EMRFD Message Archive 1327
Message Date From Subject 1327 2008-01-14 17:56:54 Chuck Adams uV signal measurements
The MAR-1 has a spec sheet gain averaging around
18dB in the HF range. This gives me a gain of
36dB before the load resistor R1 (51 ohms).
With 36dB and 50uV in, I get 199mV out from the
second MAR-1 (if I've done the math right and
didn't forget how to use the calculator). I have
a 200uA meter that looks like 147mV full scale
deflection.
If I change the .01 blocking caps to 0.22
in order to lower the low end cutoff frequency,
my question is:
1. Will removal of one of the .01 and R1 (51ohms)
adversely effect my newly configured voltmeter?
I'm not worried so much about impedance matching
at this point.
I'm interested in leaving U3 in and see just how low
I can go on the input. I am interested in sub 50uV
measurements.
In a web search, very little seems to appear on uV
measurements, especially on online circuits. I am
interested, if there is a place Phoenix, in a place
that I can walk into and lay down the cash for a
Keithley or HP uV RF meter as a calibration and
sanity check.
I'd be interested if any one has a SPICE model
for the MMIC part.
I just may wire up U1 and U2 combo and input
a 50uV signal from the Elecraft 50uV source
and see what the spectrum analyzer shows
for fun.
TIA
Chuck
--
Chuck Adams, K7QO
k7qo@commspeed.net
http://www.k7qo.net/
Moving to Arizona? Bring your own water, please.1328 2008-01-14 18:22:51 Robert Cerreto Re: uV signal measurements
Not sure where you are going Chuck.
73, WA!FXT
Chuck Adams <k7qo@commspeed.net> wrote:
Looking a the EMRFD book, page 7.7, Figure 7.12.
The MAR-1 has a spec sheet gain averaging around
18dB in the HF range. This gives me a gain of
36dB before the load resistor R1 (51 ohms).
With 36dB and 50uV in, I get 199mV out from the
second MAR-1 (if I've done the math right and
didn't forget how to use the calculator). I have
a 200uA meter that looks like 147mV full scale
deflection.
If I change the .01 blocking caps to 0.22
in order to lower the low end cutoff frequency,
my question is:
1. Will removal of one of the .01 and R1 (51ohms)
adversely effect my newly configured voltmeter?
I'm not worried so much about impedance matching
at this point.
I'm interested in leaving U3 in and see just how low
I can go on the input. I am interested in sub 50uV
measurements.
In a web search, very little seems to appear on uV
measurements, especially on online circuits. I am
interested, if there is a place Phoenix, in a place
that I can walk into and lay down the cash for a
Keithley or HP uV RF meter as a calibration and
sanity check.
I'd be interested if any one has a SPICE model
for the MMIC part.
I just may wire up U1 and U2 combo and input
a 50uV signal from the Elecraft 50uV source
and see what the spectrum analyzer shows
for fun.
TIA
Chuck
--
Chuck Adams, K7QO
k7qo@commspeed.net
http://www.k7qo.net/
Moving to Arizona? Bring your own water, please.
---------------------------------
Looking for last minute shopping deals? Find them fast with Yahoo! Search.
[Non-text portions of this message have been removed]1330 2008-01-14 19:11:26 Allison Parent Re: uV signal measurements 1331 2008-01-14 19:11:58 Chuck Adams Re: uV signal measurements
> Why remove R1?? I thought the MAR specs were valid only with a 50 ohm load?That was one of the reasons for the question. Just how important
> Without the 50 ohm load, will the MAR break into oscillations.?
>
> Not sure where you are going Chuck.
>
> 73, WA!FXT
>
was R1? With D2 being the load for 1/2 a cycle at such a low
voltage I'd have to figure the effective R_L value.
I'll experiment with the circuit with several possible configurations
and see what I get. The research of the literature just wasn't giving
me enough information to see what people had done with
such circuits. I admit ignorance, but I did read through sections
of about 10-15 books to see if there was any detailed information.
I'm hoping that some one in this group had played with uV levels.
I have the old article from Ham Radio using several MC1350 IF
amps to make a sensitive voltmeter, but am hoping for simplicity.
Hope this helps explain my silly question.
Thanks Bob,
Chuck
--
Chuck Adams, K7QO
k7qo@commspeed.net
http://www.k7qo.net/
Moving to Arizona? Bring your own water, please.1332 2008-01-14 20:01:25 kerrypwr Re: uV signal measurements
without the 50 ohm on its output.
It would be like connecting an end-fed 1/2-wave antenna to the 50-ohm
output of your transmitter.
But I also don't see what you want to achieve; the instrument will
still have 50-ohms impedance at the input socket (unless that is
disturbed by the absence of the resistor downstream; MMICs are
feedback amplifiers and the output impedance can affect the input
impedance via the feedback network).
A voltmeter with 50 ohms input resistance is no voltmeter at all;
voltmeters have high impedance to avoid disturbing the circuit under test.
Ten megohms is easy to achieve in DC voltmeters with resistors; I can
testify from recent experience that it's difficult-to-impossible to
achieve at audio frequencies and it's "forget it" at RF.
Commercial uV meters of the kind you mention have input impedance of
the order of 1k; even that can disturb RF circuits and that is the
main reason that the RF voltmeter is not (in most circumstances) as
useful as a 50-ohm power meter.
And the two instruments are, in most cases, the same; the difference
is in input impedance and in scale calibration. Examine a Boonton 42
and a Boonton 92; almost the same instrument but the 42 has a diode
power sensor and the 91 has a diode voltage probe.
Even the AD8307 is, according to the data sheet, a voltmeter; "......
log amps do not respond to power but to the voltage applied to their
input".
This voltmeter has an input impedance of 1.1 k but is so designed as
to measure voltage across 50 ohms and output the result as a DC
voltage that can be read as dBm.
An essentially 50-ohm instrument like the MAR-1 meter, AD8307 meter
etc, could be made into a voltmeter by using a FET buffer "up-front";
accuracy & frequency limit would be more "construction" issues than
"design" matters.
I also have a design that uses a transformer on the input; 10 turns on
a stack of three small beads an a single-wire primary. 50 ohms at the
primary (a power meter) reflects into the secondary and provides 5k
input impedance. I can't find the original website, though. :(1333 2008-01-15 04:52:54 Allison Parent Re: uV signal measurements 1334 2008-01-15 13:19:41 kerrypwr Re: uV signal measurements
I agree that 10-meg can be obtained with a single resistor; I was
thinking more of multiple-resistor arrangements and switching but I
didn't make that clear.
My problem was in designing a switched divider, similar to a DMM
input, for an AC voltmeter; it proved more difficult than I expected.
I eventually adopted one megohm instead of ten (which was overkill)
and that, plus a lot of attention to layout and some special resistors
(of a variety that we all have in our junkboxes but usually don't twig
to their special nature :) ), gave the desired result.
I agree with the suggestion of a pre-amplified AD8307; that's mainly
what I was getting-at. It's something that I have considered at times
but I have a couple of HP RF millivoltmeters (411A & 8405A) and a
Boonton 93A and it would, therefore, be a project that's interesting
rather than necessary.
I don't use my RF voltmeters as much as my power meters; as I said,
they are not usually as much value as a power meter in most work. And
a 'scope is not a bad voltmeter anyway!
I thought the transformer idea interesting; I agree, of course, that
there are bandwidth limitations but it's a simple project that would
be fun to try.
The OP didn't specify a frequency range so I threw the transformer
idea in for him to consider. I can't find the web site that I saw it
on, though. :(
I forestalled any problems with input cable capacitance etc in my
AD8307 meter by doing-away with the cable as you can see in the Photos
section;
http://ph.groups.yahoo.com/group/emrfd/photos/browse/7e10
A FET buffer or the ferrite transformer could be made as a plug-in
addition to the AD8307 power head.1335 2008-01-15 13:23:13 Jim Kortge Re: uV signal measurements
-snip-
>It was here: http://k8iqy.com/testequipment/-20dbmcoupler/-20dbmcoupler.htm
> I also have a design that uses a transformer on the input; 10 turns on
> a stack of three small beads an a single-wire primary. 50 ohms at the
> primary (a power meter) reflects into the secondary and provides 5k
> input impedance. I can't find the original website, though. :(
>
on my web site. I use that little probe all the time. While it isn't good below
about 4 MHz, from there on up to 100 MHz it is quite good. I just ran another
spectrum plot of this device this afternoon and will post that plot and that of
the standard -20 dB coupler to the files section of this group. Maybe I'll put
copies of the pictures from the web site there also so the plots and pictures
are together.
72,
Jim, K8IQY1336 2008-01-15 14:40:13 kerrypwr Re: uV signal measurements
http://k8iqy.com/testequipment/-20dbmcoupler/-20dbmcoupler.htm"
That's it!!!!
Very ingenious, Jim.
It's been on my Gunna List for some time; so long a list, so little
time. :)1338 2008-01-15 22:54:16 KD5SSJ Re: uV signal measurements
Cash
----- Original Message -----
1339 2008-01-16 00:51:45 kerrypwr Re: uV signal measurements
My meter uses Bob Kopski's circuit;
http://www.arrl.org/qex/020506qex003.pdf
Mine is built Manhattan-style but the circuitry is Bob's design.
I built the heads quite a few years ago, before I even saw the seminal
W7ZOI design, but I didn't get-around to designing the meter unit and
the heads just lay-around in a drawer. When I saw Bob's design I
thought "Why re-invent the wheel?" and used it with excellent results.
Some of the cal adjustments are very "touchy"; if I were building it
again I would use multi-turn trimpots.
The head contains the AD8307 and a 78L05 which regulates the "raw" 12
V. The connecting cable is ordinary two-core-plus-shield
audio/microphone cable; one for the 12 V, one for the DC output from
the AD8307 and, of course, the shield for common/ground.1340 2008-01-16 11:04:15 Jim Kortge Re: uV signal measurements
> Nice to meet you, Jim.And you also Kerry.
>Based on you comment, I re-ran the sweeps out to 1 GHz. Those plots are now on
>
> The swept results are very good; it seems to be still going beyond 100 MHz.
the EMFRD Yahoo site also. I had done plots several years ago when I built those
two devices, but probably didn't keep them or they got left behind during a
computer change. Easy to do them over.
>Yes, until some other higher priority item causes an interrupt and you go off
>
> On that basis I may move it up the Gunna List a little!!
processing that one! Been there and done that way too many times. :-)
72 and thanks for the comments Kerry,
Jim, K8IQY