Homebrew 455khz Ceramic IF Filter

So today after reading some more of SolderSmoke – Global Adventures in Wireless Electronics” and feeling somewhat enthusiastic about advancing some projects i went out to my work bench and soldered up a test rig used to determine the motional parameters of crystals and in my case, ceramics.


Circuit went together rather uneventfully, built in what i call my Ugly Manhattan style more slum lord millionaire than trump towers LOL. But it works and it does the job it was designed to do. And thats all that matters. Schematic can be seen in the above picture, but there is a better one to follow.


So i plugged in the first ceramic and took a measurement with the switch open and as expected, the 455khz resonator gave a nice readout of 455.68khz. About what you would expect.


Next we hit the switch, or in my case, pushed the bit of solder wire onto the groundplane with a pen and got the second reading. 448.12khz, now with these 2 frequencies in hand i could plug them into some software and get the values for Lm and Fo which are needed to design the filter.


With the two values i plugged them into Ladder Crystal Filter Design and got the values for Lm = 128.6mh and Fo = 447.2085khz which i then plugged into Dishal to design the filter as i like its interface that little bit better.


Immediately i saw a huge issue, the centre freq is all wrong, its way to low, if this was going into a homebrew rig, that would not be so much of an issue, just change the frequency of the LO. But, dropping this into my ICYAWOOD is not going to work, the IF there is right on 455khz, and so that is also where i need my centre frequency. Not sure what to do at this point. I do have some resonators that are closer to 458khz with their resonate frequency, they might have a better chance, but im thinking the low Q of these ceramics is the real issue and even if i come up a few khz, that is still not going to get me closer to 455khz. Back to the drawing board for now.


2015 DX Statistics

We all like a bit of statistics in amateur radio. QSO’s, stations worked, countries worked, regions, zones, continents, prefixes, you name it, someone is keeping a list for it. For me, the CQ DX Marathon kind of covers the major things i want to know about myself and my station and what its achieved in the calender year.

New OpenDocument Drawing

Thankfully my logbook keeps track of all manner of information for me, I know i have worked 77 DXCC entities this year, yes, 77, the log has an error in it and i need to fix a VK0 into a VK1 and 25 zones. I have a fairly even split between Phone and CW and most of my contacts come from the 15m band, with a bit of 40m and 10m thrown in for good measure.

In all, this is not a bad result, I have 1 antenna, a 40m coil loaded shortened dipole that is 6m off the ground. Hardly a DX machine antenna, its actually pretty gimped, and I also run low power, 10w as per my licence restrictions. When you take those factors into account, and living in VK, this is a very good result indeed.


As for grand totals, my logs says I have worked a total of 96 DXCC, this should be 98, there are 2 log entries missing from there, but i cannot work out which ones. Of these I have 58 confirmed on Log Book of the World, a few others on eQSL and one or 2 more on ClubLog. Good enough for WIA F-Call of the year, perhaps. But i do not do this for the recognition, i do it to have some data to compare against. Last year in the CQ DX Marathon i worked 57 countries and 22 zones, in 4 months, Sept-Dec. So i have improved on that score, but used the whole 12 months. All in all, stats can be fun, can be interesting, and putting in a log for CQ DQ Marathon  is a good way of keeping track of how you do. If only more people would send in their logs.


Determining Motional Parameters of Ceramic and Crystal Resonators

Ok, let me just say I am no expert on these things. But this has been my experience thus far. What got me on this bandwagon was the need for a CW filter for my IC-718. As you most likely already know this is a low spec, low cost radio and a narrow CW filter for this radio costs 50% of the value of the radio itself. In local money, that is close to $300 for the crystal filter from Icom, or you can buy Collins mechanical filters for about $230. Hardly cheap at all, and for a radio I intend to upgrade very soon, it would be a waste of money. So i decided to attempt to home brew a CW filter in the mean time. I mean what is the worst thing that can happen here, the filter is crap and i wasted 10 bucks and learned a whole lot in the mean time.

So before i started on this, i needed to know some parameters the filter uses in the radio. The IC-718 uses a 2nd IF of 455khz, with an input and output impedance of 1500 ohm. Ok, so we know the basics, now i went and scoured the net for as much information as i could find about how to go about homebrewing a filter.

Next on my list of things to do, what kind of filter am i going to be building. Ceramic ladder is very common in home brew rigs, especially in the 9mhz and 10mhz ranges where crystals are plentiful and cheap. But my radio uses an IF of 455khz, and while you can get crystals for this frequency, they cost $30 US each way too expensive for a home brew experiment, but, 455khz ceramic resonators are plentiful and cheap but are low Q and Q is everything when it comes to filter building. A typical crystal might have a Q of 10,000 to 100,000, where as a ceramic might be lucky to have a Q of 3000. So, things were looking iffy at this stage.

So i started to play with some computer simulation and filter design programs, Ladder Filter Programm “Dishal” and Ladder Crystal Filter Design (easy to find on google) and I start plugging in various bits of information to see if it is even possible to make a low Q filter from ceramics and it would seem that making something usable might actually happen, in theory at least.

With a pile of cheep Chinese ceramics in hand i began the task of finding 8 closely matched resonators. Now these things have an accuracy of + or -10% ir something nuts like that. First job was to find the resonate freq of each resonator. Out of 100 i did not even get a group of 8 that were within 100hz of each other. But you can see in the picture above that i did manage a couple of groups of 7 and a few of 4 and 5. Well, enough to make a couple of groups of 8 anyway. Why 8? well i intend to make an 8th order filter, the higher the order the steeper the sides of the filter is and the greater the attenuation.

Now to design a filter you need to know some of the motional parameters of the crystals or ceramics, namely you need Fs series frequency and Ls series capacitance. There are a number of ways to do this. I started with the above schematic as described HERE  and found no matter what i did, i could not get a good enough reading using my scope and signal generator to make any usable data.


Which leads me to where i am currently at with this project. The above schematic is an oscillator and buffer amp, and by using the values of capacitance in the circuit, you can determine the upper and lower frequency of the resonator and with those numbers can determine Fs and Ls for the resonators under test. I have not built this rig just yet, but plan to over the next few days to get to it, then measure my group of 8 resonators and average the Fs and Cs values and then use the averaged value to design my filter. I will update my blog once we have gotten to that stage. Anyway, failure is always an option 🙂



While activating for the WWFF/VKFF programs are my main focus and enjoyment, I do also hunt other activators as time and conditions allow for. This is the 4th award in this series of global WWFF awards, being for working 10, 44, 88 and 132 unique references. These contacts were made to all mainland VK areas, 1,2,3,4,5,6 and 8 as well as Japan, France and Germany. Being an F call I have limited DX opportunities as most of the DX work 20m and I do not have access to this band. But I do OK and take the opportunities as i get them.

My goals for the new year are to get the next award in this series, 176 and to go back through some of my less successful activations and get them to quota and earn the Activator 22 award and VKFF 30 activator and put in a big effort for the John Moyle Fields Day and the 2016 Mass Activation Weekend. The wife and I are already itching to go camping, and over the next few days we will start to plan out out first trip of the year. Bring on the WWFF, bring on the activations, bring on the DX and bring on the camping trips.

H 132 2015 VK4FFAB 250


“5 By 7” CW Transceiver: The Receiver

“5 By 7” CW transceiver you might ask. Well i have been kicking about a name for this project for some time. For a long time it has been called Robs Improved Pixie as it pretty much started with a collpits oscillator and had grown from there. A lot of what is going into this rig has been built up Manhattan style on scraps of single sided FR4 and found to have done the job nicely.

So what started out a couple of weeks ago as learn some CAD software to design the CW Audio Filter, has grown into a full blown all out assault on laying out circuits using Sprint Layout, and etching the boards and learning from all the little mistakes you make along the way. All going well I will etch these boards tomorrow and use up the last of the toner transfer paper I have, a sum total of 2 sheets. I have more on order from China but they will not arrive till mid January, as there is no way Im paying $25 for 10 sheets from Jaycar, when i can get 50 sheets for $10 off Ebay.

Oh and the name, “5 by 7” while it sounds like a signal report, its really about the size of each board. Part of the design of this has been that each stage needs to fit on a 5cm x 7cm piece of single sided FR4 PCB. I just want to also mention that none of this is really revolutionary, it has all been taken from various sources on the internet and Experimental Methods in RF Design and other books. All i have done is pick and choose bits that are easy to build and work well enough and are easy enough to build as a stand alone module.


First stage is the Bandpass filter and RF preamp, there is a switch missing off the 180 ohm resistor which allows for muting the preamp for large signals. This circuit was taken out of the DC Receiver section of EMRFD.


Now i know this is 2 5 by 7 boards in one, it was 2, but it was jumper madness and I might yet make this a double sided board and send off to have 10 of these professionally done and sell off the extras. I should also mention that board edge SMA connectors and coax are used throughout to join most of the stages together, in other places 2 pin jumpers are used.

The Arduino Nano is configured as a DDS VFO, with LCD screen and controls the AD9850, the sine wave is boosted by a buffer amp to an amplitude suitable for the ADE-1 Mini Circuits mixer, mixed with the RF from the Bandpass RF Amp, giving audio out. This circuit is a common one found on the internet. The code came from AD7C originally, but i have added a number of features to it to suit my own needs.


Audio from the mixer is run through an audio diplexer to narrow up the bandpass and improve selectivity. It is also amplified somewhat before being sent to the filters. From here things get uncertain, as im not sure how much gain the audio filter will take without causing distortion. This circuit was taken from EMRFD.


Audio from the diplexer then comes to the OP Amp Audio filter. A quad OP amp is used to cascade 4, 2 pole filters to make an 8th order Butterworth CW filter 200hz wide. Centre frequency is 600hz, where i like to listen to code and the bandstop is -35db down at 600hz.


Finally, narrow filtered audio is pre-amped and finally amplified in the last stage and should drive a speaker with very loud audio. Both circuits here are very common ones that can be found all over the internet.

There are a couple of things missing, virtual ground to power the negative rails of the OP Amps, TX RX switching and LO out to the TX. A couple of small things i will need to look into and solve. Maybe by the end of January next year i will have this all built as well as the transmitter, to go with the home brew antenna tuner i have built already. Home brewing, is such fun.


Etching Your Own Boards

Etching your own boards is not all that hard. Once you have created the circuit and laid it out, its just a matter of printing it in mirror onto some toner transfer paper and then ironing it onto a nice clean and sanded sheet of FR4. Sanding with 400 grit is essential, the toner needs a rough surface to cling onto. From here it is just a matter of putting the board in etching solution and waiting till the exposed copper is gone.


Board ready for etching solution.



Board etched with a coat of lacquer.


CW Audio Filter

When it comes to direct conversion receivers, to get good selectivity you need to use good audio filters. There is no IF so you cannot use narrow crystal or collins type mechanical filters to narrow up the bandpass and filter out all those close by signals. For CW use this is important, because you might have 20 different signals in the typical 3K bandpass of most receivers, and listening to those sorts of crouded conditions, sounds like machine gun fire.

So for my direct conversion CW rig, I had decided to experiment with a number of different filter types and find the one that i find works the best. The first one to roll off the production line is an active filter using OP Amps.



Analog devices have a really nice online Filter Wizard that takes all the drudgery and though out of designing filters. As you can see in the above image, the filter is designed with a 600hz centre frequency, this is where I prefer to listen to CW and find it a pleasant listening experience. The filter width is ultra narrow. being only 200hz wide, so 100hz either side of the centrer frequency i should hear a fair bit of attenuation. This is a contest type filter, for very crowded bands. I will also build a 400hz filter for more casual listening or band searching and a 2K SSB filter for SWL’ing.


The schematic above has been optimized for low noise, one of the neat outcomes of doing this is all the capacitors become 100nf. And it becomes so much easier to buy a bulk of 100nf 5% caps than have to make sure you have a number of different values on hand. All resistors are 1% types of which some values might need to be rounded off. I am not sure yet on weather rounding up or down makes a big difference, but this is something i will have to determine experimentally at some point.


Something i did learn is that OP Amps need to have both positive and negative supply voltages for them to work right. After showing my work to VK2DDS , who pointed out I had things all wrong with power, I was sent on my way to learn about OP Amps and virtual ground. Now there are a number of ways to do this correctly, supply the OP Amp with negative volts, or use a Virtual Ground. By using another general purpose OP Amp  configured as above feeding the negative supply line of the OP Amp with 1/2 the voltage of the positive supply line, you and up with the situation of the voltage being able to swing 12v <-> 6v <-> 0v. This is the same as 6v <-> 0v <-> -6v because the voltage of ground is relative, it can be whatever you want it to be.



So now the final piece of the puzzle is the design a board, etch it and test it out to see how the thing actually performs. For this I am using a TL074 Quad OP Amp. The design work was performed initially in KiCad, a Linux circuit board design program. It has a funky way of working and one that I am not all that fond of, but, it is also extremely powerful and is software i will persist with as it will be handy when i design something a little more complex than this filter.

Now, that’s not to say i did it all in KiCad, in the end i swapped out to a windows program called Sprint Layout. This is more a drop and join visual sort of method i am more comfortable with and for small jobs very easy for complete beginners to use. So the board layout is all done and all that is left now is to etch the board, I use toner transfer for this and then build and test the filter and see if it worked up to expectations and design parameters.


New QSL Cards

I have not been doing much DX of late, or radio in general for that matter, but still managed a few new cards from the VKFF Mass Activation Weekend. A nice one in thas was a park to park with VK8GM. VK8 is almost rare DX within Australia, not a lot of operators there and a very sparse population, makes it hard for contacts, but it was nice to get another VK8 park in the log.

Cliff VK2NP was in Sea Acres for another park to park contact, it was his first park activation and from all reports he had such a ball that he will be doing more activations in the new year. Anyway, the VKFF Mass Activation Weekend was a huge success and one i thoroughly enjoyed participating in. Thanks to Paul Simmons vk5PAS for organising the event. I cannot wait till the next one, should be great.

_cfimg-1558266613333855599 _cfimg1775802686455966664 _cfimg8253643352014594548


50A Anderson Plugs

Doing a few little jobs today while listening to the ARRL 10m contest on the radio. Now that i have solar panels i really need a way to charge the battery while out portable, so I knocked up this Anderson Plug and bolted it to the battery. I will be able to radio and charge the battery at the same time which will be nice.


Anderson plug after being soldered and crimped. YES BOTH


And now bolted firm to the battery.