part filter bandpass testing

Made a transformer to match the 100 Ohm output impedance of the Adruino controlled AD9850 DDS board  to the 50 Ohm input of the filter. the output is loaded with 50 Ohm (parallel two 100 Ohm resistors)

 Fired up the RF generator and used my $20 dollar ham fest bought  20Mhz oscilloscope as the detector.

look like the  band pass is around this frequency.

the reading was barely there on the scope . the sensitivity of the scope at max 5mV/cm and the 20Mhz range at the edge. Boy, this is  going to be unusable data collecting.

                I just step through the RF generator range at 100 Hz steps from 19.994 to 19.999 mhz.
the scope blurr shows it blinked very fast indicating the filter is sharp.
                I must add an amplifier to this  to buffer and amplify the signal.

Took a quick reading  on the input side of the filter and the output side readings. te input was 150mV and the output was 60mV.



        20 Log 60mV/150mV=-7.96 db  loss

time to build an amplifier with 50 Ohm in and 50 Ohm out. remembered VK5TH site has one circuit.

http://vk5tm.com//homebrew/dds/dds.php

click on the schematic to enlarge and click back to the blog

click data t o view

Toner resist on copper board

June 30 2015
Finally  received the 100 20Mhz crystals from Malaysia after a month wait at 9 cents a piece with shipping included.
         I begin the project by making a G0UUR board  for testing the crystal frequency  and make a list and choose close matched crystals. I needed 8 crystals  for the band past filter.

First test of the oscillator stability 

 



First test of the oscillator circuit on the etched PC board and found the oscillator level a bit low  and erratically  quit. It was caused by two voltage divider capacitors being too large a value  from the base to the emitter and ground . I changed the values from 1000pF to 220pF and then the circuit worked very stable and the frequency didn't drift.

Ready to tackle the 99 crystals ( short shipped one ) and write down the value and label the crystal.

Got about 75 done!

    Used Open Office Calculator to sort out the Khz descending values and found  eight  crystal that is 58 Hz in differences.  This should be okay to make the  20Mhz bandpass filter.

  Now I needed a test board to build the filter , the matching input attenuator and a load with an amplifier to monitor the output.
             I whipped up a pattern with ExpressPCB
software and ironed onto a pc clad  the toner.

                                                                                       a quick etch and the board is made.