Does it matter if the circulation plate is at the front of the feedhorn or does it have to be some mm deeper in it??
I'd put it 20-30mm, or so, in to keep clear of the phase centre of the feed.
Also, does it matter how for away the plate is from the lnb antenna-pin?
I'd keep it a wavelength or 2 away.
I'm surprised I managed to find this, not seen for a long time. Now at 42.7Mkm and presumably on the LGA. It doesn't seem to be locked to a DSN ground station.
RX on 8414.716MHz
0.2Hz BW, 4 x averaging
If you put all the parameters of each stage into the calculator and try different configurations, that will tell you what is best. We really need the full block diagram of your system to understand how to optimise it.
Attached are pages from a 1987 Avantek catalogue which show similar numbered amplifiers. They are packaged differently to yours but it suggests yours may have an NF <3dB.
The most important amplifier is the first one. The overall noise figure is established in the early stages. Draw a block diagram pf the system with the gains and noise figures of each stage. Then enter into a cascade noise figure calculator like this one: https://www.everythingrf.com/r…se-figure-gain-calculator. Then you can play with the architecture by moving blocks in the system and re-entering into the calculator. That will let you see how changes to each stage affect the overall performance. If you have a lot of attenuation (i.e. long IF cable) you will need reasonable gain before it. Normal LNBs have 50-60dB gain.
What is the architecture of your whole system? If your post-mixer amplifier NF alone make a big difference, it suggests you need more gain before the mixer. However, if you do the cascade noise calculation what's needed will become clear.
I got one, but i dont know the gain and the noise factor of it and the needed voltage, maybe someone can tell me?
At a guess, judging by it's age (The Avantek name disappeared ~20 years ago), it's probably around 3-5dB NF. Amplifiers of that age were often designed for 15V but it's hard to be sure. I'd test it on 5V and increase the voltage until the gain did not change.
Now at 67.0Mkm communicating with Cebreros
-145dBm at the feed of the 1.2m dish
Almost twice the distance, at 41.8Mkm, compared to my last post. Still very strong at -132dBm at the feed of the 1.2m dish with the modulation still visible.
Not showing on DSN Now but a good signal.
RX tuned to 8406.707MHz
66.5Mkm
A lot weaker tonight. It's possibly switched back to the LGA.
8426.0005MHz
16.1Mkm
I use a 34mm copper pipe instead of 28mm, maybe that is not so good...
34/35mm is fine for the milsat band. Here is one of my early X band set-ups, based on 35mm pipe and a 76 to 35mm reducer as the feed. You might just be able to see the squeeze of the pipe, it does not need much to get the differential 90 degree shift. I now use 28mm which covers 7-9GHz reasonably well.
I've added a picture of the PF feed using 35mm pipe. The scalar is based on Luis Excel sheet linked from post #2.
You can scale the design to match the waveguide. It is explained a little more in this article: https://www.qsl.net/ct1dmk/cp_feed_dmk06.pdf
Paul uhf-satcom makes polarisers for 8.4GHz and they do still work at 7.5GHz.
I made my polariser by squeezing the 28mm pipe in a vice, a little at a time, to optimise the null from a signal which had the opposite polarisation. The best description of the operation of this type of polariser is in these 2 articles: W2IMU 10GHz Part 1 & 2 here https://www.microwavers.es/ind…oteca/category/16-antenas
Getting stronger, now at 16.1Mkm
Nice work and welcome to the forum.
I used the Excel sheet from here https://www.qsl.net/ct1dmk/mw_low.html to calculate the scalar rings for a PF feed. It worked well. There is also information on squeezed pipe polarisers. I use one but it is literally a squeezed pipe (using a vice).
Have a look at qwer's posts in the X band data and GS sections for signal levels.
Do you have the feed below the focus? If you don't have inclination adjustment, you may find to better to put to the side of the Ku feed.
Stronger tonight, perhaps on MGA vs LGA before (~6dB difference). DSN Now was showing -136dBm on DSS65 Carrier earlier this evening but the their display is off now.
Waterfall 0.4Hz BW, 2 X Averaging
Now at 17.0Mkm
-133.5dBm on DSS55, Madrid at 181.4Mkm (almost twice the distance compared to the last post)
Waterfall 0.4Hz BW 4X Averaging
1.2m offset dish, cone feed, squashed pipe polariser, 0.8dB LNA
Congratulations to the OSIRIS-REx team for the touch and go collection (TAG) yesterday.
Unfortunately ORX was out of sight to Europe at the time. This plot was taken whilst it was downlinking to Madrid a few hours before the TAG.
8445.609MHz, -121dBm on both DSS54 & 55, Madrid.
339.2Mkm
Now at 17.8Mkm
8426.007MHz 21.4MHz IF, tuned to 21.405MHz
0.4Hz BW, 4 X Averaging and 0.2Hz, 2 X Averaging
Thanks for your reply, I understand. ST are a pain as they tie people up with NDAs and it is very difficult to get any information on many of their chips. I'm presuming that there was a lot of noise on the supply to the TCXO which translates to poor LO phase noise and a degraded SNR. That will be especially noticeable on narrow signals and things like 32APSK etc as you have shown.
Hi satesco That's very interesting. Is Y1 a TCXO? If so, I presume the extra capacitor is to decouple its supply. Is that correct? or is it decoupling the line with the diode with the connection to Y1 being ground?
I don't have a -X only 6983/6903 which have a 30MHz crystal with the oscillator on the main chip. Perhaps it is worth adding some more decoupling to the 6983/6903
8420.349MHz 63.6Mkm
At last, I found it. Only possible due to low Doppler change.
1.2m dish, 8426.012MHz, 21.4MHz IF, 0.4Hz BW, 4x averaging
0.7Hz BW, no averaging added for comparison
19.1Mkm
