Designing A Crystal Ladder Filter

I am building a direct conversion receiver for VLF.

I require a filter with a 3db bandwidth of 400Hz

I will describe the software used and show the resulting filter electrical and physical characteristics.

1. Software

The software being used is freeware and is called dishal2025

Here is a link to the pages of G3JIR where the software of DJ6EV is described and linked. Use this page to download the dishal2025 release .and  become accustomed to the program using the numerous descriptive links.

G3JIR Website Dishal2025

2. Crystals

I have decided to use an IF of 10MHz (or determined by the crystals) in my receiver. I purchased a quantity of 50 crystals form a supplier in Western Australia.

The initial stage of design is to obtain the electrical characteristics of each crystal

Here is the main dishal2025 screen (this is showing my entries - when using for the first time fields will be blank)

Select xtal and then G3UUR method revealing the following screen

This shows the basic frequency shifting and measurement to determine crystal properties.

The following circuit should be used

Each crystal is tested and the resulting properties are recorded. I put each crystal in a poly bag and write results for Lm, Cm and series frequency of the bag.

Here is a good reference site explaining the process

WA5BDU

The dishal help file is in German but here is a link to an English version of the help file

dishal English Help

3. Filter design Stage 1

The next thing to do is use the QER(G3UUR) function.

I am going to use six crystals and I take the mean of the series frequencies making sure I am using crystal of similar characteristics.

cp is the capacitance provided by the supplier of the crystals.

lm is as measured

Series frequency is as measured

Calculate provides the fm, ck and zin/zout

4. Filter Design Stage 2

Now using the main dishal window enter your needs and results

cm as calculated

series frequency is the median of the six crystals

cp is series capacitance of crystals (provided by supplier)

B3db is the desired 3db bandwidth (400Hz)

PB ripple I desired 0.2db

number of crystal I used 6 and this determines the filter response

display frequency span select to suit

Hit calculate and the program shows the frequency response and capacitance values need to make the filter

The filter design uses this format for six crystals

My results are

Ck12 289pf

Ck23 370.8pf

Ck34 383.2pf

Cs1   370.8pf

Cs4   1175pf

These values are important and should be generated using multiple capacitors to get as close to calculated as possible.

4. Impedance Matching

The final stage is to determine a good impedance match.

The calculated input/output impedance is 55.2Ohms

From the dishal main window select  LC-Match

The following window will appear

You will probably find the transform from R2 to50Ohm R1 is already selected, if not enter R1 50Ohms.

Hit calculate and the results for either equivalent circuit will be provided.

I use the program Mini Ring Core Calculator to design coils and the link is here

https://mini-ring-core-calculator.updatestar.com/

Using the above result's the inductor wound on a T37-6 core requires 9 turns for 0.26uH and I use a small variable capacitor. This can be adjusted to get symmetrical filter responses when measuring real world filter responses.

5. Final Design

Here are two pictures of the completed crystal filter

6. Measured results

Using a NanoVNA here are two images showing the measured frequency response of the completed filter than was designed with a 3db 400Hz bandwidth

7. Final Comments

Dishal2052 is an extremely accurate freeware program for designing crystal ladder filters.

It is very important to characterise your crystals and it is very important to build using exact as possible capacitance values by paralleling numerous capacitors.

I hope this blog has provide the reader with the tools to dip your toes into this valuable design process. Once done it is seen as a straight forward process to designing good filters.

The Alexanderson Association Grimeton SAQ Transmissions July 5th 2020

SAQ Transmissions

Grimeton Radio Station was listed as a world heritage site because, its original equipment is an exceptionally well-preserved example of the development of communication technology. In fact, it is still in working order.

Grimeton (call sign SAQ) still transmits a message to the world twice a year: at the Alexanderson Day (Sunday at the end of June/beginning of July) and at Christmas Eve. The frequency is 17,2 kHz CW.

I received fading signals from SAQ on the second transmission of 5th July 2020 which began with a Startup and tuning at 13:30 (11:30 UTC) with a transmission of a message at 14:00 (12:00 UTC)

I used a homebrew E probe based on a design by DL1DBC  ( http://dl1dbc.net/SAQ/homebrew_e-field.html#_Testing_the_Homebrew ).

I built a VLF upconverter using a 33.4MHz TXCO - SBL-1 Mixer and a MiniCircuits T1-1 RF Transformer as the Galvanic Isolator which passed through a stage of 30dB amplification before the receiving  RTL-SDR V3 using SDR# software.

Sample rate was 2.4MSPS, no AGC of course combined with a large FFT resolution of 262144 which pulled the weak signal out of the noise.

Here in an image of the signal and superimposed is the youtube video showing the operator.


[  RIGHT CLICK IMAGES AND OPEN IN NEW TAB ]

I recommend checking the Alexanderson Association Website for videos of the whole events and lots of information about the 1920's transmitter.

https://alexander.n.se/?lang=en



Here are annotated image of other VLF stations received here in Kallista

Here are pictures of the E probe construction.

Here is a picture of the VLF Upconverter

As I build other probes I will provide further Blog entries

Today - afternoon pass of meteorMN22 

Nice shot of Tasmania. 

Ant               QFH 

RX                RTL-SDR V3 

software      SDR#

Meteor M N2-2 this afternoon. Channel 4 (infrared) reversed.

Bushfires show up well (bright white spots) along the East coast of Australia.

Select image to get original size.

Here is my first image from the Korean GK-2A (GEO-KOMPSAT-2A)

Setting up a receive system for the Korean GK-2A (GEO-KOMPSAT-2A) geo stationary weather satellite

I have a dish being used as a garden which is 1.2m diameter

Quorum Communications supplied me with a downconverter for 1691MHz in 1994 for review and testing, I will dig it out of the junk box and see if I can get it going using the Linux software supplied by sam210723


Here is the garden dish...…...

Here is the refurbished dish...……..

Feed constructed from coffee tin, PCB interface plate to which the can is soldered and PVC pipe fittings for feed support

The finished dish installed and aligned to the satellite...…………...

The downconverter outputs 1691MHz at 137.5MHz

GK-2A LRIT Frequency:1.69214 GHz

Thus the output frequency should be 138.64MHz

The following image shows the signal being received at 138.650MHz. The downconverter has a noise figure of 0.97dB and a gain of 24.5dB

The SNR of the signal in SDR# is around 16.5dB

Now to set up the Linux decoding software.

To be continued...………………………….

Meteor M N2-2 today at 05.38 UTC

Using QFH antenna - no pre-amp
RTL-SDR V3
SDR#
LRPT Decoder V.54
Auto decode using TCP connection

Incredible results

Best image I have received