A homebrew Ezitune

Today I had some free time and I decided to build something.But what would it be that could fit in my 2 hours break?An idea came at once!I always wanted to have a tuning accessory like the famous S.E.M. Ezitune, why not build one?

The Ezitune device is nothing more than a noise bridge which can be permanently connected between the transceiver and antenna tuner. There is a switch to bypass it during normal transceiver operation and switch it in for antenna tuner adjustment.I have seen them on ebay from time to time (they are really rare) at prices around 50 GPB and one German distributor (WIMO) sells a newer version of it by the same name.However, I find them too expensive for what they do.

The main concept of a noise bridge is to measure an unknown impedance using a receiver as a null indicator. The unknown load is connected to the one port of the bridge and the receiver on the other. The bridge has two controls , one for resistance and one for reactance (Xc or Xl) and generates a wideband noise at a level of around S9. In order to find the unknown impedance at a specific frequency we tune the receiver to that frequency and using the bridge’s controls we try to achieve a null at the receiver (that is, the lowest indication on the S-meter). When the null is achieved, the unknown impedance value is equal to the readings of the bridge controls (resistance + reactance).This is based on the bridge’s property that “a null is achieved when the known impedance equals the unknown”.

The idea behind Ezitune (as I understand it at least… I never managed to find its schematic) is to use this feature to make its magic. Instead of  providing a measurement range using a variable resistance/reactance it sets the “known” impedance to pure 50 Ohm (no reactance) which means that when the null is achieved the “unknown” impedance will be pure 50 Ohm (or very very close to that).Isn’t that exactly what  we are trying to achieve when we adjust our antenna tuner? Trying to make our whatever-we-have-as-antenna to look like a 50 Ohm load to the transceiver.Ezitune makes it extremely easy.Let’s build one!

All we need is a simple noise bridge, with components which are already lying around in some bin.I have one in mind from previous google searches: the New England QRP club noise bridge (original article here and the construction details can be found here ) . (Note: The bridge is not my design and I do not pretend it to be, it is a Dave Benson’s design.I am posting the schematic here just in case the original article is removed)

New England QRP Club noise bridge

New England QRP Club noise bridge

In order to convert the above bridge to an Ezitune device, all we have to do is to remove R1 and C1 and replace them with a 51 Ohm resistor.During my initial tests I also omitted  C6 and the bridge worked fine. The reason I omitted this capacitor is because I think that it will introduce unnecessary reactance to the antenna port. Of course, it may help to prevent any DC voltage that (may) appear at the antenna port to reach our receiver.

The finished device is shown below

The completed homebrew Ezitune

The completed homebrew Ezitune

On the top right, just on the left of the toroid core, you can see the “known” impedance of 52 Ohm (a 30 Ohm + 22 Ohm resistor in series, as I ran out of 51 Ohm ones). The transformer is made of 4 turns of the flat ribbon cable on an FT50-43 core.The right coaxial cable, leaving out of the green ribbon is connected to the receiver, while the coaxial leaving out of the yellow ribbon is connected to the antenna tuner.

The next step is to see whether it is working as intended and whether is it really easy to adjust the tuner with this. My antenna is a 15m long doublet fed with 450 Ohm line and I can get a pretty good match from 40m – 10m using the MFJ-901B versa tuner. I cannot get 1:1 SWR but it is always less than 1.7:1 . On some bands (30m and 12m) the best I can get is somewhat 2.5:1 SWR but at the moment I am not really active on those bands.

For my first attempt I chose 20m where the tuning is very sharp and every time I change bands it needs very delicate touches to adjust it.I played a bit with the controls into positions I know for sure do not work for 20m (I know by heart the rough settings for 20m). The noise was a constant S9 on the receiver. As the controls were approaching their “good position” I noticed a slight dip which suddenly became deeper for a short time and then back to S9. I went back a bit and played with the other control. The null was getting deeper. I adjusted the controls alternatively and bang!The S-meter was now at S2!Wow!The adjustment of the tuner was really fast! The original noise of S9 was reduced to S2 when the tuner controls were where they should be. The interesting part is that using the bridge is easier to see and hear the dips and adjust the tuner for better match.

My method of tuning the antenna is to use my Autek RF-1 antenna analyst to find the tuning settings per band and note them on my notebook. When I change bands I preset them, connect the RF-1 and adjust the tuner knobs for the lowest SWR. On 20m the best I can get is 1.4 – 1.5. This does not have to do with the tuner/antenna but the RF-1 takes 1 or 2 measurements per seconds and it is easy to miss the optimal point.Otherwise you have to go really slow (this means time!). I am happy when I get 1.5- 1.6 and I usually stop there.

After I adjusted the tuner with the bridge I unplugged it and connected the RF-1 to see the SWR. It was 1.3 !Much better and much faster.

I tried again on 40m and 17m. There I noticed something interesting. Even though I could not achieve a perfect null (the S-meter indicated S5 during null) the result was good as it was 1.5:1. So, that means that the lowest the S-meter indication the better the match (the lowest the SWR) but even if it is not S1, the match is probably good. I will have to test if the same SWR value results on the same (or almost the same) S-meter indication across bands.

What is left is to put it in a box and add a bypass switch so that I can have it permanently connected.

Not bad for a 2 hours project, right?

73,Nick, SV1DJG

Dual Band QRP transceiver – Local oscillator

The last 2 weeks I am playing around with the local oscillator for my transceiver project. I would like to cover both 40m and 20m – if possible – and get better than average performance. This does not mean I do not care at all for the oscillator part, but would like to keep complexity low for the moment and get by with some compromises.

My initial desire was a DDS with a clean-up PLL loop in order to remove the DDS spurs and get a very clean LO signal.There are a lot of similar designs around and I also have one in my mind. However I have never designed a PLL from scratch and I judge that this could wait for a while. I decided to use directly a DDS followed by an amplifier in order to drive the diode ring mixer.I am aware that this is not the best solution due to the spurs generated by the DDS, especially the ones close to the generated frequency, but at the moment I want to try this solution and maybe in the future I use a PLL.

The DDS module I am using is one I got from eBay for 4.5 USD. It has an AD9851 DDS and claims to generate frequencies up to 40Mhz.It comes with a 30Mhz crystal clock and with the aid of the AD9851’s internal x6 multiplier clocks the DDS at 180Mhz. I have also a similar module based on AD9850 clocked at 125Mhz which I will may try for performance comparison.

AD9851 DDS module

Low cost AD9851 DDS module

I have read rumors on the Internet that all those DDS boards are sold cheap because of a factory error (they put on the wrong output filter ? ). I am not sure about this, but I can assure that they do not have any flaws or damaged DDS chip.

For the moment I am able to control the frequency using an Arduino board connected to my PC and sending commands through HyperTerminal.Later I will use a stand alone atmega328 to do that.I will provide more details on the connection and programming of the DDS module on a later post.

Programming the module is straight forward.The sellers on eBay provide sample code which can be modified easily. One major modification I did is to perform all the calculation using integer arithmetic because the Arduino’s float and double data types are of very low precision and I was getting a large frequency error.After my modifications I get exactly the frequency I program +/- the permanent error due to the board crystal oscillator (around 60 Hz).

In order to cover the 40m band (7.053 – 7.200) the LO required is 4.006-3.859 with reverse tuning, and for the 20m band (14.100-14.347) the LO required is 25.159 -25.406. Using this scheme, there is a sideband inversion for 20m and no sideband inversion for 40m. this means that the same IF filter will work for both bands. That’s very convenient!

The Eden9 uses a diode ring mixer so it requires at least +7dbm drive to keep insertion loss at normal levels.The DDS alone cannot drive a DBM so I built a buffer/amp to increase the drive level for the mixer. The initial buffer design was taken from here and I spent some time on it because I had some problems matching it to the DDS board filter and getting sufficient output on 50 Ohm load for both bands. After discussing it a bit on the EMRFD Yahoo!Group [Ref 1]  and following some advice from the members I modified the circuit but still was not up to the task. Finally, I followed the advice of  Jerry, W5JH, and built a broadband W7ZOI amplifier which I borrowed from my EMRFD book page 2.20 , Fig 2.57, the schematic is shown below:

w7zoi amplifier

w7zoi wideband amplifier schematic

The only addition to the original amp is the step down transformer at the input, which provides a match of  the filter’s 200 Ohm to the amplifier’s 50 Ohm . Thanks Jerry for the idea!

DDS amplifier during testing

DDS amplifier setup during testing

Using this amp I managed to get clean signal of 1.5V Peak on the 4Mhz range and around 1V Peak on 25 Mhz range on a pure resistive 50 Ohm which seems adequate for my needs. I left some extra space on the board in case it is required to add any filtering before the amp if  there is a problem with harmonics  from the DDS.

In reality, connecting the amplifier on the DBM , I noticed a slight degradation of the signal quality on 25Mhz. I am not sure, but I think that my homemade DBM may not be a good match in that frequency. (I will take some measurements to verify it, soon).

Ignoring that for the moment, I connected the newly constructed LO to the EDEN9 IF module and tuned around (still without any front-end filters) and the stations were coming very loud on both bands.This is a good indication that the LO drive is good and despite the signal impurity it seems to work OK.

I will know for sure after I add the front-end filter whether the oscillator is clean enough, as I will be able to test it on all the suspicious ranges and see if the harmonics and spurs are a problem and if yes, how serious.

So, the next block is the front-end filters, which I am going to work on this week.



1)   DDS module with buffer / amp message thread ( http://groups.yahoo.com/group/emrfd/message/8004  ) (messages are visible only to members of the group)

Eden9 IF subsystem – Core of the dual band QRP SSB transceiver

The Eden SSB IF subsystem was designed by Ron Taylor, G4GXO, as part of the Eden Valley Radio Society’s “Eden  70Mhz  SSB transceiver”. It is a very compact IF subsystem (IF at 11.059Mhz) and it only requires a LO and front end to become a complete transceiver. There is a very active Yahoo Group  for that project, where you can find the schematics of the subsytem. Ron, has since updated the design to use a single NE602 for both TX and RX which makes it easier to adjust.

Since it was presented on SPRAT #144 I wanted to build a 70Mhz transceiver since I do not have any equipment for that band.  The Sporadic-E seasons come and go and I still have not started it! Instead, I thought I could use it is the core of a small portable HF QRP transceiver.

As I will write later, I have already built a portable 20m QRP SSB transceiver designed by N6QW, the “Pocket Sized 20m QRP SSB” presented in QRP Quarterly, Summer 2011. The version I built puts out only 0.5W and I never -as of today – built the 10W amp. I have used this transceiver for QRPp QSOs with little success, but it will stay as-is as the testbed for true QRPp SSB QSOs.

To complement that, I would like to build an equally portable but dual band SSB only transceiver for not so challenging QSOs.I chose the 20m/40m combination. If this proves to be unrealizable, the 40m band will be dropped.

So, for the first part, I decided to complete the receiving line.Starting with the Eden9 subsystem as the core, I decided to add a LO and the input filters and see what I get. Then I will complete the final PA and low pass filters.

I have completed the Eden9 subsystem in a small prototype board.

Eden9 subsystem components side

Eden9 subsystem components side

Eden9 subsystem bottom side

Eden9 subsystem bottom side

As you can see, the components are really squeezed, but works very good.As part of the build, I decided to not use any shielding in order to evaluate the possible degradation in performance. When the photos were taken the transmit part was not complete yet, it is the blank area on the upper right.

Without the input filters, I connected my signal generator and recorded some activity on 20m.

You can hear some samples here.Please do not crank up the volume, because the sound is too loud at some parts.

(the noise comes from my not-so-optimal antenna setup)

http://soundcloud.com/sv1djg/ha2rp-ssb-cq-128 http://soundcloud.com/sv1djg/it9vpt-ssb-cq-20m http://soundcloud.com/sv1djg/cw-pile-up-on-20m

My next step is to add a DDS-based LO in order to use a uProcessor for controlling it. There are a lot of cheap DDS modules on eBay based on AD9850 and AD9851 that sells for as low as 4$. I have  a lot of them in various forms with both AD9851 and AD9850 that I will use on this transceiver.

I am now in the phase of building and debugging a DDS-based local oscillator. I will post updates on that soon.


FtTuthill80 Transceiver – Kit arrived

Today Mr.Postman had a nice gift for me! My FtTuthill80 transceiver kit just arrived! I had been waiting for a month in order this batch of kits to be dispatched.

FTTuthill80 kit arrived

FtTuthill 80 transceiver kit just arrived

This transceiver, designed by Dan Tayloe (N7VE) and kitted by the Arizona ScQRPions QRP Club, is a quite different and interesting design. I bought it not because I lack any equipment on 80m, but because Dan Tayloe’s designs are very different than what most of us are used to  and therefore (the least I can say) extremely educative. I think that Dan has created a unique “school” of receiver/transceiver design and his results are becoming classics ( I was not quite fast to grab an NC2030 kit some days before …)

More info about this transceiver can be found on the Arizona ScQRPions QRP Club page and also at the the dedicated Yahoo! Group

I now have to inventorise the kit and put it on my schedule for this weekend.

73, SV3DJG

SW40+ QRP transceiver – Ready for action!

Finally, the SW40+ project is over. I did not updated the status for a long time, although a lot of things happened!

After completing the assembly of the kit (toroids and final transistor) I was ready to make all the adjustments. In the mean time, the enclosure had arrived and I chose to install it first and then adjust it. The enclosure kit comes with a BNC antenna connector and at the time, all my other equipment uses PL-259  so  I bought some adapters to use until I make a final cable with a BNC plug.

After connecting my main antenna to the SW40 through the adapter I started adjusting the receiver. It sounded a bit deaf, meaning, that there was not quite big difference whether the antenna was connected or not. This puzzled me a lot but I decided to continue with the transmitter part.(This proved later to be a big mistake!).

I connected the transceiver through my power meter to a mini dummy load and  started keying the transmitter. There was no output at all! No matter what I did, the transmitter was dead.

It took me about 3 hours of thinking and signal tracing to make the discovery. The BNC to PL-259 adapter was faulty…actually, it was shorted! All the transmitter testing was performed with a shorted output! This resulted in a fried output transistor, and I had to order new ones from Small Wonder Labs!

After replacing the output transistor (and disposed the faulty adapter!) all the adjustments were finished within minutes. The transmitter is capable of delivering something more than 2.5 watts  (as displayed on my power meter, with unknown accuracy !) but I set it to about 1.5 Watts for safety.


The SW40+ finished

I have used it a bit, mainly on receiving (yes, my CW skills are still bad…but they are improving fast!) and it is quite a pleasure! The receiver is sensitive enough and can hear almost everything my main rig hears (which means that they are both either very good or very bad!).

There are two thing that someone has to get used to. First, is the tuning pot. The tuning range of the VFO is small but for a single turn pot, it stil feels that it tunes too fast! Also, it is a bit non-linear, because, there is a small range at the beginning and a bit smaller at the end that does not actually changes the VFO frequency. It is not very serious though and there are a lot of mods to fix that as well as the tuning range.

The other thing is the lack of AGC , which for someone that has never used an AGC OFF switch on his transceiver, is a bit strange! On the SW40+, most of the time when tuning and the GAIN is on the max setting, very strong signals appear from nowhere  at  very “annoying” levels !!! It takes a little practice to learn to tune with both the TUNE and the GAIN control.


Another view of the SW40+

However, none of the above  issues can degrade the quality of the transceiver and are not at all annoying. It is  just something that take about 2 hours to get used to! The on-the-air performance is very good and it makes a very compact and portable transceiver that I am going to enjoy for a long time!

I hope to meet you on the band, soon!

73, SV3DJG

Portable QSO Recorder from an old MP3 Player

From time to time I need urgently to record a QSO or some activity from my transceiver. I have not hooked up my computer permanently to my rigs and when that time comes, I usually run to connect the cables, start the recording software, adjust the levels and so on.  I always wanted a better solution, faster and more versatile.

Last year my MP3 player had a small accident…”someone” step on it and the screen’s protective window was broken, but other than that, it is working perfectly. Due to the ugly appearance, it was retired and stored away. It is a very cheap MP3 Player,the one that each manufacturer buys by 1000’s from China and prints his logo on it. Mine was branded as “SONY”! It runs on a single AAA battery, has 1GB of memory, FM radio, but the most important feature is its ability to record through an internal microphone.

I immediately thought that it would be nice if I could use it to record directly there whatever I want and then transfer it to my PC through the USB port. All I had to do is to replace the microphone with a cable going directly to the source, and adjust the input level accordingly.

Without any guilt, I disassembled the unit and headed for the microphone.

MP3 player broken into parts

MP3 player disassembled

It was really a very small one with very small pads, but it was removed easily and the cable was connected in place of the microphone.

MP3 player microphone replaced

MP3 player microphone replaced with direct cable input

Unfortunately there is no space inside the player to put a level adjustment circuit so it will be either external or the volume of the source will be adjusted  accordingly.

QSO recorder ready

The QSO Recorder is ready.

And  voila! A portable QSO recorder is born! I tested and it works fine, but the volume level must be kept at very low levels for preventing audio distortion. This will be fixed with soon an external circuit.

Using the MP3 player’s “High Quality” recording setting  (which I think is 32Kbps, WAV format) the device can record up to 32 hours of audio! Not bad at all !

73, SV3DJG