On Jan 29, 10:23 am, "Suzy" not@valid wrote:
"K7ITM" wrote in message

...



On Jan 28, 1:52 pm, "Suzy" not@valid wrote:
Can anyone point me to a website where a home built 70 cm reflectometer
or
70 cm field strength meter can be built -- not using surface mount
components. Thanks.

Are you willing to use a printed circuit board? If so, a directional
coupler is pretty easy to make, microstrip style. You can even use a
knife to score the copper and then peel away the unwanted copper.
It's common to have one through-line and two coupled lines, one for
"forward" and one for "reverse" measurements. You terminate each
coupled line with Z0 (50 ohms, commonly) at one end and put a diode
rectifier (or other power measuring device) at the other end.
Although it's probably easiest to get a pretty good 50 ohm termination
at 450MHz using surface mount parts, you can do OK with a couple 1/4
watt metal-film axial resistors. If you used 1206 size surface mount
resistors, you should be able to solder them down with a decent
standard iron -- they aren't all that small.

If you really can't find any construction articles on making one,
maybe I could put something together and post it somewhere. They're
far easier to make than many people seem to think.

Cheers,
Tom

Thanks very much Tom. Yes I'm certainly OK with PCBs and I was aware of the
basic circuitry. But I felt that the stripline would have to be special in
some way for the 435 MHz that I need. Perhaps a special length?

The stripline (microstrip, really, if it's on one side with
groundplane on the other, which is the easiest to do) can be almost
any length. It will have zero coupling if it's an electrical halfwave
or integral multiple of a halfwave, but otherwise, it will work. It
will have the maximum coupling if it's an electrical quarter wave
long, and the change in coupling versus frequency is minimum at that
length, but it's usable even if it is very short. I'm using one
that's about half an inch long down as low as 1MHz; the directionality
is fine, and the coupling is simply very small. You can use the free-
to-download RFSim99 program to get a microstrip design for a
particular coupling, but what the program doesn't tell you is that the
coupling they list is for a coupler an electrical quarter wave long.
To a pretty close approximation, the coupling will go as the sine of
the electrical length, so for example if you use the design for a 20dB
coupler but make it only 1/8 wave long, the response will be about
sin(45 degrees) or .707 times as much, or 3dB lower: it will be about
a 23dB coupler at that length. That means the coupled port power will
be 23dB lower than the input power, so if you had 10 watts going in,
the coupled port would receive about 0.05 watts, which is 1.6V rms.
That likely is a bit more than you need to drive a diode detector.
There can be some advantage to keeping the voltage low enough, if
you're using a diode detector, so that the detector is in the "square
law" region: DC output _voltage_ is proportional to RF input
_power_. On normal FR4 board stock 1/16" thick, leaving one side as a
ground plane, a pair of traces 1/8" wide, separated by a 1/8" space,
will give you about a 25dB coupler and pretty close to 50 ohms. It
does depend on the particular dielectric constant of your FR4
substrate, which varies from manufacturer to manufacturer and by
particular material type. In that microstrip environment, the
wavelength of 450MHz signals will be on the order of 40cm, or about 4
inches for 1/4 wavelength, so you can see that your coupler doesn't
have to be very long, physically.

If you don't mind working with SMA connectors, you can get SMA jacks
that mount directly on the edge of a 1/16" thick PC board, and that
makes it very easy to make a connectorized version of a coupler. You
can get SMA-to-BNC (or to other series) adapters--mpja.com have them
at reasonable prices, for example--to get to other environments.

If I wanted a coupler for a power/SWR meter for 450MHz, and assuming I
was going to use Schottky diode detectors, I'd first think about the
range of powers I wanted to handle, then select a coupling to give me
perhaps 0.2V RMS at max power at the coupled port. If space
permitted, I'd make the coupler an electrical quarter wave long, since
its sensitivity would then be least affected by frequency (the very
flat top of the sine curve). I'd use a DVM to read out the detected
DC voltage (and make sure my DVM or digital panel meter or whatever
wasn't affected by the 450MHz signal directly!). A 200mV full-scale
3.5 digit meter would then cover to very low SWR at full scale, and to
fairly low power. The one other thing I'd want to do is insure that
the meter--the directional coupler part, specifically--was optimized
to 50 ohms, which would require a known good 50 ohm (assuming I wanted
it to be 50 ohms) load to test. Likely I'd make some pretty accurate
measurements of the board substrate thickness and relative dielectric
constant before starting, so I got the trace widths and spacings right
from the start. It's possible to trim the coupler up in impedance by
carefully narrowing the traces, and down by adding a grounded plane
above the board--adjusting the spacing to trim the impedance. Be
aware that each of those will also change the coupling slightly,
however.

Hope that helps some!

Cheers,
Tom