On Jan 31, 1:51 pm, Owen Duffy wrote:
K7ITM wrote in news:9e844e58-a673-4ec0-9a0b-ec15f8cc8f30
@c4g2000hsg.googlegroups.com:





On Jan 31, 12:31 pm, Cecil Moore wrote:
K7ITM wrote:
To me, having a
linear power scale is a big advantage, because then you can
reasonably
accurately figure SWR without having to worry about temperature
compensation of the detectors.

Can you define what you mean by linear? Straight line?
Since we can only measure voltage and current, in order
to obtain a linear power scale from a linear meter, it
is necessary to supply some pre-display computing
ability (microcomputer).
--
73, Cecil http://www.w5dxp.com

See earlier posting in this thread. See various Avago ap notes, such
as AN 969. A diode detector run at low input provides an output DC
voltage that's a constant times the square of the input RF voltage.
If the input voltage is, or is assumed to be, at some constant
resistive load impedance, the DC output is linear with RF power
input. The proportionality is temperature dependent, but if two
detectors are constructed the same and run at the same temperature,
and run in the signal level region where that relationship holds, then
the ratio of the output DC voltages is a very good approximation of
the ratio of the input RF power levels, and thus is useful for finding
the SWR if the detectors are attached to the forward and reverse ports
of a good directional coupler. Top end of the useful "linear power"
range using an HSMS-2850 single diode detector is about 10mV DC
output. If you can measure the DC accurately down to 1uV (a bit
tough, given thermal emfs, but possible), that gives you about a
10000:1 power range, or 100:1 RF input voltage range -- or about
1.02:1 SWR. Chances are very good that a home-built coupler won't be
accurately enough matched to 50+j0 ohms to worry about anything that
low anyway, even if you had a reason to care about it.

Cheers,
Tom

Tom,

This is further from Suzy's needs, but...

Operation of a diode detector in the square law region isn't out of the
question, but it takes some serious gain to drive a meter. There are some
good chopper stabilised op amps out there that have uV offset levels and
single supply rail and input to below the negative rail eg LTC1050.

Another alternative is the AD8307AN log amps for a linear dBW scale. You
could even use one on FWD and REF detectors and difference the outputs in
an op amp for a direct indicating VSWR or RL scale. I have thought of
getting one of these chips and seeing whether its response is fast enough
to drive a PEP amplifier for SSB telephony.

Back to Suzy's problem...

The instrument downstream of the sampler is not so much the issue as
building and calibrating a sampler when you have no test gear.

Suzy, if you see a Revex W560 going on VKHAM for $100 or so, it is a good
buy. It has HF to 70cm (two independent couplers, ie four coax
connectors), and works pretty well.

For a dummy load, the market was flooded with terminations from 25W to
about 60W that had been scrapped from AMPS base station equipment, and
they were sold at hamfests for $20 or so, you may find them if you look
around.

Owen

Yes, there are several linear-in-dB RF detectors out there. Linear
Technology also have them. I really like that idea; they're typically
much more temperature stable than a diode detector. But Suzy wanted
to avoid SMT.

I've used a Harris chopper-stabilized op amp with HSMS-2850 zero-bias
detector diodes, and it works well, but I did learn something about
the need to be really careful around the chopper capacitor pins on
that op amp before getting it right... But it's also not difficult to
find a digital voltmeter that will go down to pretty low voltage at
high impedance. A 4.5 digit meter on a 200mV scale does ten
microvolts, and the simulation I ran last night suggests you could see
down to about -50dBm power level with that. When you get down to
10uV, you have to get serious about avoiding thermal emfs.

I suppose it makes sense to just drive the detector hard and run it
right into an analog meter movement, and then calibrate the meter.
Actually, at that level, the detector should be pretty linear in
voltage. That actually makes it easier to detect down closer to 1:1
SWR anyway.

I posted not too long ago about a load I made with four 200 ohm 2 watt
metal oxide resistors that shows what to me is remarkably good return
loss out to well beyond 450MHz. It was very cheap to make. But
there's no guarantee that some other brand of resistor would give such
good results. It may have just been a fluke that the one I made
turned out so good. (But I'm not about to toss it out!)

Cheers,
Tom

On Jan 31, 5:01 pm, Jim Lux wrote:
Suzy wrote:
At the risk of thoroughly boring you all, I'll summarize the position to
date. I have a good workshop with power tools and I like metal bashing and
am quite happy with PCBs. I even have an electronic calliper so can measure
thickness accurately. However, my eyes will not allow very fine work like
SMD. I am looking to build an SWR meter using two 1 mA meters, one for
forward and one for reverse. I am looking for a practical (non-theoretical)
article on how to build one. I am wondering if there is one the ARRL
handboook before I go to the exp-ense of buying one here in Australia. Most
of the beautifully argued theory on here is way way beyond me. Any pointers
to a suitable article?

One might find that you can BUY a surplus directional coupler for 440MHz
fairly cheaply. Check Ebay, etc.

Then it's just a matter of wiring up electronics to the coupled ports.

:-) I've thought maybe I could just etch one next time I'm making
boards and drop it in the mail to her. Heck, I'd even toss detector
diodes and terminating resistors on it.

Cheers,
Tom

Owen Duffy wrote:
Thanks Owen. BTW, what type of coax connector? Not PAL surely!

No, but they are not all that bad. Almost no one manufactured VHF land
mobiles here with UHF connectors, but they did use PAL (Belling & Lee)
once (Pye Reporters for instance).


Do Australian TVs use "Bloody Belling Lee" connectors like we still do
in the UK? What does "PAL" stand for?


--

73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

On Jan 31, 3:19 pm, "Suzy" not@valid wrote:


"K7ITM" wrote in message
....
See earlier posting in this thread. See various Avago ap notes, such
....


Much too theoretical for me!

OK, here's non-theoretical, practical.

Get a piece of FR4 PC board material, copper-clad both sides, 50mm
wide by 110mm long (neither of these is very critical, but should be
at least that long) by 1.6 mm thick. One side will remain all copper
clad, as a ground plane. On the other side, fabricate at a minimum
three traces, as follows. To make it easier to describe, assume you
are looking at the board with the 110mm dimension horizontal and the
50mm dimension vertical.
Trace 1: the through-line. It will run the length of the board
(110mm), centered between the two sides. It will be as close as you
can make it to 2.9mm wide, uniform width from board edge to board
edge.
Trace 2: It will also be as close as you can make it to 2.9mm wide
for its whole length. The following is the center-line of the trace.
Start at the top edge of the board, 9mm from the left side. Go down
to 9.0mm above the center-line of the through-line trace (Trace 1).
Turn toward the right edge of the board and follow parallel to trace
1, staying 9.0mm center-to-center. Thus there should be a gap of
6.1mm between the two traces. Go a distance of 92mm. Turn toward the
top of the board, and extend the trace all the way to the top.
Trace 3: It will be the mirror-image of trace 2, same 2.9mm width.
It will start 9mm from the left side at the BOTTOM edge, go up till
its centerline is 9.0mm from trace 1, follow trace 1 to the right for
92mm, and then return to the bottom edge of the board.

Mount an edge-mount BNC connector to each end of trace 1, shell to the
ground plane and center pin to the trace. Failing that, do something
equivalent with coax or connectors...if you trimmed the end of a piece
of coax so the braid connected to the back side of the board and a
very tiny stub of exposed center conductor could be soldered to the
end of trace 1, that should be OK.

You'll need two 50 ohm load resistors. 1/4 watt is plenty. Since 50
ohms is not a common value, you may wish to use two 100 ohm resistors
in parallel for each of these 50 ohm guys. Solder one of them so it
connects with vanishingly short leads between the RIGHT end of trace 2
(at the top of the board) to the back-side copper immediately opposite
that point. The resistor(s) will be soldered to points immediately
opposite each other, front and back side. Solder the other 50 ohm
resistor between the LEFT end of trace 3 (at the bottom edge of the
board) and the back side of the board. Those are the termination
resistors, and they are the ones you would adjust to get the best null
when feeding power through trace 1 to a good 50 ohm termination.

Now you'll need two detector diodes (maybe Owen can help out here; I'd
use some surface mount schottkys, but...) and two small ceramic
capacitors. 100pF would be a good value, but it's not critical. All
leads should be so short you have trouble seeing that there's any lead
there at all. Solder one lead of a capacitor to the LEFT end of trace
2 (at the top of the board), and one lead of the other capacitor to
the RIGHT end of trace 3 (at the bottom of the board). Those are the
ends without resistors. On the back of the board immediately behind
where you soldered the capacitors, solder the ANODE of a diode, one
for trace 2 and one for trace 3. Arrange things so that the free lead
of the capacitor and the free lead of the diode (the cathode) come
together off the edge of the board. OK, I lied: leave enough lead to
solder another part to, there.

Now get or make a couple small RF chokes, about 100 nanohenries. The
inductance isn't critical. The way I'd do it is to wind some magnet
wire onto a small machine screw. For example, try about ten turns on
a 4mm screw. The wire diameter should be roughly 1/2 to 3/4 the screw
thread pitch. You can then unscrew the screw and if you're careful
with it, the inductor will be reasonably self-supporting. Next,
you'll install these two and a couple more capacitors. Small ceramic
capacitors, say something in the range from 100pF to 1000pF, should do
nicely. With the board turned over so the back is now facing you,
solder one side of a capacitor just a bit in-board from where you
soldered the diode anode for trace 2. Do the same for trace 3. Now
connect an RF choke (inductor) between the diode-capacitor junction
and the free lead of the new capacitor.

You can make things a bit more robust if you mount some sort of
terminal or pad on the back to solder this last junction to. One very
cheap but effective way to do it is to cut out some squares of PC
board material, maybe 5mm on a side, and glue them down to the large
board wherever you want an electrically floating terminal.

Just about done now! Just connect a wire from each of those last
capacitors (where wired to the inductors of course) to the + terminals
of the two 1.0mA meters, and return the meter - terminals to the board
back sides. Provide a case as you see fit, though it's usable without
a case; just be careful of the parts hanging off it.

After you build it, we can lead you through calibrating it, assuming
the earlier descriptions here weren't clear enough.

And I trust several lurkers will proof-read this and find all my
mistakes and places where I wrote LEFT when I meant RIGHT, etc. Oh,
and the meter connected to trace 2 will measure the power from left to
right in the original orientation; and the meter connected to trace 3
w2ill measure power from right to left.

Cheers,
Tom

The second problem is, if you want to implement a microstrip design,
how do you get the trace width right? If you're afraid of surface
mount parts, how will you control the trace width to +/- a fraction of
a millimeter? On 1.6mm thick PC board, assuming FR4 with a relative
dielectric constant of 4.75, you'd like to have a trace width about
2.78mm to get a 50 ohm line. If your trace is 3.5mm wide, you get a
bit under 44 ohms, and if your trace comes out 2.0mm wide, you get a
line that's almost 60 ohms. If you can do the PC board
photographically and have confidence that you can control the trace
width to within 0.1mm, that would work. If you're doing it by
scribing the copper and pulling up unwanted copper, I think you'll
have to be working under a pretty good microscope to get to much
closer than a mm of the desired width-- or maybe cut it on a milling
machine.

The problem that you have with FR4 is that the dielectric constant is not
well defined. Your assumption of 4.75 may be ok from one manufacturer and
one batch, but other examples can differ widely.

Jeff

Jim Lux wrote:
Heck, if you MUST use all analog designs and you're at less than 3GHz,
don't fool with diodes, use the less expensive, more sensitive, and
more accurate power measuring chips from Analog Devices.

Example: AD8310, DC-440MHz, 90+dB dynamic range (-91 to +4dBm) linear
to 0.4dB, stable over temp(-40 to +85) +/-1 dB

or the 8319, 1MHz to 10GHz, 40dB range, similar accuracy


N2PK has published a nice reflectometer design using two AD log chips as
detectors for forward and reflected power. The output voltage of each
detector is accurately proportional to log(power) over a very wide range
of applied power levels. Subtract the output voltages of the two
detectors in an op-amp, and you have a direct indication of return loss
in dB.

Details are at www.n2pk.com (better known as the home of the N2PK Vector
Network Analyser).

By the way, the Analog Devices samples service is also open to amateur
experimenters (by company policy) so by all means let's make use of it.
Many devices are only available in SMD - and many advanced RF devices
simply wouldn't function in a larger package - so as amateur
experimenters we have to bite the bullet and learn to handle SMD.

As Tom says, SMD can be the key to accurate measurements at VHF and UHF,
because the smaller packages have much lower parasitic inductance.

I wouldn't presume to tell Suzy (or anyone else) what they can or cannot
see and handle; but for many people, converting to SMD is mostly about
having the right equipment - a small soldering-iron tip, small-gauge
solder, a good pair of tweezers, and above all, some kind of optical
aid.

If they work for you, one of the best bargains would be a pair of very
strong half-moon reading glasses, worn as a "preamp" on top of whatever
eyeglasses you already wear. It's certainly worth wandering into the
drugstore to try some (pick the strongest they have). Otherwise, there
are many other sources of hands-free magnifying visors. Whoever you are,
see what will work for you.

For people who are acutely short-sighted, ignore most of the above and
simply take off your eyeglasses. At last, your day has come.

Under the magnifier, it's amazing how most people's hands become steady
and tremor-free. (That's in the absence of any medical condition that
can interrupt the eye-hand feedback loop - and also in the absence of a
much more common problem: too much caffeine!) We can't all be
neurosurgeons, but we can become good enough to handle SMD.

And then, as with most manual skills, it takes practice to become really
good at it. For occasional SMD work, it also takes practice to get back
into it after a long break.

The bottom line is that most of us CAN handle SMD... and if we don't
want to become stuck in the late 20th century, we're going to need to
develop that skill.


they also come in dual versions and versions with phase comparators..

The last time I looked, the device with the integrated phase comparator
had fallen off the regular samples service ("contact AD")... does anyone
know more?


--

73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

"Cecil Moore" wrote in message
...
K7ITM wrote:
See earlier posting in this thread.

Thanks Tom, when I said "linear power scale", I meant
e.g. a meter reading where 2000 watts is full scale
and 1000 watts is half scale. I have seen such meters
but not without a digital or analog computer on the
front end.
--
73, Cecil http://www.w5dxp.com

Just jumping in the middle of ths, but look at this watt meter.

http://bama.edebris.com/manuals/miltest/an-urm120/

It has a linear scale. I have one and it has a linear scale. Sort of made
like a Bird meter but much larger elements. It is just a diode and meter.
There is no power needed to run the meter except the sampled power comming
off the transmission line.

There were several versions made. One has a SWR scale on it. I am not sure
how the swr scale is but the wattmeter scale is linear instead of the log
looking scale of the Bird and most other meters.

"Ian White GM3SEK" wrote in message
...
Owen Duffy wrote:
Thanks Owen. BTW, what type of coax connector? Not PAL surely!

No, but they are not all that bad. Almost no one manufactured VHF land
mobiles here with UHF connectors, but they did use PAL (Belling & Lee)
once (Pye Reporters for instance).


Do Australian TVs use "Bloody Belling Lee" connectors like we still do in
the UK? What does "PAL" stand for?


--

73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

Certainly do Ian!

"K7ITM" wrote in message
...
On Jan 31, 1:51 pm, Owen Duffy wrote:
K7ITM wrote in news:9e844e58-a673-4ec0-9a0b-ec15f8cc8f30
@c4g2000hsg.googlegroups.com:





On Jan 31, 12:31 pm, Cecil Moore wrote:
K7ITM wrote:
To me, having a
linear power scale is a big advantage, because then you can
reasonably
accurately figure SWR without having to worry about temperature
compensation of the detectors.

Can you define what you mean by linear? Straight line?
Since we can only measure voltage and current, in order
to obtain a linear power scale from a linear meter, it
is necessary to supply some pre-display computing
ability (microcomputer).
--
73, Cecil http://www.w5dxp.com

See earlier posting in this thread. See various Avago ap notes, such
as AN 969. A diode detector run at low input provides an output DC
voltage that's a constant times the square of the input RF voltage.
If the input voltage is, or is assumed to be, at some constant
resistive load impedance, the DC output is linear with RF power
input. The proportionality is temperature dependent, but if two
detectors are constructed the same and run at the same temperature,
and run in the signal level region where that relationship holds, then
the ratio of the output DC voltages is a very good approximation of
the ratio of the input RF power levels, and thus is useful for finding
the SWR if the detectors are attached to the forward and reverse ports
of a good directional coupler. Top end of the useful "linear power"
range using an HSMS-2850 single diode detector is about 10mV DC
output. If you can measure the DC accurately down to 1uV (a bit
tough, given thermal emfs, but possible), that gives you about a
10000:1 power range, or 100:1 RF input voltage range -- or about
1.02:1 SWR. Chances are very good that a home-built coupler won't be
accurately enough matched to 50+j0 ohms to worry about anything that
low anyway, even if you had a reason to care about it.

Cheers,
Tom

Tom,

This is further from Suzy's needs, but...

Operation of a diode detector in the square law region isn't out of the
question, but it takes some serious gain to drive a meter. There are some
good chopper stabilised op amps out there that have uV offset levels and
single supply rail and input to below the negative rail eg LTC1050.

Another alternative is the AD8307AN log amps for a linear dBW scale. You
could even use one on FWD and REF detectors and difference the outputs in
an op amp for a direct indicating VSWR or RL scale. I have thought of
getting one of these chips and seeing whether its response is fast enough
to drive a PEP amplifier for SSB telephony.

Back to Suzy's problem...

The instrument downstream of the sampler is not so much the issue as
building and calibrating a sampler when you have no test gear.

Suzy, if you see a Revex W560 going on VKHAM for $100 or so, it is a good
buy. It has HF to 70cm (two independent couplers, ie four coax
connectors), and works pretty well.

For a dummy load, the market was flooded with terminations from 25W to
about 60W that had been scrapped from AMPS base station equipment, and
they were sold at hamfests for $20 or so, you may find them if you look
around.

Owen

Yes, there are several linear-in-dB RF detectors out there. Linear
Technology also have them. I really like that idea; they're typically
much more temperature stable than a diode detector. But Suzy wanted
to avoid SMT.

I've used a Harris chopper-stabilized op amp with HSMS-2850 zero-bias
detector diodes, and it works well, but I did learn something about
the need to be really careful around the chopper capacitor pins on
that op amp before getting it right... But it's also not difficult to
find a digital voltmeter that will go down to pretty low voltage at
high impedance. A 4.5 digit meter on a 200mV scale does ten
microvolts, and the simulation I ran last night suggests you could see
down to about -50dBm power level with that. When you get down to
10uV, you have to get serious about avoiding thermal emfs.

I suppose it makes sense to just drive the detector hard and run it
right into an analog meter movement, and then calibrate the meter.
Actually, at that level, the detector should be pretty linear in
voltage. That actually makes it easier to detect down closer to 1:1
SWR anyway.

I posted not too long ago about a load I made with four 200 ohm 2 watt
metal oxide resistors that shows what to me is remarkably good return
loss out to well beyond 450MHz. It was very cheap to make. But
there's no guarantee that some other brand of resistor would give such
good results. It may have just been a fluke that the one I made
turned out so good. (But I'm not about to toss it out!)

Cheers,
Tom

Well, hopefully Tom, I'll soon be the contented owner of a Bird Termaline
dummy load, so that will be a start...

"Ian White GM3SEK" wrote in message
...
Jim Lux wrote:
Heck, if you MUST use all analog designs and you're at less than 3GHz,
don't fool with diodes, use the less expensive, more sensitive, and more
accurate power measuring chips from Analog Devices.

Example: AD8310, DC-440MHz, 90+dB dynamic range (-91 to +4dBm) linear to
0.4dB, stable over temp(-40 to +85) +/-1 dB

or the 8319, 1MHz to 10GHz, 40dB range, similar accuracy


N2PK has published a nice reflectometer design using two AD log chips as
detectors for forward and reflected power. The output voltage of each
detector is accurately proportional to log(power) over a very wide range
of applied power levels. Subtract the output voltages of the two detectors
in an op-amp, and you have a direct indication of return loss in dB.

Details are at www.n2pk.com (better known as the home of the N2PK Vector
Network Analyser).

By the way, the Analog Devices samples service is also open to amateur
experimenters (by company policy) so by all means let's make use of it.
Many devices are only available in SMD - and many advanced RF devices
simply wouldn't function in a larger package - so as amateur
experimenters we have to bite the bullet and learn to handle SMD.

As Tom says, SMD can be the key to accurate measurements at VHF and UHF,
because the smaller packages have much lower parasitic inductance.

I wouldn't presume to tell Suzy (or anyone else) what they can or cannot
see and handle; but for many people, converting to SMD is mostly about
having the right equipment - a small soldering-iron tip, small-gauge
solder, a good pair of tweezers, and above all, some kind of optical aid.

If they work for you, one of the best bargains would be a pair of very
strong half-moon reading glasses, worn as a "preamp" on top of whatever
eyeglasses you already wear. It's certainly worth wandering into the
drugstore to try some (pick the strongest they have). Otherwise, there
are many other sources of hands-free magnifying visors. Whoever you are,
see what will work for you.

For people who are acutely short-sighted, ignore most of the above and
simply take off your eyeglasses. At last, your day has come.

Under the magnifier, it's amazing how most people's hands become steady
and tremor-free. (That's in the absence of any medical condition that can
interrupt the eye-hand feedback loop - and also in the absence of a much
more common problem: too much caffeine!) We can't all be neurosurgeons,
but we can become good enough to handle SMD.

And then, as with most manual skills, it takes practice to become really
good at it. For occasional SMD work, it also takes practice to get back
into it after a long break.

The bottom line is that most of us CAN handle SMD... and if we don't want
to become stuck in the late 20th century, we're going to need to develop
that skill.


they also come in dual versions and versions with phase comparators..

The last time I looked, the device with the integrated phase comparator
had fallen off the regular samples service ("contact AD")... does anyone
know more?


--

73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

The design from that link is for HF up to 30 MHz. I want to go to 435 MHz.
have a maggie lamp and am prepared to (reluctantly) try SMD after all, if I
have to -- in desperation.