QO-100 Amateur Geostationary Satellite

I have been thinking to make a transverter fed from 2m(or HF) IC706 until I decide on what SDR to build my ultimate all-in-one radio. I think 2.4ghz isn’t so bad for short runs of trace or dead-bug wire as far as watching for accidental capacitance and impedance in the traces. I dont know if I should make the final linear amp a separate thing and put it in a heatsink Al box right at the antenna and just feed it with a bias-T on the 50ohm antenna line. Elegant would be to run everything over one line and multiplex the power and RF with bandpasses. I think I want to make a POTY to add a S band tx antenna to the Ku LNB. As fro RX I guess for now an RTL-SDR<–bias-t<–Bullseye LNB.

Since there is not already a really integrated unit I hate just buying someone else’s project off the shelf and would rather make a few crappy prototypes myself.

How did you make your bias-T for your Ku side? I have been mulling how to wipe out the oscillator noise from whatever boost I use to get the 12v to 14/17v for polarization.

How many dB to say a POTY antenna and a ~1m offset dish is needed to send up a CW or tight digital signal?
Are wifi amps really enough to show up on the transponder? What are those, something like 30dBm or less?


I just bought a Bias-T Injector from Ebay or so for a few euro and added a Buck Converter to up The voltage from 12V to the 14V i need for the LNB.

The WIFI Amp i got is 8W, well it realy does 2-3W but it is more then enought even with a little Patch Antenna on a 60cm Dish Indoors.

1 Like

Wow, it is amazing that the input of the transponder is grabbing qrp signals like that, I just imagine terrestrial noise would wash a 33-35db signal away especially with such a small antenna/reflector.

I am using a stabilized LNB with injector from DJ4ZZ in a 60cm standard tv-dish and a RPI with openwebrx as you can watch online here. The injector is fed directly from a 12 Volt solar-system.
Did not try transmitting until now.
Wolfgang, DH2WS

1 Like

You don’t need a stablized LNB, i use a normal one from Amazon and it works great, especially with simple Software Compensation with SDR Console :slight_smile:

1 Like

I love that everyone has to go back to DIY for this satellite.
I would like to get better at RF circuit design for microwave,
I used to know one of the emeritus chief scientists of Tektronix,
his name is all over tech like bluetooth.
I had a copy of the At&t manual gifted from him, now a search for that book
an amazing resource for microwave tech returns a forever list of old At&t wireless phones.

I was boning up on some microwave design and saw this as the intro in a book talking about a hoped for near future of apparently a different 1985 timeline branch than the one we passed through. They probably have a USSR and fragmented Internet though.

Hi there, i’m using the very stable bullseye. It change my setup , now I only have the LNB and a sdr receiver or the downconverter to 144 mhz for my aged ssb radio receiver. No need any gps control on the oscillator or the software compensations in SDR Console . The TX side is made by an SDR ( pluto sdr direct 2,4ghz + rf amp) or ( analogic) up converter BU500 on 432 MHz+ rf amp.
The bullseye make the difference in a simple configuration LNB + downconverter+ssb radio.
Well done othernet , thanks for this piece of hardware.

1 Like

Have somebody tried to change TCXO frequency in Bullseye LNB? I read it is programmable… Or have anybody tried exchanging TCXO of 25MHz with another one of the frequecy of 25,7872 MHz, wich gives you 432,5MHz IF?

I know some people that bought a Bullseye from me here in germany that replaced the TCXO with a 0.5ppm 24mhz one to make it useable for DATV Receiving, they said it wasn’t to hard, so a 25mhz tcxo should also work i think for a lower frequency. The LNB isnt programmable as far as i know.

I started with standard LNB in an 80cm-dish, bias-T at 12 volts and openwebrx on an RPI3 with a NooElec RTL-SDR tuned to 739,5 MHz. Works with some drifting as a proof-of concept. Now I use a standard-LNB with a heated xtal. Drift is below 100Hz/hr now. Pointing to the EsHail-Satellite at 25,5°E was easy as the Badr-Satellites sit at 26°E with lots of TV-Signals, which I could watch during dish-setup in a standard-sat-receiver for KU-Band.
Wolfgang, DH2WS

I’ve just got started on my QO-100 project. 100cm dish positioned at ground level with clear view, playing with cheap LNB and SDRPlay RSPdx with modded bias Tee to delete inductor which popped on LNB load current. Bullseye on order, plan to work on RX first to maximise SNR and then on to TX. Hope to be QRV end Feb.

Not possible with the current design. It can only be changed a few kHz and the RF SOC will not work at such offsets.

It is programmable but requires more external circuitry. It was developed to “Fine” tune in terms of 100’s of Hz resolution at Ku for hams and has the ability to store the offsets in memory. There is nothing like it on the market in terms of stability and RF performance. The SOC has a fixed division and altering that far it will significantly will degrade Image rejection and Gain (not saying it wont function but an external down conversion is much simpler and more effective, especially as you can lock the LO of the downconverter to the LNB so the LO is coherent. Remember the SOC’s are designed for DBS and although they do have some margin they have limitations

To everyone on this thread and interested in the Bullseye LNB, I would listen to @MWAVE_JON. We can consider him to be the world’s foremost authority on the Bullseye.

Thanks Syed. The lnb really is quite special and is highly under rated. It was based on a highly successful LNB that is sold to high end operators, its not a cheap LNB… For all the guys out there who believe the total rubbish that is advertised on boxes, take it from me its complete Horse S**t on 98% of LNB’s specs out there. Good MER,G/T, quality figure depends on many factors and also environmental conditions. Any NF lower than 0.5dB over the full BW cannot be achieved FACT! unless you start cooling it to <<< -40C, the best GaaS Fets are 0.35dB NF at SPOT frequency and this does not include feed and lens losses, plus the fact its optimised for over 20 percent bandwidth. Bullseye in terms of RF performance (NF, Xpol, gain stability etc) is pretty much the the best on the market, Syed asked for the impossible which was to get better stability than a 1ppm TCXO. We actually base it on a VERY good TCXO and make it MUCH better, in production its initially referenced to GPS as this is the only timebase we could trust. I am not going to say how its done but there is a lot of math and time put into the product to achieve it once it leaves the factory. It was also designed to be cost effective and built in volume so no space for OCXO etc. It is programmable through the reference port using a custom protocol, but was never officially published and it can be digitally tuned and altered real time within about =+/- 100kHz around the Ku LO in approx 100 - 200 Hz steps at Ku!
In theory a simple tracking loop can be built from SDR to LNB to auto tune out offsets to maintain lock on narrowband channels to compensate for receiver drifts of other equipment. It also has memory where users can offset and store offsets alongside all its calibration coefficients.

RF performance in MER terms (we test on standard DVB-S2 channel) there are very few LNB’s on the market that get close.

If you are going to take it apart to butcher the LO with huge offsets (with the exception of wanting an external reference port) then don’t bother, buy a $3 LNB.

To get something other than the standard output then build a simple downconverter and use the reference output to lock the LO. You will need to amplify the LO as we designed in loss to protect the LO from transients and voltages.

You should give Syed more credit for getting this product designed and completed as even 10,000 only takes us 1 day to build so he invested a lot of personal money in the development

Hope this helps

It reall helps. Very interesting, very helpfull, thanks. But I’m still not sure whetrer it can be retubed bz the simple exchange of original TCXO with the one of the frequency of 25,787 179 Mhz on order to get down/convering from 10 489,500 MHz to 432,500, which is my idea… Could you comment this, please? Of course, if it is imposible, a simple 2nd downconverter from 739,500 to 432,500 would do the same externally. Just to understand: my idea is to use only an ordinary HAM TRX on 70cm band w/o SDR and compenzate small offset of downconverting chain via “Quick Split”, as it is called by Icom and to get all the instabilities in the limitys of 1/- 10kHz, which is a range of RIG RIT…

OK, this discussion saved me a lot of effort and disappointments implementing the idea of replacement TCXO in Bullseye LNB. I’m going to construct a simple external down converter. BTW, I have already bought a $3 LNB and replaced XO in it with a TCXO of 25.7872 MHz. Resulting offset is about 6 ppm and stability is +/- 1 ppm within 8 to 60°C, while original Bullseye LNB with original 25 MHz TCXO showed offset of -1 ppm and stability of +/- 0.5 ppm. Great. The entire lab equipment locked to Rubidium reference, which has deviation equal to 3 Hz on 10 GHz against GPS reference.

As mentioned previously the performance of the LNB will degrade as it was not designed for those LO frequencies. The image band is now at 9.624G and the RF BPF and the SOC image reject mixer will only give about 20dB image rejection. But depending on the C/N of the signal it may be enough. NF and XPOL should be ok at 10.4G and the PLL will Lock with no problems. One issue you will face is the gain will be low as the output Lband stage in the SOC and the output S11 were not designed down to 400MHz, however it may well work for your use case.

One better method for you would be to modify the Bullseye LNB to accept an external reference through the second F-port, this way you can adjust the offset from inside the home and get better thermal stability on your reference. Just remember that the phase noise of the reference, if poor, will degrade the phase noise of the Internal PLL within the loop bandwidth of SOC.

It may work well enough for your application

One better method would be to use a wideband LNB that is used in Europe on some MDU installs.
These have 10400 LO and give a continuous output 300 to 2350 on 2 outputs on Vertical and Horizontal.

In theory you can also sacrifice one polarisation on this LNB and feed an external reference on one and the LNB would perform very well in this case as the SOC is different than the standard universal SOC.

In this case no downconverter is needed!