The DXR-700 Conversion / Transverter Project  

Well I have done 1296, 2424 transverters.... What will we do now??  5.7gig of course!!

Some of us here in ZL have been  fortunate enough to be on the receiving end of  DMC DXR-700 6 odd Gig  Digital Microwave Radio units minus the digital modulator and control boards.

The logical thing here to do is to bring it down on frequency to 5760.

This page endeavors to detail what a group of three of us have been doing here in Auckland to get these going.
The team involves Harry ZL1BK, Keith ZL1BQE and myself Simon ZL1SWW. Each of us have areas of expertise in the field so we were tasked with self appointed jobs..

Be Sure To Check For Updates As I Will Do Them As I Get Time...

 

The Local Oscillator

Not just one, but two are employed in this unit, one for the transmit mixer and the other for the receive mixer. They are in the form of a PLL  that is tripled up to get the required frequency. The reference starts at 10 MHz.
Keith ZL1BQE - the PLL gun, was able to adapt a piece of code we were  using on another 2Gig PLL project I was playing with, and using conditional assembly, we make a few tweaks in the assembly code and we can use the same PLL code for both types of PLL 2Gig or 5.7Gig. It just needs reassembly and reprogramming the chip.

The PLL is able to be controlled in 1MHz increments, by a PC in the same way we have done it before in other projects such as the Maxon PLL controller on the ZL1VK web site.

In our design criteria, we decided that the micro should have a pin for selecting two frequencies. This pin is sampled on a processor power cycle.
The initial thought here was to have it so the micro will send 6.5Gig frequency command to the PLL. This is to allow a constructor to get the PLL up and running with just the chip connected to the unit in the way I will describe later. It should show the PLL lock LED blink briefly and then go out, you can do a finger check on the VCO to knock it out of lock to see the LED light.

Here is a basic layout of the PLL / Tripler chain as we see it. We have no documentation on these units so we have to make our own.


The Micro PLL Controller info..

The micros used in this project are Motorola 68hc908QT1 units. They are a small 8 pin device that has in internal clock.. Keith ZL1BQE, our PLL gun has kindly made up the code for this.

Serial comms is done by a software UART routine that derives its data rate from the internal oscillator.
For the serial comms routine to work, an accurate clock frequency is needed or the data bit times will be out and will result in garbage on the computer screen.
For this reason, we can supply on request, the micros pre programmed as each unit we have to calibrate the oscillator trim value at location 0XFFC0. There is a wide variation in values. We plug this value into the assembly code and recompile and burn the new code into the chip.
.
We tried to make the system clock off the reference oscillator but it proved to be unreliable. Both Keith & I tried for while to get the serial routine to run off the reference but had some weird irregularities that we couldn’t iron out. This was done not for this PLL but it was another of my PLL projects at 2.4 Gig. The codes uses conditional assembly and a few small value changes allows this code to be used with an MA1019 PLL controller as well as the LMX2326 as used on the DXR700 PLL.

Within reason, the code allows the user to input any reasonable frequency span and it will faithfully push it to the PLL chip. Weather it will lock is a different story. Eg, you could put in 4Gig and the commands would be  sent to the chip.

Serial comms happens on the same pin and hence we use a small adapter to change from RS-232 / V.24 levels to a 5v signalling. Pin 7 is the data in/out pin for serial comms.

Operation is simple, only 2 commands
 'R' will return the programmed chanels
 'W' followed by 8 digits will program chanel 1 with 1st 4 digits in MHz and
             W60005616 ch1=6000MHz ch2=5616MHz in 6G version

In a default condition, the micro sends SPI instructions to the PLL to set the final frequency to 6500MHz. This allows the user to check that the PLL is working before modifying it.

The PLL runs at a third of the final frequency of the LO. The final stepping rate is at 125KHz so that means we set the reference division ratio to 41.66667 KHz so from a 10MHz reference we have a division of 240.

SPI data is sent via 3 pins, Pin 2, 3 & 4. Functions are  (Pin 2) Enable, (Pin 3) Data and (Pin 4) Clock.

There is a link on pin 5 that allows two presets. The preset frequencies are 5616 (link in) and 6500 (link out).

Your are, of course, not locked to these values as you can send commands to put whatever you like in each location. Values are stored in flash and committed in real time with each value change.
The jumper configuration will be read on a power cycle of the micro.

The pinouts of these chips is somewhat unusual when it comes to the supply rails. Pin 1 = +VCC and pin 8 is ground.

We decided in transverter use, we normal work at integer values so we can send commands via the serial pin in 1MHz increments.

The  Cavity Beneath.

Once upon a time, in this units life, it once had a control board for controlling the PLL and the other digital functions. The two halves originally mated together to plug into this board.
The connectors for this board are offset just by  about the thickness of the connectors.

I have designed a group of PCBs that have been through several iterations until we came to a happy medium. One has to be careful when designing it, that you make sure that the parts will not clash when the two parts are pushed together.
There was a thought that we put all parts on one board, but it was found that there were no supplies of 12 way  IDC headers or sockets for a ribbon cable connection to the RX unit. So we are back to using two boards.

The PDF link here shows the pinouts for the two halves.


PCBs and details are in the zip file supplied.. For people without Protel, use the PDF for RX and TX but chack the scale is right before etching as no guarantee can be placed on board size accuracy.

Here is the PDF of the Schematic for the Micro and Level translator. It is not a direct corellation to the PCB layout as it doesn't include the voltage regulators.

The  TX Mixer.

The  RX Mixer.




The  PLL Controllers

As mentioned above, there have been a number of goes at PCB design. As they say, third time lucky!! I finally came up with something that fits as well as being simple to build.

The boards are built in a surface mount style due to keeping costs down as we need to be able to successfully solder the pin headers in. I guess that we could have done it by elevating the pins so you can solder them, but that creates dilemmas when fitting it in the case.

In addition to this, the board mates down to the chassis plate with not too much room to spare so it was decided to do surface mount to stop solder connections shorting out.

One has to note from the PDF that shows the pinouts is that the RX and TX pin numbering is different. The functional locations of the pins, however, are the same.

For Supply of pre programmed micros for the PLL controllers, please contact me at the email address shown here..
The cost is NZ$20 for a group of 3 chips, 2*PLL controllers and 1*Sequencer chip. There is extra cost for Postage and Packaging.
Please see the Sequencer section on my website for details. Look at the bottom of the page for the unit with the PWM output.

The PCBs are arranged in 2 ways..

TX Board.

Basically it contains 2 micros. One for PLL control and the other for TX sequencing.

Each board has its own regulators so that they can run as autonomous systems for testing purposes. There are 2 LM317s and a 78L05 for the Micro logic.

There is a pot which is used for setting up the control voltage that trims the TCXO PLL reference frequency. It should sit somewhere around 2.45 volts. Fine adjustment is done by limiting the voltage range the pot can sweep across as well as using a 10 turn unit.

There are various pads on the board that bring out the TX enable line and VCO voltage etc. This is shown in the pinout PDF.

TX board voltages .

We have found that the TX section runs best at about 7.5 to 8.3v. We thought initially that the unit was to run at 10v but am unsure due to the lack of info.

The PLL supplies should remain at 10v. We used LM317s here as they are cheap and easily available..

RX Board.

This is very similar to the TX unit from a PLL control standpoint as well as having the TCXO trim pot.

The only thing missing here is the TX sequencer micro.

PCB Patterns etc..

The units were designed using Protel and I have included the Protel PCB files as well as a PDF. I cannot guarantee that the PDF is to scale as that depends on your printer settings.

The PDF will be in mirror format, evident by the writing being backwards.

A Cheap and Reasonably Successful Way to Etch…

1. After printing out the correct way, and checking the scale if using the PDF, cut out the printout to just bigger than the pattern.
2. Prepare the PCB copper by scrubbing with steel wool to get it nice and shiny all over. Rub down with isopropyl alcohol to de grease.
3. Place the toner side of the printout down and iron on with a good hot iron. You can heat quite a bit on G10 / FR4 board with no probs.
5. Soften the paper by submersing the PCB so that the paper softens and can be rubbed off with your finger.
6. Carefully use a toothbrush to remove paper dags that maybe left behind, as this will impede the etching process.
7. Check for broken etches and repair, if any with a Dalo Pen etc.
8. Throw it in the etchant solution and baby sit it until all is etched correctly.
9. Once etched, you can then scrub off the toner to reveal the shiny copper.
10. Drill the holes for the pin headers..

VCO Mods
As the PLL stands, we cannot get the VCO to run low enough for the output of 5616MHz. What we had to do is to lengthen the tracks on all parts of the VCO "Airstrip" or "Runway". All tracks have to be symmetrical as asymmetrical layouts will affect phase noise.

Carefully move the varicap diodes by using solder wick. Wick the hotside first as this is easier to remove solder. The ground side has a lot of thermal bulk and will take a long time to heat. Gently with a knife blade, heat and prise up the diode leg. Now heat the ground side and remove  the diode. DON'T SNEEZE WHILE DOING THIS, YOU WILL LOSE IT!!! ;-).

After making sure the polarity is right, solder them in at the end of the track. Diode should be reverse baissed so that the varicap action will work.

	The un modded VCO
	The modded VCO. The Loop of kynar is there to reduce Phase noise. In addition to this, you may need to add a 0p5 cap  across the base feed, and one across the diagonal cap as seen in the pic. All this is a bit of a play around as it is possible to get low frequency "wrigglies" running over the LO output. We don't want the wrigglies as it will show up as rumbling on FM and bad warbling on SSB.

 



Protecting the RX IF Port.

I added a little attenuator pad at the back of the RX IF port of the converter to aid in protection of the output MMIC as it is possible that it could see a short burst of RF from the 2m exciter IF. Across the output I put a couple of 1N4148’s back to back across part of the pad to limit input voltage to the MMIC. I have yet to see if it affects the noise floor.

 

The  Filter Conversion.

As we have a 144 Meg IF on TX, there is the possibility that the TX IF may appear in the output. We would need to filter this out.

If you were lucky to get a filter unit with your converter, it is reasonably easy to bring it down to 5760.

This is done by adding small extensions to the stubs inside. This was done by chopping up sections of hardline to project a bit higher than the original stubs.

The finished length of the stubs should be between 9.2 to 9.4mm long. Any longer makes it hard to tune.

1. Start by getting a piece of hardline and use a file to gently file one side to almost reveal the inner but not all the way. The reason for this for ease of cutting with a sharp hobby knife or similar.
2. When done, run the knife down the side to get a slit that runs just through the copper to the dielectric.
3. From there, measure 10mm long pieces and cut by using the knife and roiling the hardline on a flat surface to create a score mark all the way round.
4. Once scored, gently rock the free end until you feel it give.
5. Slide the short bit off the end, leaving the dielectric behind on the larger bit. May have to slightly open the slit a smidgen to get it off.
6. Repeat for all the stubs as shown in the picture.
7. De burr any edges.
8. Press fit all the bits on to the stubs.
9. Once all on and snugly fitting, file off all stubs to 9.4mm.
10. Reassemble the unit.

Tuning.

We are only adjusting the screws on the topside housing. This is the housing part that has the SMA connectors on it.

1. As the unit is out of tune, remove the second tuning screw on the bulky side of the filter housing, not the screws on the modified side.
2. Put a small pickup I that hole that is connected to a Spectrum Analyser or RF pickup detector.
3. Loosen the locking nuts on all the associated tuning screws.
4. Inject a 5760Meg signal into the port nearest the end with the screw removed. The reason for this is that the rejection ratio is high and it is likely you will not see a signal at the other end of the filter.
5. Adjust the screw to get a peak at 5760MHz. Tighten the locking nut (not too tight as it may shift the screw as it lifts it in the thread) to hold the screw in place.
6. Install the removed screw assembly and then put Spectrum Analyser or RF detector at the other SMA port.
7. Peak up all remaining screws to get a peak, lock the nuts but not too tight as you will then fine tune them all for max output again.

A correctly tuned filter will have about a 20MHz bandwidth.

 

Making a Common TCXO Reference on TX + RX Units.

This mod is especially useful when trying to use in SSB as the TX / RX frequency should be the same for easy transceiving.

The PLL units on each converter have a "Fine Tune" adjust where the VTCXO unit has a voltage trim point to bring it on frequency. (Remember, we are tripling so any discrepancy in the reference will be magnified 3 times at the final output frequency).

Harry ZL1BK, has been able to get the units to run off one of the VTCXOs, in this case, the TX module.

As seen on the pinout, there is a ref frequency out that can be effectively paralleled to the RX unit providing the following is done.

1. Remove the power to the VTCXO by pulling a Zero Ohm resistor that will remove the power.
2. Remove the 5 legged device near the VTCXO, this is a buffer from the VTCXO to the Reference pin.
3. Bridge the existing in and out pads with a resistor to back feed the clock signal.
   
   

   Unmodded Ref OSC section, the H5L Unit is a buffer chip.	Modded ref osc - the 0 ohm resistor takes the supply off the  VTCXO. The resistor allows you to back feed the TCXO reference off the TX board. Once the mod is done, all you have to do is add the link to allow the TCXO signal to pass from TX to RX units.

   
   Mod for Removing VCO Beating Problem. 
   
   PLEASE NOTE THAT THIS IS NOT REALLY A RECCOMENDED WAY OF DOING IT NOW AS THERE IS A MUCH MORE SIMPLE / LESS PARTS COUNT. As this project has evolved to it current state, this information is here for interest sake  and also for other pertinent information. Please check the update frequently as I will update as I do more on this.
   
   It has been noted that there has been a warble between the two PLLS. The only way to cure this so far has been to disable the opposite PLL to stop beating with the other unit.
   I devised a small circuit that disables the VCO by pulling down the base of the VCO feed transistor. This device is here for some noise immunity on the supply rails but it serves well as power switching for the VCO.

Please see the piccies below for what I have done. All the pics are for the RX side but the TX side is much the same but componnt location is slightly different wrt the power supply.
I supply the little circuit by a handy 5v supply point just under the PCB. All this is used for biassing the base of the second transistor on the board. I needed the inversion so I used two devices with a 100 ohm in the collector of the second output transistor. Grounding and support for the little board is via the -ve pin of the SMD cap the board is sitting on.

The little VCO power control board. Any NPN Gen  purpose switching transistors will do here.	Redirecting the tripler ERAs and the GaAsFET power supplies to be switched. Note the small cut on the PCB to isolate the power.	The base Connection to the VCO supply transistor.

 


HOME