USB synthesizer

At the annual radio market in Rosmalen I bought a USB synthesizer kit.
See SDR-KITS
It uses the si570 to generate signal between 3.5 MHz and 180 MHz.
This unit can provide all frequencies I need for the all-band version.

Power output is a bit low. I will amplify it using a MAR amplifier. That should not be a problem.

The synthesizer is USB controlled. I will use a PC to control it. I might add a micro controller board later.
However, I am considering to make it a PC controllable Transverter.
To do that I need some USB controllable I/O ports. I could even make a PC controllable transceiver.

The power amplifier I will use is 5 Watt amp box73 it is a 5 watt amplifier that works from 1.8 to 144 MHz. So I can even add 50 Mhz an 144 MHz.

More design considerations later.

allband version

I have investigated the possibility to build an allband version.

A device Si570 could be used to generate all signals. The CB set can be set to a fixed frequency. A microcontroller, LCD and rotary encoder can be used to control the Si570.

It should be possible to build one for most HF bands. However 28 MHz will be difficult. The Mixer will leak some 27 MHz. A suppression of 40 db is possible but it is hard to accieve. The 28 MHz bandpass filter won't provide enough attenuation.
I think that 24 MHz will be difficult too.

For 14 MHz and higher a preamp between the mixer and the bandfilter may be needed for reception.

Using a Lo signal that is higher than the reception frequency is needed for bands > 10 MHz. LSB and USB will be switched. I better use it for all bands. it will reduce software complexity. An additional advantage is that the Lowpass filters can be used to reduce mirror image reception.

80 meter version

The design can be adapted to 80 Meter. making a suitable LO signal may be a bit difficult though.

Two 23.5 MHz xtals are required (27-3.5). These are non standard as far as I know. 30.5 (27+3.5) MHz can be used instead. Those are nonstandard frequencies too.
using 30.5 inverts LSB and USB.

23.5 MHz can be generated by dividing 47 MHz by 2. You could try to use a VXO at 20 MHz. Then double this signal and mix it with 7 MHz.

A band 300 KHz wide band filter for 80 meter can be found in the Elecraft K2 design.

Almost ready

The transverter is almost ready. I could add the bias current temperature stabilizer.
I might do that in the future. It may be good enough though. I have reduce the power to 5 Watt so it is not getting that hot anymore. Five watt is enough to drive my lineair amplifier.

I tried to make contacts using it bare bones. So far I have not had any luck. It should be possible though. I have a lineair amplifier that can provide 200 watt PEP using 5-10 watt drive. I will use that amplifier to boost the signal. It should make things easier. In the future I may build a PA for the lower bands using IRF510 fets.

I will try to make a movie clip showing its receive capabilities.

New output transformer

I made a new output transformer for the PA. I used less turns and more ferrite.So I was able to use wire with a larger diameter.It worked out fine. I can get a bit more than 10 Watt output, measured after the low pass filter. Input current was 1,8 Ampere. This result is OK.

The transformer has 2*2 turns primary and 9 turns secondary. SM0VPO uses 3*3 turns primay and 16 secondary.My transformer is made by 3 2 hole ferrite cores. Brand is unknown. This is the third transformer I tried. My maximum power is limited to 12 Watt theoretically. That is OK I will use the PA at 7-8 Watts maximum. The SSB signal sounds good on my receiver. I didn't notice splatter on nearby frequencies.

I really have to implement temperature compensation. Now It suffers from thermal runaway (DC bias current increases when temperatere rises). For SSB speach it is not a problem. It is a problem when a carrier at full power is being transmitted (AM mode). AM is not needed ,I (mis)use it for tuning the antenna though.

Low pass filter and fan regulator

I just built the Lowpass filter using info from the ARRL handbook. I have used the 5 pole Chebychev filter. That sould be enough. It seems to work although I can not measure its characteristics yet. The output signal is not attenuated much. the output looks clean on a 100 MHz scope. I have to wait until my spectrum analyser project has finished. Than I can be sure.

The PA does not have enough cooling, so I added a fan. The fan starts when a transmission starts. After the transmission stops, it stays on for a few secons. I made a circuit to accomplish that.

After pushing the PTT switch I hear a tjoep signal on my SW receiver. I will have to find out why that happens. It is a bit anoying.

I found a similar power amplifier in a previous project. That PA has a better output transformer. It puts out 5 watts too but the efficiency is much better. I 'll rebuild the transformer using the same ferrite core type.Then I may get out a bit more power. That would be nice.

amplifier news

This picture shows the amplifier.

It is not 100% the same as the SM0VPO amplifier.
I did not use a thermistor in the the bias current regulator. When I find a 1K thermistor I will add it. Now I just use a 1 K resistor. The bias circuit for the driver is different too. I don't remember why I changed that.

The amplifier works but I can only get 5 watt output. I add a few capacitor to smoothen the base and the collector voltages. They are sinus shaped now. Without those capacitors the signals cointains spikes.
A problem is that collector voltages reach their maximum at 7 Watt. That should not happen. I tried different ouput transformers but the results are the same. The output voltages divided by the input volatges just doesn't match the turn ratio. Input current of the finals is 1.5 Ampere for 5 watt output, so the efficiency is low too.

I will have to investigate what is wrong meanwhile I keep using amplifier as it is stable. I did not see it oscillate, even when SWR was high. The output looks clean.
Then I checked the audio. It souns good on a shortwave receiver.

Next task is to improve power and efficiency and build a lowpass filter.