plotting antenna pattern from Sun noise describes Rex's measurement of beamwidth of a 13cm band antenna by measuring received noise power during a transit of the Sun over a stationary antenna.

Narrow pattern does not prove high gain, but it is a very strong indicator of directivity which is a prerequisite for highest gain.

Owen

[VK2JDS]

HI Dave, some very good points there : the minikits relay is rated at 0.5dB loss at 2.5GHz, while i have one in the shed after a failed experiment with the plate feed (they dont like interfacing to a circ-plate type feed at 20 watts) i stopped using it.
the one i am using at the feedpoint is a broadcast microwave type rated to 3 GHz, 100 watts. amphenol 318-10430-4, its 48 volt so needed a charge-pump circuit to operate it off 12v, its the best of what microwave N coaxial relays i have and was donated by a fellow EME'r, the relay at the radio end of the heliaxes is a cx830a mpx-k-26 which thru-loss measured insignificant but data is difficult to find. the problem i am worried about using the 318 relay with the preamp switching into the 50 ohm load is that the relay then places a dead short directly across the receive port. no, i havent re-measured the return loss on the ports since dish-mounting, so will do that this weekend.

i did try it on sun noise, but during the working week its gone before you get home, i was hoping to use a radio star as the sun seems wide . I have done some radio astronomy with this dish in the past on the hydrogen line, unfortunately centA is right over the jolly radio hut! so its got to be something north or go hunting for something around galactic centre if i get time.

measuring return loss using the cheap avair meter requires using the absolute minimum power you need to get the meter to full scale on transmit, before switching to reflected. if you use the amplifier and you have the avair at the feed or even next to the amp doing loss measurements rf gets into the meter case.
as an example i was tuning the amp and the avair was sitting disconnected on the shelf near the amp, it read half scale without anything in the connectors!

[VK3HZ]

It really shouldn't matter what you do with the Rx port when transmitting. The isolation between ports should be >>20dB so it'll have virtually zero effect on the Tx port and power / pattern / circularity ...

Sounds like the Avair might be at its limits - an HP432/5, directional coupler and a handful of attenuators could be a good investment!

Regards,
Dave.