Maker Pro
Maker Pro

Field Strength Meter (LTspice)

J

John Woodgate

Jan 1, 1970
0
I read in sci.electronics.design that Paul Burridge
[email protected]>) about 'Field Strength Meter (LTspice)', on Sun, 28 Dec
2003:
ISTR somewhere that this *is* the preferred way of doing it. Saves on
RF transistors if nothing else! However one assumes at least *some*
degree of RF pre-amplification is needed to get a potentially very
weak signal beyond the knee of the diode characteristic.

There ISN'T really a knee. It only appears that there is because of the
way the I/V curve is plotted. If you plot log(I) against V you get
nearly a straight line, with nothing unusual happening at V = 0.6 to 0.7
V.
At low signal levels, a diode detector (or any rectifying junction) has
a *square law* transfer characteristic. If you put a voltage-doubler
diode rectifier (the two-diode circuit you've been shown by me and Jim)
at the *input* of an op-amp d.c. amplifier with your meter at the op-amp
output (through a series resistor), you'd have a working circuit.
 
P

Paul Burridge

Jan 1, 1970
0
[...]

What's the function of L1 and L2?
How do I plot the voltage of Node8 WRT Node7 in LTspice? All I've ever
done
before is reference voltages against ground.
Are q3 and q4 acting as some variation of the long-tailed pair config?
What is the relevance of modelling the meter coil as having 10u of
inductance when the voltage across it is DC?

If these questions could be contained in a general description of how
the complete circuit
works it would be *most* helpful.

Thanks.
 
P

Paul Burridge

Jan 1, 1970
0
There ISN'T really a knee. It only appears that there is because of the
way the I/V curve is plotted. If you plot log(I) against V you get
nearly a straight line, with nothing unusual happening at V = 0.6 to 0.7
V.

I don't doubt it, since the linear plot gives an exponential curve. So
you're saying diodes in the real world work in such a way as logs
prevail and there's no knee? There was me thinking log plotting was a
simply a human contrivance to save us having to use enormously long
graph paper. :)
At low signal levels, a diode detector (or any rectifying junction) has
a *square law* transfer characteristic. If you put a voltage-doubler
diode rectifier (the two-diode circuit you've been shown by me and Jim)
at the *input* of an op-amp d.c. amplifier with your meter at the op-amp
output (through a series resistor), you'd have a working circuit.

Yeah, I guess it might work but probably pick up everything within 50
miles. 50 yards would suffice!
 
F

Fred Bloggs

Jan 1, 1970
0
Roger said:
A chip called LM3189 (or 3089) can be used to build a nice field
strength meter.

Use the first half of the chip and take the output signal from pin 13
to a meter through a suitable resistor, as described in the National
Semiconductor datasheet for these circuits.

It gives indication from 10uVolt signal to around 1Volt signal.

It may be a bit too ambitious a project for mr. Burridge.
Im am not sure how easy it is to find this chip either.

But he should absolutely get the Handbook for radio amateurs from
ARRL. It is cheap to buy (considering its size and the enormous amount
of useful information) and it is available in most public libraries.

It contains, among all other things, descriptions of all kinds of DIY
measuring equipment related to radio and general electronics, and it
is written for people who don't have a lot of background knowledge in
electronics. The text is easy to understand and it has well checked
and tested circuits.

Excellent! The meter drive output will provide 100dB of measurement
range- quite a bit better than the 20dB range of that simple '3904 ckt.
The front end will probably have to be mixed down to ~10.7MHz but that
is a good standard IF frequency for readily available filters and
coupled inductors.
 
F

Fred Bloggs

Jan 1, 1970
0
Kevin said:
But not necessarily as relevant as might be initially assumed. The
common base circuit is a bit more subtle than this. A fundamental reason
for the common base is the lower resistance at the emitter. This results
in a re.cbe limit, not an Rs.ccb limit (simplified). There is still the
same large gain from the emitter to collector, which still results in
miller gain to any capacitance from emitter to collector. This
capacitance may be less than the base-collector capacitance, but it is
still Miller amplified, but in this situation, the source resistance is
in || to rbe frequency response wise, so is not usually significant to
response. If the source resistance is low, Miller capacitance is not
relevant.

Hmmm- there is no Miller effect in the common-base configuration, the
base is signal ground and therefore the two junction capacitances are to
ground. The stray Cce results in a negative capacitance contribution to
the input, aka positive feedback, since I=(Ve-Vc)jwCce and Vc=+AvVe, you
have I=Ve(1-Av)jwCce or Cin=(1-Av)Cce a negative capacitance effect.
 
J

Jim Thompson

Jan 1, 1970
0
Hmmm- there is no Miller effect in the common-base configuration, the
base is signal ground and therefore the two junction capacitances are to
ground. The stray Cce results in a negative capacitance contribution to
the input, aka positive feedback, since I=(Ve-Vc)jwCce and Vc=+AvVe, you
have I=Ve(1-Av)jwCce or Cin=(1-Av)Cce a negative capacitance effect.

He! He! He! Kevin got caught in-error again ;-)

...Jim Thompson
 
R

Roger Johansson

Jan 1, 1970
0
Fred Bloggs said:
Excellent! The meter drive output will provide 100dB of measurement
range- quite a bit better than the 20dB range of that simple '3904 ckt.
The front end will probably have to be mixed down to ~10.7MHz but that
is a good standard IF frequency for readily available filters and
coupled inductors.

These input sections are not tuned in any way. It is up to the user to
put filters before this section, and after it.

I heard that these IF stages work from DC to well over 30 MHz, and I
see nothing in the schematic which contradicts this.
These chips have been used as field strength meters for 0-30MHz, if I
recall correctly.

One can of course put a variable or fixed tuning section before the
input, if a certain frequency range of a certain frequency is of
interest.
 
A

Active8

Jan 1, 1970
0
Active8 wrote:

On Sat, 27 Dec 2003 21:50:19 GMT, [email protected] said...

Excellent!

P.S. Paul, go ahead and lay out Analog's circuit as a PC board.
It's a winner.

Sorry, I've only just found how to get LTS to recognise this but
it's fine now. There are a couple of questions I have but I'll run
a few sims first and see how it performs. I wonder how the hell he
came up with this design? It looks pretty unconventional to my
inexperienced eyes!

It's as conventional as they come. Common base configurations were
used a lot in the early days of transistors when you couldn't just
go out and buy a better transistor. Hell, there were only three or
four different transistors at all in the beginning. You needed to
push the transistors as hard as you could to get them to work at
high frequencies. I remember GE taking out a large magazine ad for
one of their improved transistors that would actually oscillate all
the way up to the top end of the AM band at 1.5 megacycles.

Jim



You all have failed to mention one key feature of this design. The
common base config reduces the Miller effect capacitance.



But not necessarily as relevant as might be initially assumed. The
common base circuit is a bit more subtle than this. A fundamental reason
for the common base is the lower resistance at the emitter. This results
in a re.cbe limit, not an Rs.ccb limit (simplified). There is still the
same large gain from the emitter to collector, which still results in
miller gain to any capacitance from emitter to collector. This
capacitance may be less than the base-collector capacitance, but it is
still Miller amplified, but in this situation, the source resistance is
in || to rbe frequency response wise, so is not usually significant to
response. If the source resistance is low, Miller capacitance is not
relevant.

Hmmm- there is no Miller effect in the common-base configuration, the
base is signal ground and therefore the two junction capacitances are to
ground. The stray Cce results in a negative capacitance contribution to
the input, aka positive feedback, since I=(Ve-Vc)jwCce and Vc=+AvVe, you
have I=Ve(1-Av)jwCce or Cin=(1-Av)Cce a negative capacitance effect.[/QUOTE]

He! He! He! Kevin got caught in-error again ;-)

...Jim Thompson[/QUOTE]

Does this make me a troll? ;)

HH,
Mike
 
A

Active8

Jan 1, 1970
0
I don't doubt it, since the linear plot gives an exponential curve. So
you're saying diodes in the real world work in such a way as logs
prevail and there's no knee? There was me thinking log plotting was a
simply a human contrivance to save us having to use enormously long
graph paper. :)

Heh, heh. It does. It's an added bonus. Think about it this way.
deciBels are a logarithmic function of a ratio. Plotting a log
function on semi-log paper gives you a straight line. Plotting log
freq against decibels on log-log paper results in a straight line
for filter slopes.

Ok how many have log "paper". I still have a tablet of Smith charts
somewhere.
Yeah, I guess it might work but probably pick up everything within 50
miles. 50 yards would suffice!

That's what the series diode is for. Current limiting. Scaling. A R
in parallel with the meter scales voltage.

Mike
 
J

Jim Thompson

Jan 1, 1970
0
On Sun, 28 Dec 2003 19:39:54 GMT, Active8

[snip]
Ok how many have log "paper". I still have a tablet of Smith charts
somewhere. [snip]

I have still have some multi-cycle log and log-log paper around (and
Smith Charts).

At one point in time I had copied some paper patterns into hp printer
language so I could just print a sheet when I needed it, but I can't
locate them now... probably on an archive CD... in a fit of
housekeeping I copied all my old 5-1/4" and 3-1/2" stuff from DOS days
onto CD's, making it now harder to locate old stuff :-(

...Jim Thompson
 
A

Active8

Jan 1, 1970
0
[...]

What's the function of L1 and L2?
How do I plot the voltage of Node8 WRT Node7 in LTspice? All I've ever
done
before is reference voltages against ground.
Are q3 and q4 acting as some variation of the long-tailed pair config?
What is the relevance of modelling the meter coil as having 10u of
inductance when the voltage across it is DC?

If these questions could be contained in a general description of how
the complete circuit
works it would be *most* helpful.

Thanks.
Looks like D1 protects the input Q against reverse Vbe from strong
signals though I've usually seen back to back diodes for that.
Ditto for D2 but it's also part of a doubling detector. Look at
what JT gave you and redraw this new circuit substituting a diode
for Q3. Cool, eh?

L1, L2 for DC bias while blocking RF. C1-C3 for DC blocking.

Q3 in emitter follower since you want to drive a low impedance with
current. It gives current gain, not voltage gain. D3 as a detector
if you want to call it that. It makes the meter read DC current.
Pulsating DC, mind you, C5 provides filter action to fix that and
it bypasses R5.

C6 bypasses R6.

Q4 is a NULL adjust to zero yer meter, i.e., make nodes 7 & 8
equal. C4 keeps stray RF from messing with Q4.


WHere do you get 2.8V batteries or is that a Thevinin equiv? Maybe
it's a couple of 1.5V batts run down. I'd want to use rechargables
and NiCd's are 1.25V each or 2.5 per pair. Oughtta still work.

HH,
Mike
 
A

Active8

Jan 1, 1970
0
On Sun, 28 Dec 2003 19:39:54 GMT, Active8

[snip]
Ok how many have log "paper". I still have a tablet of Smith charts
somewhere. [snip]

I have still have some multi-cycle log and log-log paper around (and
Smith Charts).

At one point in time I had copied some paper patterns into hp printer
language so I could just print a sheet when I needed it,

those would be cool to have. I think one of those Gerber Viewer
sites has a HPGL print util.
but I can't
locate them now... probably on an archive CD... in a fit of
housekeeping I copied all my old 5-1/4" and 3-1/2" stuff from DOS days
onto CD's, making it now harder to locate old stuff :-(

...Jim Thompson
I've done that too, but I still have MIMP from Mot on 5-1/4"
somewhere in storage. Remind me to use it for a frisbee. Even in
the incomplete stage, my smith app is easier to use.

HH,
Mike
 
J

Jim Thompson

Jan 1, 1970
0
On Sun, 28 Dec 2003 19:39:54 GMT, Active8

[snip]
Ok how many have log "paper". I still have a tablet of Smith charts
somewhere.
[snip]
Mike

I have still have some multi-cycle log and log-log paper around (and
Smith Charts).

At one point in time I had copied some paper patterns into hp printer
language so I could just print a sheet when I needed it,

those would be cool to have. I think one of those Gerber Viewer
sites has a HPGL print util.
but I can't
locate them now... probably on an archive CD... in a fit of
housekeeping I copied all my old 5-1/4" and 3-1/2" stuff from DOS days
onto CD's, making it now harder to locate old stuff :-(

...Jim Thompson
I've done that too, but I still have MIMP from Mot on 5-1/4"
somewhere in storage. Remind me to use it for a frisbee. Even in
the incomplete stage, my smith app is easier to use.

HH,
Mike

I found the following sites:

http://www.mathematicshelpcentral.com/graph_paper.htm

http://sorrel.humboldt.edu/~geodept/geology531/graph_paper_index.html

and, print your own...

http://perso.easynet.fr/~philimar/graphpapeng.htm

.... but realistically all my CAD programs create whatever is needed so
I don't plot much of anything by hand anymore.

...Jim Thompson
 
J

John Woodgate

Jan 1, 1970
0
I read in sci.electronics.design that Active8 <mTHISREMOVEcolasono@earth
Ok how many have log "paper". I still have a tablet of Smith charts
somewhere.

You don't need log paper. Use squared or linear graph paper and make
ten divisions equal 1 decade on your log scale. Then 2 is at 3
divisions, 3 is at 4.77 divisions, 4 is at 6 divisions, 5 is at 7
divisions, 6 is at 7.78 divisions, 7 is at 8.45 divisions, 8 is at 9
divisions and 9 is at 9.54 divisions. You may not be able to mark the
points accurately to 2 places of decimals, but you can get quite near
enough with a thin pencil.
[snip]

That's what the series diode is for. Current limiting. Scaling. A R
in parallel with the meter scales voltage.

In parallel??
 
P

Paul Burridge

Jan 1, 1970
0
That's what the series diode is for. Current limiting. Scaling. A R
in parallel with the meter scales voltage.

Thanks for the clarification, Mike. That possibility *did* occur to me
about one-tenth of a second after I'd hit the 'send' key. Do'h! :-(
 
P

Paul Burridge

Jan 1, 1970
0
Looks like D1 protects the input Q against reverse Vbe from strong
signals though I've usually seen back to back diodes for that.
Ditto for D2 but it's also part of a doubling detector. Look at
what JT gave you and redraw this new circuit substituting a diode
for Q3. Cool, eh?

L1, L2 for DC bias while blocking RF. C1-C3 for DC blocking.

Q3 in emitter follower since you want to drive a low impedance with
current. It gives current gain, not voltage gain. D3 as a detector
if you want to call it that. It makes the meter read DC current.
Pulsating DC, mind you, C5 provides filter action to fix that and
it bypasses R5.

C6 bypasses R6.

Q4 is a NULL adjust to zero yer meter, i.e., make nodes 7 & 8
equal. C4 keeps stray RF from messing with Q4.


WHere do you get 2.8V batteries or is that a Thevinin equiv? Maybe
it's a couple of 1.5V batts run down. I'd want to use rechargables
and NiCd's are 1.25V each or 2.5 per pair. Oughtta still work.

Thanks a lot, Mike. Did all this just sort of 'materialise' from your
looking at the circuit or do you have some prior familiarisation with
this kind of configuration? I'd really like to be able to acquire an
intuitive grasp of such circuits by cursory inspection if that can be
done.
 
A

Active8

Jan 1, 1970
0
On Sun, 28 Dec 2003 19:39:54 GMT, Active8

[snip]

Ok how many have log "paper". I still have a tablet of Smith charts
somewhere.

[snip]
Mike


I have still have some multi-cycle log and log-log paper around (and
Smith Charts).

At one point in time I had copied some paper patterns into hp printer
language so I could just print a sheet when I needed it,

those would be cool to have. I think one of those Gerber Viewer
sites has a HPGL print util.
but I can't
locate them now... probably on an archive CD... in a fit of
housekeeping I copied all my old 5-1/4" and 3-1/2" stuff from DOS days
onto CD's, making it now harder to locate old stuff :-(

...Jim Thompson
I've done that too, but I still have MIMP from Mot on 5-1/4"
somewhere in storage. Remind me to use it for a frisbee. Even in
the incomplete stage, my smith app is easier to use.

HH,
Mike

I found the following sites:

http://www.mathematicshelpcentral.com/graph_paper.htm

http://sorrel.humboldt.edu/~geodept/geology531/graph_paper_index.html

and, print your own...

http://perso.easynet.fr/~philimar/graphpapeng.htm

... but realistically all my CAD programs create whatever is needed so
I don't plot much of anything by hand anymore.

...Jim Thompson

That's what I'm talkin' about. Between PSpice, MathCAD, and
whatever my app with be named, there's not much need. It's probably
cheaper to buy quad paper than use toner or ink, but some log paper
would be handy for the concept phase. Polar paper. Cool.

Thanks, those links are great! Music score paper, too.

HH,
Mike
 
A

Active8

Jan 1, 1970
0
I read in sci.electronics.design that Active8 <mTHISREMOVEcolasono@earth
Ok how many have log "paper". I still have a tablet of Smith charts
somewhere.

You don't need log paper. Use squared or linear graph paper and make
ten divisions equal 1 decade on your log scale. Then 2 is at 3
divisions, 3 is at 4.77 divisions, 4 is at 6 divisions, 5 is at 7
divisions, 6 is at 7.78 divisions, 7 is at 8.45 divisions, 8 is at 9
divisions and 9 is at 9.54 divisions. You may not be able to mark the
points accurately to 2 places of decimals, but you can get quite near
enough with a thin pencil.
[snip]

That's what the series diode is for. Current limiting. Scaling. A R
in parallel with the meter scales voltage.

In parallel??
I'm backwards. Series scales voltage. || scales current for current
measurements, and the combination thereof, IIRC and my mind isn't
FUBAR'd from the latest mini-crisis, scales voltage and matches the
meter to the measurement app for lack of a better way to put it.

HH,
Mike
 
A

Active8

Jan 1, 1970
0
Thanks a lot, Mike. Did all this just sort of 'materialise' from your
looking at the circuit or do you have some prior familiarisation with
this kind of configuration?

I looked at the circuit and figured it out based on what I know of
RF design. The differential pair looking thing was easy enough to
figure out, though I can't describe or quantize the near FTL right
brained parallel thought processes involved.

I've never seen common base amps used in a meter like this, in
fact, the sniffer and the simple 1 transistor active ( and maybe
one with an OpAmp rather than a transistor) sniffer are the closest
I've seen for relative FS measuurement.
I'd really like to be able to acquire an
intuitive grasp of such circuits by cursory inspection if that can be
done.

It can. BTDT before. It's sometimes called "designer dissapeared
and left me this POS." Also known as "building on the work of
others", "embrace and extend" (MS parlance), and copying.

I received a bad electric shock before I was old enough to
remember, another not long after that that I *do* remember. Rigged
two dry cell batts together with and old telephone speaker and mic
when I was six along with the obligatory wire around a nail e-
magnet and Captain Crunch Microphone (see "electronics for 6 year
olds".) Tore apart radios. Plenty of old junk like door bells (the
single metal bar gong and solenoid) make-break bells (vibrators, I
guess,) and phones to tear apart. Found the 1st electronics book in
the school library in 6th grade. Then the magazine sub started. Don
Lancaster was around back then. At the time, I lived next to a
lighting technician for Haji (later Hybrid Ice. They disbanded and
if rumor has it correct, the old stage manager is bringing them
back together. Haven't seen him in months. He wrote some songs.
Stutters like mad especially if you get him excited, but he says he
can sing. I hope they make lots of money so I can charge them a lot
for custom designs and whatever. Why compete with the Japs?)

Anyway, Neil, the lighting tech, had a physics degree. He turned me
on to algebra to make ohms law more useful and poly-paks to make
parts aquisition easier. He and his wife took me and sometimes mom
to concerts and he and the other techs explained things like
dimmers, how to clamp onto house power at the box, tape loops, and
digital light sequencers (johnson counter / triac crap).

So if you've only been at it for a while don't stress too hard, I
guess. Kids have gone to college not knowing shit about electronics
and probably don't when they get out. I know a tech with a BET that
can't design a 2 resistor divider. Pretty sad.

I suppose experience will help, but there's that right brain thing
that helps. Fear not, I think it can be developed. Supposedly the
bicameral mind can be link enhanced.

Right now, I need a circuit to rectify certain people. I think I'll
call it a cloaking device. Maybe fake my death.

dahdidahdi dahdahdidah Calling all crackers.

HH,
Mike
 
J

Jim Thompson

Jan 1, 1970
0
On Sun, 28 Dec 2003 23:05:39 GMT, Active8

[snip]
Right now, I need a circuit to rectify certain people. I think I'll
call it a cloaking device. Maybe fake my death.

dahdidahdi dahdahdidah Calling all crackers.

HH,
Mike

didididi didi dahdah didi dahdidah di

(Ex K7ZAE, last used 1968 on two meters)

...Jim Thompson
 

Similar threads

L
Replies
4
Views
2K
Lord Snooty
L
P
Replies
20
Views
3K
John Woodgate
J
P
Replies
61
Views
4K
Frithiof Andreas Jensen
F
J
Replies
0
Views
991
John Smith
J
J
Replies
7
Views
2K
Paul Burridge
P
Top