PA for 1296 MHz

Since i was finished with the transverter for 1296 MHz. I try to build a PA 45W this was not working. I start lock at ebay.com  an found a PA original from N9RIN. I by this and put it on a heatsink, it works grate. I run a few QSO with good reports. This pa was original set for class A bias 1.3A. My ide was to lower this to 400mA but when I try this I blow the transistor. It was a hard fight to get replacement for the PTF10021 infineon transistor. I bye 2 pcs from Alibaba, Kina this was not working. But on 18 December I received a pcs from Kynix.

 Perfect 23cm test was on the 19 December an now all was working even with lower bias 400mA, I measured 45.4 dBm out, not bad. The voltage to the PA is 28V so I tried to up converted it from 12V, there was a good article in Radcom nr2 2017, bayed from ebay.com, and add some filtering. Cant measuring any spurious on spectrum. A coax relay CX-140 used and a isolating extra relay Panasonic RK1-12. I has an already built RF pre-amplifier as I remember it is from  YU1AW, yu1aw. 

Now all is finished, before Christmas, hop on many more qso. next year.

PA and Driver for next QRP Transciver

I has built a driver for the PA. I started to lock at “Building a Transceiver” by EI9GQ and G4LFM. Side 118 it has a promising setup of RD06HHF1. The test showed god result. Next step was to put it together with the PA. I has little more to do with the optimize power transfer from driver to PA, but it will work just to optimize performance. In The plot from VNA it was 30dB atten from the output to VNA, its means gain in driver is close to 20dB.

PA For next QRP station

I by this PCB from K5BCQ a long time ago. Now I hade time to build and test the performance and it locks weary promising.  I measured max 42.7 dBm out.  But I think 42dBm 16W is more ok to not stretch the PA to much. The flatness is better then -0.7 dB from 1.8 – 52 MHz. It means that the work with the output transformers is grate.

I waiting for driver transistor to put in front of this module. I will put info here as son I has test the driver.

Link PA description K5BCQ

A BP filter for my next QRP Transceiver

I came across this YU1LM interesting side a time ago. Really interesting result. I build tree of this filter to check the performance, and get very close to the description but I don’t use so accurate values, that’s way I think it is worse loss in the passband. But I will go for this type.

 I has also tested pin diodes and other diode to use as switch, BAT18, BAP70-3, BAP51-02, BA592, BAP50-04 and 1N4148, I chose BAP70-03. I have read that some has problem with IM product when using diodes my test don’t show up worse with diode compared to without.

 I use 2 generators and a hybrid combiner a generator G1 in passband -12 dBm and G2 to different frequencies to test for IM -70 - +17dBm. If there is IM product it is worse without filter, so I can’t blame the diodes for that.

dokument http://yu1lm.qrpradio.com/INPUT%20HF-6m%20BP-LP%20FILTERS%20FOR%20SDR%20TRANSCEIVERS%20-%20YU1LM.pdf

The Measurment.

Stay tuned I working on PA for QRP. 

de SM4HNG

Si5351

Small wonderful VFO and BFO

I get this PCB from SM5CJW nearly a year ago. 5 Component, 2 resistor I2C pullup  1 decoupling 100nF, the Si chip and Xtal 25 MHz. Work grate directly.

Earlier i use Adafruit  module, and NT7S sw for Arduino.

Stay tuned I working on a QRP Transceiver.

My DS15001

I order this before the summer. Last weekend it was time to solder the kit. After turning on the Oscilloscope I see the test signal a short time then the screen was dark. The adj is connected to encoder did not work in both direction. After google about this, I was shore the small pcb. was mounted upside down with the encoder. The behaviour was different but don’t work. After checking the schematics and measuring around the encoder it was shore the PCB was broken in one track to the encoder. After a strap all run nicely. I fixed a 9V NiCad Battery and a box for that.

Here is the troubleshooting Trouble

A video about the Osciloscope Video

Electronic

Electronic Compass

In the winter i spent time to test a sensor BNO055. This was a success it works from start, has built in tilt compensating and calibration. Only a few rows of code. It has two register  to read the current calibration state for system all sensors and magnetic sensors all values  > 0, gives correct directions.

Above photo, first digit is magnetic sensor status second is system status 0 – 3. My next try vas to integrated with Fox receiver and it also works., no effect of metal box.

I also try a sensor LMS303, but this this I can’t get to work.  I has cod to make tilt compensation and calibration. It doesn’t  give me 180 deg. Difference from N to S direction and also W – E so I lev it for now.

The BNO055 has been on two competitions, and the first direction after start was tree spot on and the other 2 just a few degrees off, it was a 80m fox-run.

See adafruit BNO055

POWER METER

This power meter is now finished. A port is LTC5532 and port B and C is classic AD8307. I make my own PCB for this, today you can by AD8307 from Ebay sold for a few dollar.

I connected a LCD display and Encoder. Draw 5V from Arduino Uno to all tree PM boards. Output from A to A0, B to A1 and C to A2.

Then little sw doing the work.

LTC5532 can’t find a formula so I use my generator for input and read levels by ADC and put a table on my serial port than past in to the code, like this.

if(u0>3.4604)

  P1 = 8.0;

else if(u0>3.3822)

  P1 = 7.8;

else if(u0>3.3040)

  P1 = 7.6;

…………..

else

  P1= -21.2;

Then port A is working.

Port B and C. use my generator put in 0dBm and -40dBm read the level in V0 – V40/40 = x V/dB around 23 – 37 mV/db, put it in the code.

if(port==1)   //Cal Port B

{

   for(int i=0; i<10; i++)

     a1 = a1 + analogRead(A1);

   u1= a1/10;

   u1 = 5.0 * u1 /1023;

   pmSet.cal1 = u1;

}

 

To get stabile values I average 10 times.

P2 = (u1 - pmSet.cal1)/0.0242  + pmSet.atten[1] + freqComp[1][pmSet.freq[1]];

pmSet.cal1 = u1 att 0dBm input, calibrated in menu.

pmSet.atten = attenuation infront off power meter 0 – many dB integer values, set in menu.

freqComp[1][pmSet.freq[1]] = freq compensation set in code, measurd by generator @ 0dBm on varying frequencys (16 pcs.).

double freqComp[3][16] = {{0,0,0,0.3,0,-0.2,-0.1,-0.4,0.1,2.9,1.4,4.0,3.8,6.7,12.5,10.4},…….

Unit can be set dBm or W, in menu. Convert to W

if(pmSet.unit == 1) //Unit W

{

  P2 = pow(10,(P2-30)/10); 

}

Jumping over to new project, magnetic compass and PA 1296 MHz, stay tuned.

Antenna Analyser

This has now come to the END. The Analyser is checked and compared to a RIG-EXPERT Analyser. We measured antennas from 160m to 6m and there is no difference in the WSVR result compared to each other.  It’s the first time I do a measuring instrument from scratch. There is some decisions to do to get it accurate enough. It is also possible to export the sweep to a mobile phone to store the result as a documentation of the measured antenna. The CPU memory is used to 96% the IDE report! All settings is done from the encoder.

When measuring antenna and demonstration of the analyser its good to have different known loads as 50 ohm, 100 ohm and a simulated 80m antenna. See below picture.

Transverter 1296 MHz

All boards is no in the final box, The first QSO was from my QTH to JO78FM a distance of 77,1 km. Output is around 2W, Antenna DL6WU 37 el homemade Yagi 23m up in the tower.  Report  59.

I try an IF attenuator with pin diode and attenuators, but I was not satisfied with the result, so I made a small board with attenuators and relays. Its adjusted for my transceiver Kenwood TS-2000 with 5W output power.  But can be simply changed.

I also try a relay I think it was CX120.. but it was not ok for 1296 Mhz frequency. I by some from Ebay, Omron and MATSUSHITA RK1-12V, that solves the relay problem.

I has start planning for a PA to connect to this transverter, stay tuned!

It is possible to lock the VCXO 96 MHz to a reference frequency, not implemented yet. I must build a reference unit first. 

Haven’t you lock at W1GHZ its time to do that, grate inspire.

Here is how I made IF-Board and SM-Board.