[NLRS] 10GHz transverter info

[Doug Reed]

At 11:35 PM 10/25/2003 -0000, Scott wrote:
>Looking for your experiences with the various vendors.  Down East Microwave
>has a great variety, but the prices make me gag.  I'm cheap, so I will
>probably buy boards and buy parts separately.  I have another hobby that is
>expensive and I don't want to spend gigabucks on radio gear that I will only
>use in contests or an occasional road trip to a mountain top....
>Scott N0EDV

Hi Scott. 

Most of the NLRS group started with DEMI transverters. Most of them have
added amps and are now running 1 watt or more. I'm in the second group that
bought DB6NT transverters last winter. The majority of the units were
purchased assembled although quite a few kits were purchased. I'm cheap so
I bought the kit. Unless you have a lot of time, buy the assembled unit.
That is particularly true if you buy the DEMI and don't have a laboratory
full of equipment. I liked the DB6NT because it runs 200mw out of the box.
I figure I'm only 6-10 dB down from the majority of the folks and signals
are seldom so weak that it becomes an issue. 

Over all, I'd say the DB6NT kit is extremely well designed and fairly easy
to assemble, assuming you have some SMT experience. With the right tools,
it's an easy kit. But the right tools include a 7X microscope, very fine
solder and a fine point soldering iron. With a large magnifying lens around
2X-3X you'll get it done but not as easy. The majority of the parts are
large enough that you can probably see them to solder them if you're under
age 40..... But you'll always do a better job soldering with magnification. 

I took it very slow and worked carefully. It took about five evenings after
work, about 3-4 hours per evening. Its not that anything was difficult, I
was just very methodical. I hate correcting mistakes, particularly with
SMT. Most of the tuneup is done with a voltmeter, just like working on
commercial FM mobile radios. Like I say, very well designed. 

I only had two problems with tuneup. The first time I tried to tune the LO,
I kept turning the coils but couldn't get the LO to start. After several
minutes I checked the layout and found I was tuning the 640 MHz multiplier
stage instead of the LO stage. (Almost wore the threads out.) When I
finally started tuning the correct coils, it worked perfectly up to the
10GHz pipe cap filters. I started tuning the pipe cap filters but couldn't
get any tell-tale response. I eventually gave up and took it over to Bob
W0AUS and tuned it on his bench. It started to work as soon as I tuned the
screws half way through the jam nut. The only reason I had any trouble was
that I never tuned the screws down far enough. 

The only other trouble I've heard of was not changing the LO tuning slug
from ferrite to brass. It has to be the brass slug to get the LO started.
On my unit the brass slug was loose in the coil and the LO frequency would
jump around. I cut a sliver of plastic bag and stuck it in the hole before
I threaded in the slug. That solved the tuning jump. I do think I'll take a
cue from the Japanese and melt a little beeswax on top of all the tuning
slugs. That will help with vibration and its easy to remove for retuning. 

My only other caution is on the 10GHz RF cables. MEASURE THE LOSS of the
cables before and after you tweak them. You'll need some 10GHz SMA
attenuators, some coaxial adapters, and a HP-431/432 or equivalent. Use the
transverter as an RF source. Measure the RF output thru about 20dB of
attenuators, then install the cable to be tested and check the loss. I
suggest testing all the cables, connectors, adapters and relays in the
10GHz side of things. 

When Bob and I started checking cables, we found that the 24" piece of
RG-213 I was going to use had about 20dB of loss. We tried several other
cables and they were much better but not long enough. We checked all the
right-angle N connectors I had and not one of them showed less than 20dB
loss. Every straight connector I had was low loss, every angle connector
was very high loss. 

When we finally started checking UG-141 cables for the internal jumpers to
the relay, we found that most would measure under .7 dB of loss but after
bending the cable to fit the new application, the loss went up to 1.5dB or
more. It is entirely possible the high loss is simply due to my not taking
time to make each connector line up perfectly. In fact it is fairly likely.
But it sure was annoying! 

I ended up using a 24" piece of UG-085 for the main antenna cable from the
transverter to the feedhorn. It was the best I could find and had about
1.2-1.5dB loss. And I had at least 2-3dB more loss inside the cabinet. This
needs to be my main point of improvement before next year. I bought the
DB6NT for it's receiver noise figure and RF output. But my RF feedlines cut
my RF power by half or more and the receiver noise figure is about the same
as a diode mixer because of those cables. It's time to find a different
antenna relay and remount the transverter nearer to the feedhorn. 

If you are good at scrounging parts, have a well stocked junkbox, or simply
like buying things on Ebay, then you could try assembling a transverter
from parts. Bob W0AUS built one of the first 10GHz transverters in the area
using a Frequency West type RF brick for the LO, a waveguide mixer, some
waveguide filters, a couple of transfer relays and a small instrumentation
amplifier. And it is still a good transverter. I used it for two years
previously and Scanner, N0KP used it this year at Lake Superior. It works
very well although its large and ugly to look at....... 

And try to keep DC power on the transverter LO all the time you are using
it. I turn on the DC about 30 minutes before I want to use it and never
shut it off all day. If you don't do this you'll drift 5-10KHz in the first
5-10 minutes. I thought my DB6NT was very stable during all the events this
year. I think the only gesture I made was to put a Styrofoam peanut over
the crystal as insulation.....

Time to go to work.....

73, Doug Reed, N0NAS.