Converting a Qualcomm Ku-band Transverter

for 10GHz X-band Amateur Radio Operation

Jeff Keyzer, KF6PBP


    For some time now, the San Diego Amateur Radio Microwave Group has been converting Qualcomm 10GHz transverters used in the old Ku-band OmniTRACS system to amateur radio use centered around 10.368GHz.  I have been convinced by a friend of mine and a member of the group, Tony (KC6QHP), to do the conversion simultaneously with Shaun (KD6AZU), another UCSD student and Wireless Communications Club member.  In the end we hope to have two newly converted transverters, which, when combined with Tony's existing station, makes three 10GHz rigs we can play around with.  In order to document and share the conversion effort, I have created this page and will update it periodically as progress is made.

News:

8/27/00 - My rig worked well for the 2000 ARRL 10 GHz and Up Contest. Tony, KC6QHP, Royce, KF6PEO, and I met in Santa Barbara and spent the weekend operating there. We had a great time and made a bunch of 300+ km contacts to San Diego and Northern CA. I learned a lot about operating in the field, including: The display on the TR9000 is HORRIBLE in sunlight, don't get magnets too close to a TR9000 or it stops working until you demagnetize it (!), a good working surface makes a big difference, carry extra fuses, next time bring a hat, and don't drive up a mountain on two gallons of gas!

5/23/00 - The weekend before last, Tony and I put everything in a nice aluminum box and mounted the box and dish to a tripod given to me by Royce, KF6PEO. (Royce also made the box, which turned out extremely well. Thanks Royce!) Over this past weekend Tony and I painted the dish and put my callsign on it. Now people will stop asking me what Airlan (originally written on the dish) means and start asking "KF6PBP? 10 Ghz?" instead! The whole assembly has been tested extensively around town and works very well. I'm almost ready for contesting!

5/2/00 - Finally, I have made an update to this page! Due to a light class load during my final quarter as an undergraduate at UCSD, I have had time to put a lot more work into all of my microwave projects. The 1 watt X-band power amplifier has been tuned, the mixer board installed, a new 992 MHz filter installed, and the whole system has been thrown together enough to be usable.

It works! I am now active on 10 GHz during the weekly the San Diego Microwave Group net using a simple horn antenna. The LOS path from school to the SD X-band repeater is about 30 miles and signals are always full quieting.

Note: Most recent progress is highlighted.

Pics of the Unmodified Main Transverter Unit:

These images show the unit as I received it from Kerry Banke, N6IZW, of the SD Microwave Group.
Main Unit - Image 1, Image 2

Underneath the milled aluminum cover is a teflon circuit board covered with microstrip and microwave components.
10GHz subsection - Image 1

Other Images:

Miscellaneous parts to be incorported into the final setup: Image 1
Included in that image are some interesting parts - a 10.00MHz temperature stabilized reference crystal and the RF relay I intend to rewind to work on 12V instead of 28V.

Modification Process Begins - Milling the Chassis:

3/6/99 - The transverter chassis must be milled to allow SMA connectors to be placed onto the bottom of the board for 10GHz transmit/receive.  Thanks to Shaun and Tony's help, we were able to mill both boards on a Saturday morning.  We also performed drilling for the connector center pins and mounting screws.

Milling the Transverter Chassis:  Image 1, Image 2, Image 3
The final result: Image 1
(Looks pretty good, eh?)

3/9/99 - The SMA connector mounting holes were drilled and tapped, and the connectors mounted.
SMA Connectors Mounted - Image 1

Board-Level Modifications

3/9/99 - I made the proper cuts to the microstrip on the surface of the board (1pF capacitors will be placed across three of the cuts), and trimmed the SMA connector center pins to be nearly flush with the board.

Board with cuts and SMA connector center pins: Image 1

5/5/99 - After receiving a completed transverter from Kerry to use as a reference, I made the last of the modifications to the 10GHz section. These mods included one more cut and a few tuning stubs at various locations on the board.

5/13/99 - The synthesizer modifications have been completed. This entailed lifting several pins on the Qualcomm PLCC surface mount synthesizer chip and soldering one of those pins to ground. This programs the synthesizer for a 2272MHz VCO, which is multiplied and mixed to yield the 10.368MHz desired transverter output. Eventually I hope to put up a block diagram and theory of operation section for the benefit of those not involved with the SD Microwave Group.

5/2/00 - Here is some more info from Kerry on the synthesizer mods:

Synthesizer programming:
Lift pins 4,7,8,9,10,13,14,21,22
Tie pin 10 to pin 17 (15,16,17 are all ground)
Add (2) 3000 pF caps to reference supression filter
Add (1) 1000 pF cap to reference supression filter
Add two 0.5 pF caps in parallel to lower VCO frequency.

The caps were added to various locations around the synthesizer, someday I may get a good picture online.

Microstrip Filter and Capacitor Modifications

4/11/99 - The four microstrip filters have been modified and the three 1pF capacitors added.

The 10GHz subsection with filter and capacitor mods - Image 1
A closeup of the added capacitors - Image 1
A closeup of one of the microstrip filters - Image 1

Mixer Board

6/99 - The small mixer board is complete. This board takes a 1136MHz LO from the transverter, amplifies it, and mixes it with 144MHz from a common 2 meter handheld radio, supplying 992MHz as a second IF to the transverter board. The board is a surplus Qualcomm part which has been modified and combined with a Mini-Circuits SRA-11 mixer module.

The 1st IF mixer board - Image 1

1 Watt Qualcomm Power Amplifier

5/2/00 - Kerry gave me a Qualcomm PA to use with the transverter. This will take the few mW produced by the main transverter module (at 10 GHz) and give back around a watt. The PA comes with it's own power supply board which has a transmit key to control the supply voltage to the PA. The PA draws a continuous 1.2A or so when on regardless of whether any input signal is present. Thus, to increase battery life the power supply is turned on only when transmitting.

The PA had to be tuned to 10.368 GHz before it could be used in my setup. This procedure took a few hours and involves removing all tuning stubs from the microstrip on the board, then sliding a small square of thin copper attached to a toothpick along the microstrip lines until the maximum output power can be reached for a given input. A similar piece of copper is then soldered in that location and the process is repeated. Chuck WB6IGP has written an article describing this procedure. The result is a PA that gives just short of a watt out for 1 mW in and has about 29.5 dB gain.

The PA and power supply - Image 1
A closeup of the PA showing some tuning stubs I added - Image 1
The bottom of the PA showing SMA input/output - Image 1

Miscellaneous

5/7/99 - Tony and I successfully rewound an Amphenol RF relay to switch solidly at 12V instead of 24-28V. More about this process can be found on the Down East Microwave site. Paul Wade, N1BWT, has written an article about converting RF relays down to 12V, and it really is a very simple coil rewinding process.

Added SMA connectors for 992MHz IF and 10MHz reference to transverter.

5/20/99 - Tony assisted me in producing a waveguide type bandpass filter, designed to pass the 992MHz IF while rejecting the LO of 1136MHz and the 1136MHz+144MHz second image produced by the mixer board.
It's not beautiful, but it works! - Image 1

5/2/00 - The waveguide filter will not be used in the final system. Instead, a 992 MHz helical filter will be used. It is much smaller and lighter but performs the same function.

Final Assembly and Testing

6/1/99 - I brought the modified transverter and all other parts over to Kerry's (N6IZW) place, and we successfully tuned the waveguide bandpass filter to 992MHz. We also added some surface mount caps to the transverter to allow the synthesizer to operate at its new frequency. The transverter powered up and seems to be working well! We were able to transmit and receive on 10GHz, albeit with very low power and sensitivity. (No antenna!) The 10GHz section needs some tuning, but worked amazingly well for the first time it had ever been powered up. We mounted the second conversion mixer board to the transverter, but weren't able to get it working in the short time that we had. Kerry feels however that this should be simple to fix given some time. Overall it was a very productive evening and I am happy to say that months of effort on everyone's part are certainly paying off!

5/2/00 - It works! Apparently the problems at Kerry's last year turned out to be a bad cable in the test setup. The tuned PA was attached to the transverter by drilling and tapping a couple holes for screws. I wired up the PA, T/R coaxial relay, 992 MHz filter, 10 MHz TCXO, and PA power supply in order to do some tests. The temporary test setup is duct taped to a sheet of plexiglass to give me a convenient way to carry things around over the next few weeks. A friend is currently making an aluminum box to put the whole thing in. A feed still needs to be constructed. The box will also need to be mounted to a dish and tripod. This may end up being the most difficult part of this project!

Upon throwing everything together, I desperately needed a temporary antenna. For the moment I am using an SMA to WR-90 transition and a horn made of cardboard and aluminum foil that Tony made for me. It works surprisingly well and allows me to work the SD X-band repeater from the third floor of EBU-2 at UCSD without any problems.

The transverter and PA - Image 1, Image 2
The "bottom" of the transverter showing the installed mixer board - Image 1
The duct tape and plexiglass mess I have made - Image 1
The cardboard and aluminum foil antenna - Image 1, Image 2
(Yes, the polarization on the horn is wrong - this took a few minutes to figure out when I couldn't hear the repeater but could hear the beacon at the same location!)

5/23/00 - The transverter has been mounted in an Aluminum box and the box and dish mounted to a heavy duty tripod. I am now using another offset feed parabolic dish purchased for $15 at Murphy's Surplus in El Cajon. It was originally used for a 2.4 GHz low power local area network and has a much better shape than the DirecPC dish I originally was going to use.

The newly painted dish and transverter box mounted to my tripod - Image 1
A closeup of the feed - Image 1
Side view showing connectors, etc - Image 1
The inside of the transverter box - Image 1

To automate the switching between transmit and receive on the transmitter (a signal must be pulled low on the Qualcomm board and a relay energized), I built a simple TR switching board designed by Chuck, WB6IGP. This senses the RF energy present on the IF line of the transverter and switches two FETs. At the moment I have this board connected to the IF line with a coax T which is probably not the ideal way to do it. Someday I will probably improve this.

The TR switch board installed - Image 1

6/6/00 - I bought a Kenwood TR-9000 2m all-mode rig to use as an IF transceiver to allow me to work FM and SSB on 10 ghz and also on 2m around town. The output power of the rig has been reduced from 10W to 10-20mW to work with my transverter. I have tested it and it works very well. The frequency display is made of red LEDs and apparently does not work well in bright sunlight. I have not experienced this yet but will probably have to make a sun shade of some sort for the contests.

8/27/00 - I finally added a compass rose to the tripod for contesting and got some pictures of it taken (sometime last month). Kerry N6IZW provided most of the assembly and I made the brazing rod indicator. The actual compass rose is printed on a piece of overhead projector slide material which is glued to an aluminum disc. PVC fittings hold everything together along with a couple set screws. Thanks Kerry for coming up with the design and providing parts.

The compass rose assembly - Image 1, Image 2

I have also modified a Qualcomm board to function as an 1152 MHz LO and marker generator for all the ham bands 1152 and up. These modifications are detailed here. The 1152 board generates an LO at 1152 which is very rich in harmonics. I can pick up the 10368 MHz multiple from over a block away with only a short stub for an antenna on the 1152 board. This board is very useful for making sure the rig is working on receive when I don't have a beacon I can listen to easily.

The 1152 board - Image 1

I finally added labels for the controls. These won't last long but were useful during the contest.

Waveguide output label - Image 1
Side panel controls labels - Image 1, Image 2

Next up: Investigate TR switch misfires in heavy RFI environments? Move everything back to San Diego.

Last updated 8/27/00, jkeyzer@ucsd.edu.