BK Precision Frequency counter - does it work?

Hello,

I have purchased a BK Precision model 103 frequency counter that came with
an TA1 antenna. I thought that I would be able to measure signal strength as
well as signal frequency with this device (as advertised). However, even
when close to the transmitter or to the installed access point (the whole
WLAN is using 802.11g) there is no strength bar-graph displayed, but only
some meaningless, floating (at least for me) frequency numbers. Do I do
something wrong? Do I need a different antenna?
Thanks in advance for your answers...

On Wed, 14 Jul 2004 10:00:56 -0400, "Zoltan Lazar"
<(E-Mail Removed)> wrote:

>I have purchased a BK Precision model 103 frequency counter that came with
>an TA1 antenna. I thought that I would be able to measure signal strength as
>well as signal frequency with this device (as advertised). However, even
>when close to the transmitter or to the installed access point (the whole
>WLAN is using 802.11g) there is no strength bar-graph displayed, but only
>some meaningless, floating (at least for me) frequency numbers. Do I do
>something wrong? Do I need a different antenna?

Won't work. Spread spectrum means that the *FREQUENCY* is spread.
There is no single coherent signal to measure. On a spectrum
analyzer, it would look like a smear instead of a single strong signal
suitable for measuring. Even if you could get enough sensitivity to
get the counter going, you would have a constantly changing count
splattered over a 22MHz 802.11b/g bandwidth.

You also don't have enough sensitivity. Looking at the specs at:
http://www.bkprecision.com/www/np_specs.asp?m=103
the counter is <100mv at 2.4Ghz sensitive. If you literally plugged
the counter into the antenna port of the access point with a pigtail
and adapters, you might have a chance to see something. +15dBm is
what the access point outputs. Subtract -10dB for processing gain,
and we get +5dBm which is 3.6 milliwatts.
Volts = sqrt( Power * impedance) = sqrt ( 3.6*10^-3 * 50)
Volts = 425 mv
Yeah, that should get the display going, but you still won't have a
stable count. Reconnecting the antennas will give you less signal due
to coupling loss. I don't think you'll see anything on the counter.

Note that the maximum input power on the spec sheet is +15dBm. I
don't think you'll blow anything up by going over that limit, but
whatever prescaler is being used might have a problem. Dunno.

In any case, the frequency counter is not going to give you a useful
frequency count and might be useful as a signal strength meter a few
cm away.

So, what were you trying to accomplish with the counter?


--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

In article <(E-Mail Removed)>, Jeff Liebermann wrote:

>Won't work. Spread spectrum means that the *FREQUENCY* is spread.

Correct

>Note that the maximum input power on the spec sheet is +15dBm. I
>don't think you'll blow anything up by going over that limit, but
>whatever prescaler is being used might have a problem. Dunno.

I'd be cautious about the +15 dBm limit. It's likely there is an active
stage on the input that is setting this limit. Given the low price, it's
unlikly to have an input limiter dewice either.

>In any case, the frequency counter is not going to give you a useful
>frequency count and might be useful as a signal strength meter a few
>cm away.

Yeah, I computed the maximum range as about 0.43 feet or 13 centimeters

Old guy

On Fri, 16 Jul 2004 00:22:41 -0000, (E-Mail Removed) (Jean
Staehle) wrote:

>>Note that the maximum input power on the spec sheet is +15dBm. I
>>don't think you'll blow anything up by going over that limit, but
>>whatever prescaler is being used might have a problem. Dunno.

>I'd be cautious about the +15 dBm limit. It's likely there is an active
>stage on the input that is setting this limit. Given the low price, it's
>unlikly to have an input limiter dewice either.

Oh maybe. The B&K counter is actually a rebranded Optoelectronics M1
counter.
http://www.optoelectronics.com/m1.htm
I borrow a friends on occassion and have used at much higher power
levels than +15dBm. I think it uses an SAB6456A divide by 10
prescaler. This chip doesn't blow up with input levels over +15dBm.
Instead it leaks input signal all over the output and thus creates
some nasty jitter problems for the counter section. A 250MHz low pass
filter between the prescaler and the counter would have solved the
problem, but I didn't see one when I disassembled the M1 (to replace a
leaky NiCd battery).

>>In any case, the frequency counter is not going to give you a useful
>>frequency count and might be useful as a signal strength meter a few
>>cm away.

>Yeah, I computed the maximum range as about 0.43 feet or 13 centimeters
> Old guy

Ok, maybe not so useful. My version of the same bad idea was buying a
digital microwave oven leak detector.
http://www.comforthouse.com/comfort/micleakdetwi.html
I figured it could be used as some kind of wireless signal strength
meter. Well, it will detect the RF from a typical access point at
about 5cm maximum. It doesn't even have an antenna but uses just the
extra long leads on a diode. Bad idea.


--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831.336.2558 voice http://www.LearnByDestroying.com
# (E-Mail Removed)
# 831.421.6491 digital_pager (E-Mail Removed) AE6KS

In article <(E-Mail Removed)>, Jeff Liebermann wrote:


>I think it uses an SAB6456A divide by 10 prescaler.

I'm long retired and no longer have databooks, but google suggests the
SAB6456A is an ECL divide by 64/256. Assuming something comperable,
I still wouldn't like putting to much power in the input. Smoke is expensive.

> This chip doesn't blow up with input levels over +15dBm.
>Instead it leaks input signal all over the output and thus creates
>some nasty jitter problems for the counter section. A 250MHz low pass
>filter between the prescaler and the counter would have solved the
>problem, but I didn't see one when I disassembled the M1 (to replace a
>leaky NiCd battery).

Hmmm... I'd sorta expect the following stages would have such high loss
at frequencies that high that there'd be little consequence.

>>Yeah, I computed the maximum range as about 0.43 feet or 13 centimeters

>Ok, maybe not so useful. My version of the same bad idea was buying a
>digital microwave oven leak detector.
> http://www.comforthouse.com/comfort/micleakdetwi.html
>I figured it could be used as some kind of wireless signal strength
>meter. Well, it will detect the RF from a typical access point at
>about 5cm maximum. It doesn't even have an antenna but uses just the
>extra long leads on a diode. Bad idea.

Most of those devices are little more than a field strength meter - a small
antenna, a detector diode of some sort, maybe a "calibration" or sensitivity
adjustment pot, and a microamp meter. The fancier ones might have some form
of tuned circuit to select the desired signal. But they're really meant
for detecting signal levels measured in Watts. To get the sensitivity down
to the microwatt level means having some kind of amplification before the
detector.

Old guy

On Fri, 16 Jul 2004 16:38:30 -0000, (E-Mail Removed) (Jean
Staehle) wrote:

>In article <(E-Mail Removed)>, Jeff Liebermann wrote:
>>I think it uses an SAB6456A divide by 10 prescaler.
>I'm long retired and no longer have databooks, but google suggests the
>SAB6456A is an ECL divide by 64/256. Assuming something comperable,
>I still wouldn't like putting to much power in the input. Smoke is expensive.

So much for my photographic memory. I'll take apart the
Optoelectronics and see (after I borrow it again). Maybe my memory
isn't as far gone as I thought. Here's an article where someone
disected an M1 and 3000A and found a divide by 64 prescaler chip.
Hmmm....
http://virtual.xs4all.nl/software/qr...ary/oe_m1.html

>Hmmm... I'd sorta expect the following stages would have such high loss
>at frequencies that high that there'd be little consequence.

Sorta. What happens is that high levels of x10 (or x64?) signal on
the counter input cause a substantial uncertainty as to the zero
crossing point. It effectively smears the signal. It's a problem
common to all counters that use prescalers generally solved by a low
pass between the prescaler and the counter section. +15dBm (about
30mw) isn't gonna blow up anything except perhaps an uprotected GAsFet
front end. I could lookup the maximum input levels to the prescaler
chip, but I wanna get the right chip number first.

>Most of those devices are little more than a field strength meter - a small
>antenna, a detector diode of some sort, maybe a "calibration" or sensitivity
>adjustment pot, and a microamp meter.

Diode detector driving a DC amplifier and a digital display. What's a
"pot"? This is the microprocessor generation where everything is done
in software. Push the button and it "self calibrates" which is a nice
term of compensates for radical temperature drift in the cheap
circuitry.

>The fancier ones might have some form
>of tuned circuit to select the desired signal. But they're really meant
>for detecting signal levels measured in Watts.

I should hope not. If my microwave oven leaked watts, I would be
seriously worried about health effects. The limit is about 5 mW/cm^2
at about 2" from the door. That's milliwatts, not watts. Range on
the leakage detector is 0-9.99 mW/cm^2 with an alleged accuracy of +/-
1dB (about 10%).

>To get the sensitivity down
>to the microwatt level means having some kind of amplification before the
>detector.

If you don't care about DC and temp drift, you can do it with any high
electron mobility diode detector and lots of DC gain (probably with a
differential instrumentation amp). If you want, I'll tear it apart
again and post some photos. I gotta justify my overpriced digital
camera purchases with something useful. It was fairly disgusting
construction inside the last time I checked.

Incidentally, I planted it next to my Linksys BEFW11S4 access point,
moved some traffic, and it displays 0.4 mW/cm^2 at about 2". So much
for hazardous radiation from 802.11.


--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

In article <(E-Mail Removed)>, Jeff Liebermann wrote:

>So much for my photographic memory.

I think it's the second or third thing to go, but can't remember ;-)

> I'll take apart the
>Optoelectronics and see (after I borrow it again). Maybe my memory
>isn't as far gone as I thought. Here's an article where someone
>disected an M1 and 3000A and found a divide by 64 prescaler chip.
>Hmmm....
> http://virtual.xs4all.nl/software/qr...ary/oe_m1.html

I'll have a look.

>>Hmmm... I'd sorta expect the following stages would have such high loss
>>at frequencies that high that there'd be little consequence.
>
>Sorta. What happens is that high levels of x10 (or x64?) signal on
>the counter input cause a substantial uncertainty as to the zero
>crossing point. It effectively smears the signal. It's a problem
>common to all counters that use prescalers generally solved by a low
>pass between the prescaler and the counter section.

What I'd expect is that the printed circuit is far from an ideal
transmission line (yes, I know about stripline and microstrip), and
this should offer lots of loss. When you get inside the "low
frequency" counter, there are likely some Schottky diodes but the noise
margin should be quite large, further reducing strangeness.

> +15dBm (about
>30mw) isn't gonna blow up anything except perhaps an uprotected GAsFet
>front end. I could lookup the maximum input levels to the prescaler
>chip, but I wanna get the right chip number first.

Ignoring VSWR, +15 dBm is 1.26 Vrms - and you _know_ that the mismatch
is going to put a "max" at the wrong place. I will admit that a decent
design should have some overload protection.

>>Most of those devices are little more than a field strength meter - a small
>>antenna, a detector diode of some sort, maybe a "calibration" or sensitivity
>>adjustment pot, and a microamp meter.
>
>Diode detector driving a DC amplifier and a digital display. What's a
>"pot"? This is the microprocessor generation where everything is done
>in software. Push the button and it "self calibrates" which is a nice
>term of compensates for radical temperature drift in the cheap
>circuitry.

Well, yeah, but I was describing the field strength meter you built
yourself - the expensive part is the meter (out of the "junk" box, or
purchased at a swap meet).

>>The fancier ones might have some form
>>of tuned circuit to select the desired signal. But they're really meant
>>for detecting signal levels measured in Watts.
>
>I should hope not. If my microwave oven leaked watts, I would be
>seriously worried about health effects.

Again - I was talking about the field strength meter, and referring to
the transmitter power. I do agree about the health concerns.

> The limit is about 5 mW/cm^2
>at about 2" from the door. That's milliwatts, not watts. Range on
>the leakage detector is 0-9.99 mW/cm^2 with an alleged accuracy of +/-
>1dB (about 10%).

And how often calibrated? 1 dB is pretty good accuracy.

>>To get the sensitivity down
>>to the microwatt level means having some kind of amplification before the
>>detector.
>
>If you don't care about DC and temp drift, you can do it with any high
>electron mobility diode detector and lots of DC gain (probably with a
>differential instrumentation amp).

The old solution to that was a chopper amp. But I'd probably go with both
RF and DC gain. Mini-Circuits Lab (www.minicircuits.com) used to build a
drop in amplifier suitable for stripline with about 20 dB of gain that
were pretty cheap. Like I say, I no longer have my data boks, but I seem
to recall them doing 10 MHz - 3 GHz in a package about the size of of a
TO-236 cased transistor (0.2 inch diameter 'pill' actually).

> If you want, I'll tear it apart
>again and post some photos. I gotta justify my overpriced digital
>camera purchases with something useful. It was fairly disgusting
>construction inside the last time I checked.

Not built by HP or Tek, huh? ;-)

>Incidentally, I planted it next to my Linksys BEFW11S4 access point,
>moved some traffic, and it displays 0.4 mW/cm^2 at about 2". So much
>for hazardous radiation from 802.11.

I think that's the _least_ of the problems.

>Jeff Liebermann (E-Mail Removed)
>150 Felker St #D http://www.LearnByDestroying.com
>Santa Cruz CA 95060 AE6KS 831-336-2558

Hmmm, isn't that just off Ocean Street at highway 1? This dates me, but
I used to occassionally fly into Sky Park, up near Scott's Valley. I think
it was converted to a shopping center or something back in the 1980s.

Old guy

On Fri, 16 Jul 2004 22:45:42 -0000, (E-Mail Removed)
(Moe Trin) wrote:

>>The limit is about 5 mW/cm^2
>>at about 2" from the door. That's milliwatts, not watts. Range on
>>the leakage detector is 0-9.99 mW/cm^2 with an alleged accuracy of +/-
>>1dB (about 10%).

Here's some photos. The detector diode is labeled D1 and is on the
left. No strip line, no antenna, no calibration, no kidding.

http://www.LearnByDestroying.com/pic...tor/index.html
I didn't notice the funny looking trace jogs above and below D1.
That's a good indication that the board was either taped by hand
(where's my Brady and Celotape collection?) or butchered by hand on
the negative. How neanderthal. Anyway, what you're looking at is
state-o-de-art consumer RF design.

>And how often calibrated? 1 dB is pretty good accuracy.

No problem. Everybody lies, but that's ok because nobody listens.

>The old solution to that was a chopper amp. But I'd probably go with both
>RF and DC gain. Mini-Circuits Lab (www.minicircuits.com) used to build a
>drop in amplifier suitable for stripline with about 20 dB of gain that
>were pretty cheap. Like I say, I no longer have my data boks, but I seem
>to recall them doing 10 MHz - 3 GHz in a package about the size of of a
>TO-236 cased transistor (0.2 inch diameter 'pill' actually).

Actually, I'm old enough to remember packaging prototype transistors
in .38 and .45 caliber shells, crimping the ends, filling the shell
with nitrogen, and soldering them shut. Incidentally, the TO in TO-5
originally stood for Texas and Oklahoma, home of TI.

A chopper stabilized amp would certainly give the sensitivity and DC
stability. It would also be sensitive enough to detect signals from
alien civilizations. However, they don't integrate well thanks to
clock leakage problems and are serious overkill for a $30 leakage
detector. Certainly a precision instrument, but not this piece of
[insert derrogatory explative] junk.

>Not built by HP or Tek, huh? ;-)

Both Tektronix and HP (Agilent) have changed in recent years.
Tektronix is making scopes that are unrepairable and use unobtainable
modules. The service manual documentation is pablumified and useless.
Even the voltages and waveforms are missing in action. Read the
comments in sci.electronics.repair on undocumented and unrepairable
scopes. No more gold plated printed circuit boards from HP (Agilent).
I'm not sure what HP is doing lately because I can't afford to even
borrow any of their new boxes. Incidentally, I collect HP LED
calculators. I think I have about 30 of them.

>Hmmm, isn't that just off Ocean Street at highway 1? This dates me, but
>I used to occassionally fly into Sky Park, up near Scott's Valley. I think
>it was converted to a shopping center or something back in the 1980s.

Yep, that's my palatial office.
http://www.LearnByDestroying.com/panorama/office.jpg (283KB)
Sky Park airport closed in Jan 1983. I was there to watch the local
pilots buzz the field on the last day. It was quite a zoo. It's now
a park and housing development. I certainly miss the airport although
I can't say the suicidal right hook approach over the sand quary will
be missed.
http://www.santacruzpl.org/history/trans/airports.shtml


--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831.336.2558 voice http://www.LearnByDestroying.com
# (E-Mail Removed)
# 831.421.6491 digital_pager (E-Mail Removed) AE6KS