Lack of Activity

Been getting the itch to experiment with microwaves again.  Been going through my equipment for the upcoming season, and will hopefully have some new stuff to post in the upcoming months.

Useful Microwave Test Gear

About 6 months ago, I ordered a VFO kit using an Si570 DDS chip.  The unit is self contained with a 2-line display and an optical encoder switch to change frequency down to 1 Hz resolution.  If I recall the specs on the Si570, I believe it is accurate to 25ppm.  It puts out 10 dBm from 3.5 - about 260 MHz.  There are other models which go all the way up to 1400 MHz.  I got the lowest cost unit that goes to 260 MHz, which is adequate for its intended purpose.  The units are available on http://www.sdr-kits.net/

I bought it so that I can use it as a reference oscillator for various brick oscillators.  I have 2 bricks in my possession, one on L band, the other on 6200 or so MHz.

Here is a test setup showing the unit in operation, referencing my L band brick:

Software Defined Radio - What Is It

Over the past year or so, I have made mention of the SDR (Software Defined Radio) in some facebook posts.  I get questions from time to time, so I thought I would write a little bit about how this works and what it is, exactly.

I have 2 different SDR's actually.  One is a receiver only, the other is a transmitter / receiver.  Both of my radios cover the HF (shortwave) bands, from 1.6 - 30 MHz.  The radios are kits, yea, you have to build them.  They are a PC board which you have to populate the board with the parts.  You supply the case for it, but the newest versions supplies all of the connectors.

The finished radios - Softrocks have no knobs, no volume control, no tuning dial, nothing.  It is a black box with an antenna connector, a USB port, and 1 or 2 audio connectors, depending on if your board is receive only or is a transceiver and a power connector.

The USB port is used to tune the radio.  The software which you use with this board sends tuning code to the board via the USB.  It also  automatically selects the band filters for the frequency youre dialed to.

The audio connector on the board sends a stereo audio frequency signal to the sound card.  This signal is audio, but there is a slight phase difference between channels that is derived by the mixer on the board.  Actually, there are 2 mixers on the board, with a common local oscillator , however, one mixer is fed a phase delayed signal from the local oscillator (90 degrees).  The outputs of the 2 mixers therefore has a slight phase shift in the mixed down signals.

In transmitting, the transmitter basically reverses the above process, amplified and transmitted to about 1 watt, which I feed into a small linear amplifier outboard.

Because of this special stereo audio, the DSP chip in your sound card and the associated software is able to distinguish between upper sideband, lower sideband, AM, narrow band FM, double sideband (with or without carrier), DRM...anything you might want to receive.

The software (there are several different programs available) displays the entire spectrum that the soundcard sample rate allows.  For instance, if your sound card is capable of a 96 Kbps sample rate, you are able to see 96KHz of spectrum at one time on your display.  Lets say you are looking at the 49 meter shortwave band at night, you will see a spectrum analyzer display that is just under 100 KHz wide, and you will see every station within that bandwidth on the display.  The software is calibrated in dBm, and you can see which signals are stronger or weaker and what frequency theyre on.  A simple mouse click will allow you to select any signal that you can see.  The demodulated output appears at your line output of your sound card.

One of the biggest features of this technology is the filtering.  By "grabbing" and dragging the audio passband window in the software, you can continuously vary the bandpass and center frequency of the filter so you can eliminate annoying heterodynes and other interference from the sighal you want to actually listen to.  There are preset filters, which vary from 2.4 KHz to 16 KHz in the AM mode (there are other filters for SSB).  One can select one of these filters and modify them any way you want to enhance the intelligibility if the signal you want to hear.

Synchronous detection on AM is available in some programs that allow you to listen to shortwave stations in the presents of "selective fading"  Selective fading is a phenomenon where the carrier fades, but the sidebands arent faded.  This causes severe distortion during the fade, but the synchronous detector eliminates that problem by injecting a phase locked BFO to the carrier frequency, simulating the original carrier, and since the BFO does not fade, you will never know there was a selective fade going on (unless you look at the display).

All of these features are in the software.  The hardware part is just a mixer and oscillator that produces an audio baseband signal for which the software can do its work.  The hardware end of it is extremely simple.  There are receivers available for as little as $20.  The all band  receivers are about $50 or so.  When these are used with a good quiet sound card, the performance is quite unbelieveable.

I hope this answers a lot of questions as to what this technology is about.  This is a basic, non-technical explanation.  There are more in-depth websites that describe the technology in far greater detail.

2304 Yagi

Being I got some power out of the transmitter, I had an idea a couple of days ago, and thought about building a yagi antenna for 2304.

I went with the DL6WU design, but with a twist.  Instead of doing the folded dipole and balun, I went instead with the WA5VJB unbalanced half - folded driven element.  Everything else about the yagi is in accordance to DL6WU's calculations.

I have 10 elements built so far.  The longest element is about 60 mm long, and the boom length is 12 inches (305 mm).  The entire antenna is made from hobby brass, the boom is 1/4 inch square brass.  The elements are 1/16 inch diameter brass rod.

Here is a photo of the driven element.  The driven element is insulated from the boom. At first I could not get the SWR down on the driven element.  Some adjusting of the driven element length, shape and spacing to the first director, I got the return loss down to -30 dB or more.  I cannot measure below that level of reflected.

As you can see, the antenna is quite small.  The antenna appears to have decent directivity.  Although I have not yet completely tested it with a weak signal source, placing my hand in front of the antenna causes wild fluctuations in the reflected power, but moving my hand around the sides and to the rear of the antenna shows very little change in reflected power.  This kind of tells me that the RF is being radiated off the front of the antenna, which is what we are looking for.

The antenna will be extended to about 2 feet.  I wanted to try to see if I could make this work with 10 elements before I waste the additional brass (and time) with a longer antenna.

I used the VK5DJ Yagi Calculator program to design this antenna.  The driven element was designed by me, using the principle that WA5VJB used on his 1296 & lower antenna.  The length was arrived at empirically, trimming the length and spacing between the DE and 1D.  The odd shape of the hairpin does not seem to affect anything, its just how I happened to bend the DE.

I still need to package up the transverter and actually finish it.  The antenna experiment was a little bit of a distraction to break some of the monotony of the rest of the project, besides, I had some ideas after talking to N6NB, Wayne at the Packrats VHF conference regarding his Quagi design.  Since his designs were done on an antenna range, with a limited number of elements, I posed the question "How do you think your principles would work on a long boom design, such as the DL6WU design - using his director calculations, but replacing the reflector and driven element with your quad elements".  His reply was that it should work.  Although I did not go with the quad elements on this design, I  did think that the WA5VJB driven element should work with this antenna, which apparently it does.

2304 MHz Transverter - Photos

As I promised, here are some photos of the nearly completed 2304 rig.

 This is the complete rig

 One of the modified RF amplifiers with the cover off.

 The other amplifier.

 Mixer and IF Attenuator

 Bandpass filter

 Unused amplifier module.

Modified MMDS unit.

So these are the pictures of how this thing looks right now.  The amplifiers are not bolted down to the aluminum panel yet.  I have yet to install a T/R switch and I need to come up with some kind of power distribution.  This unit takes +20V, -20V and +12V to make it all work.

2304 MHz Transverter - Its almost done!

I know I promised a posting describing the filter and other modules that I'm using in this project, but I thought I would wait until I had gotten some of it actually working.  Since I am now getting a useable amount of power out, I felt it was time to write a little.  Today I built an IF attenuator to drop the 144 MHz IF level from 250 mW down to 1 mW to feed the mixer.

I then started hooking up the modules that I had built up.  The LO was fed through a commercial low level amplifier module that was part of an amplifier chain which I will describe in a couple of paragraphs down this entry.  Measured LO was about +16 dBm.  The mixer wants between +10 and +20 dBm, so this level is perfect.  From there, the output goes into the LO port of the mixer.  The IF port goes to the 24 dB attenuator that I just built, and from there goes to the FT-817.

The RF port is connected to the band pass filter.  This is a 3 pole inline resonator filter that had to be modified slightly to get it to tune 2.4 GHz.  This filter does have a cool feature that is worth talking about.  It has 2 notch filters incorporated in it.  I tuned one notch to the LO frequency, the other to the image.  This should be one clean transverter!  The LO and image look to be down better than 80 dB because of these notch filters.  The tuning of the bandpass filter is quite sharp as well.

From the band pass filter, the RF goes through my 2- mmic preamp I built from Down East.  I'm only seeing about 20 dB of gain, however, so I'm not sure if there is a problem or not.  I'm getting 0 dBm out of the preamplifier.

WA3TTS gave me some surplus 1.85 - 2 GHz hardware, including a filter and an amplifier chain.  The amplifier chain consists of 5 modules.  I'm not certain how much input power is required to drive the entire chain to its rated 5 watts, but my guess is about a milliwatt or 2.  The modules all needed to be retuned to 2.4 GHz.  The modules are built one stage per module, and are built on ceramic circuit board material and have etched tuning and matching circuitry with no tuning adjustments.  The modules that I've tuned so far have been tuned with my Dremel tuning tool - carefully grind away printed capacitors and tuning lines.  This is opposite of most microwave commercial hardware, as most surplus is designed to operate above the amateur bands, such as the MMDS unit that I added material to its internal PC filters to get it to cover the ham band.  Because these units operate below the band, the opposite has to be done - copper has to be removed.  Because the copper is either vacuum deposited or etched on ceramic, an X-Acto knife was pretty much useless.  I then attempted grinding away the traces carefully with a fine grinding bit on the dremel tool.  With the unit powered up and the covers off the modules, I connected the power meter to the output of the first module through the attenuator, and applied drive from the IF.   I then began grinding the input network in the module while watching the power meter.  As the power came up, I continued until I saw no more increase.  I did the same with the output network.  I'm seeing about 10 dB of gain thru each of 2 modules I have done so far.  Currently I'm getting +23 dBm out of the last amplifier.

Well, thats where I am with the project right now.  I still have a 5 watt PA to tune.  All of the modules have to be mounted down to the chassis, and the T/R switching still has to be built.  The receiver is working, and the transmit is working at the 200 mW level, and the next thing I think I'm going to do is install the T/R relay and try to make a short range contact with what I have so far.  The 5 watt PA will be more difficult to tune, as it consists of a pair of devices which will have to be balanced somehow.  I'm not entirely certain how I'll retune it, but once I figure out a gameplan on tackling it, you can bet I'll be posting a procedure right here.  Check back soon for the info!

2304 MHz Transverter - Significant Progress Made #1

Welcome back!  Today was a good day.  Got a LOT done on the transverter with the help of WA3TTS.  There was so much that we did, I can't write it all in one post, or in one night.  I'm going to split this up into several postings.  Question is, where to start.

First thing, and probably a very important thing is we checked the LO level and spectrum.  The level is much lower than I expected, I'm getting about -3dB at the SMA connector on the receiver brick.  The good news is it is VERY clean, all spurs are down by more than 60 dB, and the nearest spur is over 100 MHz away from the intended LO freq.  My plans have changed slightly because of this.  Since I have 2 of the Down East ERA2 2 stage amplifier boards.  I plan on taking the unbuilt board, and cut it in half, making 2 single mmic boards.  I will then use one of these boards to amplify the LO.  These boards should have about 15 dB of gain, so I should get about 12 dBm out of the amplifier, enough to properly drive the mixer.  I'm certain that if i fiddle with the LO probe, I could increase this level, but since the LO is so clean, I am almost tempted not to mess with it.  At least I know how much RF is there, it wont be difficult getting the level up to where it needs to be.

While I was there, I finished the mixer module.  We added a filter (more about this on the next post), and measured -20 dBm at the RF port of the mixer.  We only used the available -3dBm from the LO source.  A signal generator simulated the IF, tuned to 144 MHz.  The mixer output saturated at +6 dBm.  I believe that when I increase the LO level to +10 - +13 dBm, the output level will increase dramatically.

We did not test the 2 stage low power amplifier.  It is ironic to think however, that the -20 dBm coming out of the mixer matches perfectly the 2 stage gain of 30 dB, and in theory, I would end up with the +10 dBm that I'm shooting for at the output of this amplifier.  I believe it would be better, however to drive the LO to the +10  dBm level and run less IF energy into the mixer, thinking the linearity would probably be better coming out of the mixer.

As you can see from what is presented here, the rig is almost ready to hook to an antenna.  All that it needs is a good filter following the mixer, and perhaps a bit more power.  Tomorrow I'll write about a filter that we retuned and some of the really cool features of this filter.  We'll also talk a little bit about power amplification too.  I'll leave the cliffhanger here... Lets just say, I got a solid state power amplifier that will provide meaningful power for portable work.  It will need some work to get it to run on 2304, so will be experimenting with that soon also.

Stay tuned, keep the dish pointed this way, and check this frequency soon for the next update.  73 for now.

2304 MHz Transverter - A Little Bit About the TX Mixer

While I'm waiting for the parts to come to repair the ABPM, I thought I'd write a little something about the mixer that I'll be using with this rig.  The mixer I'm going to use is an old Vari-L DBM-184 flatpack double balanced mixer.  The specs on this device, although it is a rather old device shows some promise according to the datasheet.  It will accept up to a +20 dBm LO, the LO to RF isolation is about 20 dB, and is good to 2500 MHz on the RF port.  The conversion loss is about 8 dB, which is typical for a DBM.  Its 3 dB compression point is +6 dBm, which I would assume is the output power, which corresponds to an IF drive level of +14 dBm when  used with a +20 dBm LO.  I will not be driving this device that hard though, figuring on driving the device with about 0 dBm, and between +10 - +13 dBm on the LO port.

I picked up 2 of these mixers at the 2010 Packrat (Mt. Airy VHF Club) VHF Conference.

I am in the process of packaging the mixer in a brass box, just slightly larger than the mixer itself (3/4 inch square), with SMA connectors connected to all 3 ports.  This will make the unit a "module" which is how I like to build things, especially when working with the microwave bands.  This method of construction allows one to substitute different components, so one can experiment to find the best performing module for this particular combination of parts.

Basically, the planned signal path will be something like this:
From the LO tap in the CalAmp RX converter, LO will be amplified using a MSA-0386 in a homebrew amplifier run at saturation, feeding the DBM-184 mixer.  The IF will be an attenuated 144 MHz TX signal from the FT-817nd, feeding 0 dBm into the IF port of the mixer.  The RF port will feed a 3 resonator BPF designed by G0ORY, details of which appear here.  From the filter, the RF will be amplified by the already constructed ERA-2 mmic amplifier.  From that point, I'm unsure as to what the final power output will be, but I am hoping I will find an amplifier that will provide a couple of watts on the band.  What I use will depend on what the pocketbook will allow, as well as what power supply I have available.

Once again, stay tuned, as work progresses, I will post again.

2304 MHz Transverter - Transmitter Construction

After what seemed like forever, the ABPM (All Band Power Meter) arrived from Down East Microwave.  I had also ordered 2 ERA-2 amplifier boards at the same time.  I had the power meter built that evening, and went to test it, and the microwave detector does not work.  The parts in that section are extremely tiny, and I'm thinking I popped the detector chip.  I think it was probably my fault.  I also believe I'm losing my touch, as I'm having more difficulty working with surface mount parts due to their size.  I think I need to invest in one of those big magnifyers with the light in the head if I plan on continuing to work on this sort of thing.

Anyway, the power meter will read from about -20 dBm to +10 dBm in 2 bands - HF thru 500 MHz, and from 500 MHz to 10 GHz.  The low frequency sensor is working perfectly.  I got this shipment yesterday.

Today, I built one of the ERA2 dual mmic amplifiers.  I am going to use this as a transmit amplifier to get to the 10 milliwatt level.  10 milliwatts is a magic power level, as many higher power amplifiers are designed to input 10 milliwatts to get you from 1 watt to 100 watts, depending on what you are planning.

The ERA2 preamp boards are tiny as well, but I didnt have too much trouble building this one.  The 0805 chip caps are small, but fortunately, in these amplifiers, they're not too close together, and the circuit for a mmic amplifier is very simple.  These boards are nice high quality boards, made of Teflon, and have plated through via's.  These amplifiers provide about 30 dB of low level amplification, and will output up to 20 milliwatts at the 1 dB compression point.

This is a photo of the amplifier board.  As you can see, it is quite small.  Of course, I have no way to test it until I have the mixer finished.  I am currently working on that now, but I'm short one SMA connector to make that happen.

I still need to finish the mixer, and I also need to make a bandpass filter to place after the mixer that will pass 2304 MHz, but block the 2160 LO and 2016 MHz image.  I am going to use a 3-resonator type filter similar to the one I made for my 1296 rig.  Things start to get rather small on 2304 MHz, the resonators are less than an inch long on this band.  Its a good thing I have a micrometer.  I'll have the filter swept before I use it on the air.  If it's not selective enough, I'll use a cavity filter instead (or in addition to).  2304 is an awkward band in that it is more difficult to homebrew components such as the filters.  Cavities for 2304 are about 1 1/2 inches high, which makes them somewhat on the large size if you are trying to build small.  The resonator filters are less than an inch, and they require some level of precision for them to work well.

Once I get the mixer and filter built, I'll post again, and let you know how it all plays together.  Once I get to the 10 milliwatt level, I may attempt to make a contact on the rig if I can find someone with the capability to operate that band without having to drive too far.

Stay tuned, the next update will be coming soon!

New Project - 2304 MHz Transverter - Tuning the Front End

OK, I got the ambition to hook it all up today and start tuning the receiver.  I used a very crude setup.  The only signal source I have that will generate a signal near 2304 MHz is the second harmonic of the local oscillator in my 1296 transverter.  I used that as the signal source.

I had to set up the signal source about 20 feet away from the workbench to keep the RF from being picked up directly on the 13 cm circuit board.  I put a homebrew 1/4 wavelength antenna on the input side of a 0 - 50 dB attenuator, then fed the other side of the attenuator into the RF in on the converter.  I used my FT-817 as the IF.  Tuned to USB, I easily spotted the signal very close to 144 MHz on the IF.  The signal was about an S-2 with 0 dB of attenuation inline with the little 3 cm long antenna.  I began snowflaking the hairpins, adding small strips of copper to the hairpins where I saw an increase in signal and soldered them in place.  There were a couple places that gave huge increases in gain, in the middle of the filter.  After about 2 hours work, I was able to get the signal up to about an S-7 with 20 dB of attenuation in line with the little whip antenna I was using as a pickup.

The frequency stability appeared to be quite good, there was some drift when I first powered it all up, but that quickly settled down after it was on for about 5 minutes or so.  I did not have to touch the VFO dial in the IF rig after  that time.

The next step will be in building the transmit side of the transverter.  I am still waiting on my order to come in from Down East- one of W1GHZ's power meters and a couple of MMIC low level amplifier boards.  I need the power meter in order to tune the TX filter which I have yet to build.

Oh, I'm using this article as a guide to my conversion: http://www.qsl.net/g0ory/2.3g/31732/31732.html  No sense reinventing the wheel, being Adam already did the hard part.

Stay tuned, when I start on the TX portion, you can bet I'll talk about it here.