The 1090MHz filters arrived and i measured this variant plus one without filter.
Closer to the filter the response looks quite ok:
Far away from the passband the supression degrades. Especially very high frequencies pass the filter quite well.
LNA 1090M span 3GHz
This is the assembly with 47pF capacitors and a 27nH coil for supply of the MMIC. Below 200MHz does not pass the capacitors well. There is even some remaining gain at 3.4GHz although its a 1.6mm thick FR4 PCB.
This weekend i did some new trials with QO-100. In the picture you can see parts of setup i used. This time the TX setup was purely classic. You can see it in the grey box in the foto below. FT-817 on 2m as IF followed by an upconverter made from some old UMTS measurement equipment block (silver box in the middle). It does dual upconversion with high-side LO. The LO consist of an ADF4351 each (the two smaller boards above the converter). The reference is a 96MHz OCXO of G8ACE design (left of the PLL boards). The converter block generates up to 200mW on 2.4GHz. I drive a WLAN PA (black box) that probably generates 1W. And the final PA is a MRF21030 PA (on the big heat sink on the right). It provides about 10W in SSB. The antenna is my 60cm offset dish with the DJ7GP patch (the one that has the round patch element). Left to the grey TX dish you can see a black 40cm camping dish with Octagon LNB and TCXO modification. But instead i used the BATC websdr this time. The reason is the amount of cable i currently have. I use 3 power supplies (28V, 13.8V, 5V). Reducing the number of required cables is one of the next steps. However i was pleased to do some nice QSO.
Since quite some while there are 5.7GHz PAs intended for video transmission from drones to ground. Those PA are made with 5GHz WLAN IC and very cheap (below 30Euro including shippment). These PA can be used for 6cm amateur radion but need some modification upfront to generate a usable amplifier.
The PA comes with heatsink, small fan and a short power cable.
The PCB side has a metal cover that has a sticker with some technical data. The PA is supplied with 12-16V. It is intended for 5.7GHz operation. The sticker also mentions 2.4GHz but the PA will not work there. The specified output power is 2.5..3W although the vendors usually state 5W. You will not reach 5W.
Removing the metal cover shows that the PA PCB contains two Sige 5004L WLAN amplifier IC, some splitter/combiner, an input attenuator and a power supply circuit. The connectors are Reverse SMA (RSMA) and they are mounted really poor. I would not even try to use the PA with those connectors and the air gaps to the PCB.
Now the silly RSMA connectors are removed. The pads need to be cleaned. Also clean the ground ring to prepare re-assembly of the shield later on.
Solder some new SMA connectors. The grounds of the SMA connectors shall touch the top side of the PCB. You need to shorten the ground connectors a little bit. Likely you do not have connectors intended for the PCB thickness. You may solder the other side for mechanical reason. Solder the inner pin of the SMA last in order to prevent breaking the RF trace.
Some attention needs to be put on the thermal conductivity between PCB and heat sink. Removing the PCB from the heat sink shows that cost optimizations safed the thermal compount which makes the heat sink almost useless. So add some thermal paste to the marked areas below the amplifier IC. Probably the switcher IC needs some cooling as well.
The PA PCB contains a input attenuator. The default attenuation is around 18 to 20dB. If you have less input power you might want to change it. For my measurements i removed the 3 resistors and soldered a 0-Ohm upside down to the middle resistor position. After everything is modified and tested you will likely want to re-assemble the metal shield.
The PA has about 27-28dB gain and my sample achieves 3W output at 7dBm (5mW) input and a current of 1.3A at 12V. Saturation might be somewhere at 4W with 10mW of drive and 1.5A current.
I was working the contest from home and part time. The 23cm transverter with 10m IF and the HiQSDR work pretty well. I used the Quados4 antenna since the 36ele ordered was still stuck in customs. ODX was HA5KDQ and in total 41QSO and 6900 points raw score. I also made some QSO on the higher bands on request. 3x 9cm, 1x 6 and 3cm each.
Last days we have strong tropo conditions. Currently i run a Openwebrx with FT8 monitor on 2m and although i only use a small vertical monopole the maximum distance of reports are >800km. Some 2m FM relais from Hamburg was audible very loud. The 70cm band was full of repeaters (i think most of the signals were DMR). This evening i gave 70cm a try. I was surprised to hear LA1UHG beacon from JO59FB (>900km). Other beacons heard: DB0VC (JO45), OZ7IGY, OZ5SHF, DM0UB. I worked SM7LCD in JO86 over 600km with only 30W and Quados-6 antenna.
I have an older OSLO LNB which, was my spare for the websdr. Now i want to use it for QO-100. To improve the temperature stability a TCXO was fitted in place of the original crystal. Find some pictures below.
The type is a TXC 7N-26.000MBP-T from TXC. It works from 2.7 to 5.5V. Luckily the LNB works with 5V from an 7805 regulator. The TCXO was supplied by Michael DG0OPK. He suggested this type and already modified a OTLSO some years ago.
Some warning: This might not apply to current types of the Octagon LNBs. As far as i know you will likely get a 25MHz crystal version if you try to buy one. It will not be possible to use the 25MHz LNB with a lower reference frequency.
LNB with NXP TFF1017 IC could be an alternative as far as i know. I have no experience with this type. Here is some reference:
First i removed the crystal. It is necessary to use hot air because otherwise the pads might be damaged. But they are still needed. Also the 0 Ohm resistor over the two traces that go between the two pads needs to be removed. The trace connected to the former crystal pad closer to the RF section needs to be disconnected. This crystal input will be left unconnected.
OSLO LNB with crystal removed
Now a insulated copper wire can be connected to the pad of the former 0 ohm resistor that leads to the converter IC. The capacitor to ground at the crystal pad close to the 7805 regulator needs to be replaced by a short. The other crystal capacitor should be replaced by 100nF.
OSLO prepared for mounting TCXO
Now the TCXO can be soldered. It is important to avoid shorts of the pins of the TCXO towards the PCB. I soldered the TCXO slightly elevated. It is important to keep distance from the border of the PCB because the cap of the LNB still needs to be mounted. The insulated copper wire can be connected to the output of the oscillator which is pin 5 (right of the row of capacitors). The pad with the remaining capacitor towards ground needs to be connected to the output of the 5V regulator. It is the regulator pin with the 0 ohm resistor.
The pins of the TCXO are assigned as: 9=VCC (left of the capacitors), 4=GND, 5=Output.
Below i add some short videos i made. The first two show transmit and receive operation of the PlutoSDR in the 6cm amateur band. I used the great sdrangel software. The opposite side is my normal 6cm rig consisting of a FT-817 and a DB6NT transverter.
The third video shows 3 of the DM0TUD microwave beacons that are located close to my home. It is from right to left: 13cm (FT-290), 3cm (FT-790R2), 9cm (FT-817).
German BNetzA released notifications 414/2018 and 415/2018. Those extend the permission for class E operators to work in the 13 and 6cm band as well give a new permission for 4m band access in DL. On 4m class A will be allowed from now on to Dec.31 2019 to work between .150 and .200 with 25W ERP max.