The following discussion about how to measure the NF (Noise Figure) took place in the Moon-Net reflector. I found the subject so interesting that I decided to put together all the messages in this page.
On 1 February 2000 OZ5IQ wrote:
SRI and still wondering, how some of you is
talking of optimizing NF measurements on the workbench. This is shows only how
the preamp behave, at that certain setup !
Why ? , the impedance of the W6PO GaAsfet design commonly used, will cause a
returnloss of 6-8 dB or so. This means, that a miss noiseoptimization, is the
result of the workbench setup.
To measure on the bench, you have to isolate the preamp and generator with a IE.
10dB or better 20 dB pad. And still it shows only how good it can be.
Conclusion: If this kind of inputcircuit is used, the preamp must be
optimized with the REAL lenght of coax from the generator (read antenna) . The
W6PO design is good, but must be optimized ON SITE. So you have to use the
"cold sky method" . In top of your towers !
On 1 February 2000 DK9ZY wrote:
Hi Kim, you are right.
To avoid that you have to use your original antenna cable for measurement,the
cable lenght can be set up by multiple of 1/2 lambda (incl.velocity factor).
This ensures that you do not get additional (unknown) phase shift.
On 1 February 2000 K2TXB wrote:
Hello Kim. I've known this for a long time yet I've never
tried to set up and optimize for conditions at the tower top. I wonder if
anyone has made any measurements to see just how much NF degradation can happen
when the preamp is moved from the workbench to the tower? Some numbers
would be useful so it can be ascertained just how important this is.
As for using an equal length coax to make the measurements I am not convinced
that is good enough. The output impedance of the NF meter (with pad) will
be very close to 50 ohms resistive. When that is the case the length of
coax is unimportant (except for loss). Most antennas are not that perfect.
I think you would have to find a way to duplicate the resistive and reactive
components of your antenna in the lab.
I think the only 'amateur' way to effectively do this might be to check the
preamp performance in the lab at 50 ohms, and then put it on the tower, and
radiate the signal from the NF meter into the antenna. Then the preamp
would have to be adjusted (on the tower) for best relative NF. Of course
this is far from easy if your array is high off the ground, which is why I have
not tried it. (Surrounded by 80 foot trees my eme antennas must be high in
order to have much moon time.)
On 2 February 2000 Graham wrote:
The conversation seems to have wandered between getting good
results from test equipment and what might change results in real world
situation....very confusing...along with some "simulation" curve balls
along the way...
1..A good NF pre amp is very unlikely to have a return loss as good as 6
to 8db...you *might* get that if you have traded NF for other parameters..
2..gain change, giving false absolute NF, is much more likely than
absolute NF change, in pure instrumentation measuring terms the gain change is
significant in quoting a NF ...in relative or practical terms it will be
of little or no importance...(unless extreme return loss changes take place as
in use of inline filters etc. again even this has little NF affect, mostly gain
and stability issues arise). The only exception where gain change at masthead
fitting is of any importance is in a y factor test using load and antenna noise
source switching...
3..adding lengths of cable in the on bench measurement system has nothing
to do with "unknown phase changes", or "simulating antenna
feeders", or simulating anything in fact....other than simulating an
attenuator :-) which MIGHT improve the return loss change of the noise head
between ON and OFF conditions and help to provide a more reliable NF result on
test equipment, but a high ENR head and a 6 or even 10 db pad is a much better
solution...adding another half db is neither here nor there....it will do
nothing for the tuning of the DUT then being moved to another set up (even
another on bench set up)
4...adding multiples of 1/2 lambda cables anywhere is a waste of time
because a) the theory of impedance repeats applies only to a lossless and
constant impedance cable, neither of which exist...b) all it does 'on bench' is
add attenuation and improve return loss stability further(maybe)
IN addition, even if it were true, and useful, which it isn't...all it would do
is 'repeat' the impedance at the head output at both on and off states...huh !
what for ?..
On 2 February 2000 F6DRO wrote:
I agree with Graham ,lots of myths are still around about
preamps.There is one good and clear reading that everyone MUST read:it was
published in DUBUS written by Rainer DJ9BV.Title is , as far as I recall
rigth:Myths and facts about preamplifiers tuning everything you must know and
understand is there
On 2 February 2000 WB9UWA wrote:
I am certaimly not the expert in
this field that many of you are, but my antenna is small enough a
accessible enough to try tuning for best noise figure with the antenna
connected. I found that even in the quietest part of the sky, I could not
improve RX over what my bench tuning provided.
Please bear in mind I am in an urban environment, but I made every effort to find a quiet sky. My noise is never below that provided by my dummy load at the antenna location. At least in this location, hunting for the last .5 db of noise figure is really splitting hairs. I have yet to see an improvement for mast mounted pre-amp versus radio mounted. My feedline is 100' 7/8" hardline measured loss .25 db. My relay switching permitts using it both ways. When my pre-amp blows I am virtually uneffected.
On 2 February 2000 DK9ZY wrote:
Hi Graham,
Your statement is only valid if both sides are real 50 Ohm.
++++++++
3..adding lengths of cable in the on bench measurement system has nothing to do
with "unknown phase changes", or "simulating antenna
feeders", or simulating anything in fact....other than simulating an
attenuator :-)
++++++++
But a LowNA is far away from 50 Ohm.So the impedance of your load,for the LNA,
varies with the lenghth of your cable.
Except it's lambda/2 or multiple.(=180?/360?)
On 2 February 2000 K2TXB wrote:
Graham wrote:
> Subject: VHF: Measuring best NF
>
> The conversation seems to have wandered between getting good results from
test equipment and what might change results in real world situation....very
confusing...along with some "simulation" curve balls along the way...
>
> 1..A good NF pre amp is very unlikely to have a return loss as good as 6 to
8db...you *might* get that if you have traded NF for other parameters..
Agreed.
> 2..gain change, giving false absolute NF, is much more likely than
absolute NF change, in pure instrumentation measuring terms the gain change is
significant in quoting a NF ...in relative or practical terms it will be of
little or no importance...
I know that gain will change more, but are you sure that NF cannot change too?
When you tune the input of a preamp, you affect NF as well as gain. When you
change the input impedance to the preamp it changes the tuning (as evidenced by
the gain change), so why will that not also affect NF?
Have you made measurements to verify this? If so, what if any NF change
did you see? Some people have reported NF improvments (at UHF) by
optimizing on the tower. I suspect that most people here would not be
fooled by a gain change. For one thing since the test was made using a NF
meter (radiating the noise), the NF meter would most certainly not be fooled.
Also he reported gaining 2 db sun noise, and I suspect he knows how to measure
sun noise. Of course the closer an antenna's impedance is to 50 ohms real,
the less this effect would be.
> 3..adding lengths of cable in the on bench measurement system has
nothing to do with "unknown phase changes", or "simulating
antenna feeders", or simulating anything in fact....other than simulating
an attenuator :-)
Agreed.
On 2 February 2000 G3SEK wrote:
Russ Pillsbury wrote:
>I know that gain will change more, but are you sure that NF cannot
change too? When you tune the input of a preamp, you affect NF as well as
gain.When you change the input impedance to the preamp it changes the tuning (as
evidenced by the gain change), so why will that not also affect NF?
>
>Have you made measurements to verify this? If so, what if any NF
change did you see? Some people have reported NF improvments (at UHF) by
optimizing on the tower. I suspect that most people here would not be
fooled by a gain change. For one thing since the test was made using a NF
meter (radiating the noise), the NF meter would most certainly not be fooled.
The NF meter most certainly can be fooled - totally - and so therefore can the
person reading it!
The NF that the meter indicates is a function of both the true NF *and* the gain
of the amplifier under test. The NF meter only measures noise levels at the
output of the amplifier - when the noise level changes, the meter has no way of
knowing how much is due to a change in NF, and how much is due to a change in
gain.
With most FET amplifiers, the true NF is not affected very much by the change of
impedance at its input, but the gain is affected a lot... much more than most
people imagine! This can be verified from the NF circles on the device data
sheet, and circuit analysis of the complete amplifier.
A change of impedance in the noise source between the "on" and
"off" conditions will totally screw up the NF measurement, because of
the gain change that it causes. Any method that involves RF switching at the
preamp input *must* involve at least a small change in input impedance, so it is
automatically invalid. For a long time I believed that "invalid" meant
"ignore the numbers, but you can still adjust for minimum NF." But
that is NOT TRUE - the analysis by DJ9BV in the DUBUS article showed that you
can't even tell correctly where the minimum NF setting is!
It is far more important to optimize the amplifier using a noise head whose
impedance doesn't change (and therefore doesn't affect the amplifier gain) than
it is to have exactly the same input impedance as the antenna. It's also much
easier to adjust the preamp accurately to the minimum NF when you're working
indoors under good conditions. Having done that, the potential increase in NF
when you install the preamp at the antenna is quantifiable... and small.
On 2 February 2000 W9IP wrote:
Several years ago, on 432, I amplified the noise head output
with a couple of MMIC's and fed the output to a dipole. Then I mounted a
microammeter up at the top of the tower where the preamp was. There is no
way to *measure* noise figure like this, but you can certainly *optimize* it.
After carefully tweaking my preamp on the bench, I was surprised to see how much
adjustment it needed at the antenna. I picked up 2 dB of sun noise this
way... definitely worth the trouble
On 3 February 2000 K2TXB wrote:
Ian White, G3SEK wrote:
> The NF meter most certainly can be fooled - totally - and so therefore
can the person reading it!
>
> The NF that the meter indicates is a function of both the true NF *and* the
gain of the amplifier under test. The NF meter only measures noise levels at the
output of the amplifier - when the noise level changes, the meter has no way of
knowing how much is due to a change in NF, and how much is due to a change in
gain.
I thought most people used the automatic HP noise figure meters that measure
gain along with NF. Since they measure the gain, is seems that gain
changes are automatically calculated into the NF measurement. Is this not
true?
On 3 February 2000 W7EME wrote:
As it seems nice to simply increase the input r of your
antenna to drop the nf, this actually increase the noise of your source.
A fantastic work sheet can be found at Micronetics:
http://www.mwireless.com/Noise/sheet/nfm.html
Simply fill in some blanks and they will tell you how to sub your input and
output LC instead of dummy loading, which does not make much sense anyway.
With this spread sheet you will see as sm5bsz had said earlier that manufactures
claims to device performances are often eronious (spelling?).
On 3 February 2000 GM4JJJ wrote:
Here is what Leif SM5BSZ had to say on this subject a while
ago on Moon-Net for those who missed it :
========
Tuning for quieting in a FM receiver is much safer and more accurate than tuning
for optimum NF on a noise figure meter.
The procedure, as described by Johannes, has one problem. That is the stability
of the local beacon. Very small movements of the antenna will cause variations
in the signal level so it is not easy (if possible at all) to use this method to
compare several pre-amps or to see the effect of replacing some component.
The FM-quieting method using the cold sky and an accurately controlled injection
signal allows precision measurements without any sophisticated instruments.
The noise figure meter uses a large bandwidth to allow the comparison of two
different noise levels within less than 0.1dB. The measurement time is very much
longer than one over the bandwidth ( 1/B ) The measurement becomes incorrect if
some external signal enters due to inadequate screening and then tuning for
optimum S/N is not a good idea.
On the noise figure meter the reading differs by 0.05 dB if the noise
temperature of the pre-amplifier changes from 20.8K to 24.4K (0.30 TO 0.35dB
noise figure.)
The AC volt reading at the FM audio output is a non-linear function of S/N.
In the steepest region a 1 dB change of S/N may give 1 to 4dB change of the AC
volt meter reading (depends on IF/AF bandwidth, and if you can disable the
squelch). The same change as in the previous example in the pre-amp will change
the system temperature from say 60K to 63.6K if the antenna is good and really
needs a good amplifier. S/N changes by 0.25dB and the AF volt meter will show a
change somewhere between 0.25 and 1dB depending on the FM receiver.
With some time constant on the AC meter it is not difficult to reach the
accuracy required to see the 0.05dB difference in noise figure - and it is
usually much easier to tune this way because the conventional noise figure meter
will need a longer time constant.
One fine thing is that the receiver performance will be optimised at the
frequency of operation with the impedance actually present at that frequency. If
there are spurious oscillations far from the operating frequency due to severe
mismatch outside the design frequency range, it will show up with the FM
quieting method - but not with the noise figure meter.
To inject a controlled signal into the receiver, I prefer to use a 4 port
directional coupler inserted into the feed line (and it is always there). With 2
extra cables to this -40dB coupler, I terminate one of them in 50 ohms and
inject a signal into the other. The injected signal is sent towards the antenna
or towards the receiver. The ratio of the signals give me the SWR of the
antenna, and the level of the signal injected towards the receiver gives the
noise temperature of the whole system, everything included. To relate the signal
level to the noise floor I use the FM quieting method. (In transmit mode the two
cables give forward and reflected power for measurements in the shack)
To calibrate, just replace the antenna by a 50 ohm termination.
On 3 February 2000 G3SEK wrote:
Russ Pillsbury wrote:
>
>I thought most people used the automatic HP noise figure meters that
measure gain along with NF. Since they measure the gain, is seems that
gain changesare automatically calculated into the NF measurement. Is this
not true?
No - the gain measurement is part of the noise measurement, so it's subject to
the same error.
The NF meter can only measure noise power at the output, like so...
Noise temperature of amp = T_amp (referred to *input*)
Gain of amp = G
k = Boltzmann's constant
B = measurement bandwidth
Condition 1: noise head off
Noise temperature of head = T1
Total noise temperature referred to input of amp = (T_amp + T1)
Noise head measures power at output of amp =
Pn1 = k*(T_amp + T1)*B*G
Condition 2: noise head on
Noise temperature of head = T2.
Noise head measures power at output of amp =
Pn2 = k*(T_amp + T2)*B*G
The NF meter measures Pn1 and Pn2, and it is calibrated to "know" the
ENR of the noise head, which is related to T1/T2.
The NF meter can then solve the two equations above for the two unknowns, T_amp
and G.
That's how it's supposed to work. But if an impedance change at the input of the
amp between condition 1 and condition 2 causes both T_amp and G to change, the
indicated values of both parameters will be incorrect.
On 3 February 2000 K6LEW wrote:
Lot's has been said along this thread. There are some
very useful Appilcation Notes available from Hp (Agilent). See HP
Application Note 57-1 and 57-2. These should put this thread to rest as
they are available to all on the Internet.
Let me just add one thought. Done right, optimizing NF & G on the
bench will yield excellent, equatable results when the DUT is placed at the
antenna. If we are fighting over tenth's of db, then we might want to look
at the formulas in the HP Application Notes to see if that level of effort makes
sense. If the bench work is not accurate or can't be due to equipment
limitations, then the "in system" suggestions made are the best
alternative.
Have we forgotten our days of "valves", 416Bs and Nuvistors?
What would any of us have given for a GasFet in the 50's and 60's. I can
remember when a NF of 1.0 at 432 was considered "the hot setup!"; we
struggled but we did make contacts. Now it seems we are in a thread where
a few tenths of a db with NF numbers well below 0.5 are common place.
Perhaps a thread on system NF reduction, DSP, Phase Noise reduction, etc. might
be more practical. Just a thought.
On 3 February 2000 K6JEY wrote:
Lots of interesting comments about nf.
Here is another way of measuring
system noise figure that might be easier than some ways: Set up a Sinadder
on the output of the receiver and inject a modulated tone into a remote signal
generator with an antenna connected to it and aimed at the EME array. By
adjusting all components in the receiver system for best you SINAD, also
have best noise figure, but no objective measurement of what that noise figure
is; just that it is optimized. Same thing can be done with a distortion
analyzer.( A Sinadder is just a simplified automatic distortion analyzer)
Noise figure test set up accuracy is another large topic......1/10db
is all anyone will certify.
On 3 February 2000 DK9ZY wrote:
Just another comment :
-the simplest way to tune to best NF on the bench(?). -
- Built a one transistor cystal oscillator as testgenerator.
- Use a multimeter on AC-current ( approx 1 - 2.5mA,find the current-range which
shows max. indication on the scale of the multimeter using the RX noise).
- Tune in FM (Squelch open) to maximum quieting.(Refer also to Leif's proposal.)
- Thats it.
If you have another LNA with known NF and a switchable attenuator it's possible
to determine the NF of the D.U.T.,i would say with an accuracy of +- 0.2dB at
using 1dB steps on the switchable attenuator .
On 3 February 2000 G4FUF wrote:
The discussion appears to have digressed form the statement
made by Kim OZ5IQ which made the implication that the preamp NF could be
improved by retuning at the antenna. The premise of the statement was based on
the preamp having a poor return loss (RTL).
Now lets examine some FACTS:
1/ The preamp RTL is poor because the optimum impedance transformation
from 50ohm to Z.opt has not been found. This is common with the type of devices
in question. The reality is that the RTL can be easily made 20dB by the correct
transformation and at the same time find the minimum NF.
The reason is simple. The preamp is "tuned" to some intermediate
impedance which corresponds to the lowest loss transformation with "the
component parts used". In basic terms, to correct this issue requires using
sizeable parts to keep the losses low. So, the preamp will get larger in most
cases.
2/ The VSWR down a piece of lossless transmission line does NOT change.
The only reason we notice this "phenomena" is because the line is not
lossless- i.e there is an additional resistive loss which modifies the
inherent reactive ratio's.
3/ The VSWR "uncertainty" when measured is equivalent to the
line loss. In RTL terms, if we have 1dB line loss the RTL
"uncertainty" is 2dB.
4/ Retuning the preamp close to the antenna where only a short coax
section is in front ONLY results in the loss in the coax section being
MINIMISED. This is a dangerous situation, since movement of the antenna, and a
change in source impedance dpending on if it intersects other close structures
can have VERY decremental effects to the SYSTEM NF.
5/ Putting random lengths of cable into the system result in a phase and
gain shift to both the "noise" and "signal" components,
hence the SNR and NF are NOT modified. What you are normally observing is a gain
shift in the device. Therefore, (in most cases) the system NF is worse by the
gain shift ratio (ie 2nd stage.....) + the loss in the extra piece of cable you
have added in front!
6/ If by adding cable or adaptors changes the NF of the preamp on a noise
measurement system then look at the measurement system - the preamp is not
at fault only the test method.
On 4 February 2000 F6DRO wrote:
1/ The preamp RTL is poor because the optimum
impedance transformation from 50ohm to Z.opt has not been found.
This is WRONG noise matching differs from power matching simply because
optimum noise source impedance differs from optimum power transfer (situation
improves when frequency increases).Low noise preamps cannot be matched at the
input ,there is an itentionnal mismatch at the input.
Some feedback techniques may enable better matching but it is very tougth to do
because the feedback techniques are often causing stability problems.
4/ Retuning the preamp close to the antenna where only a short coax
section is in front ONLY results in the loss in the coax section being
MINIMISED. This is a dangerous situation, since movement of the antenna, and a
change in source impedance dpending on if it intersects other close structures
can have VERY decremental effects to the SYSTEM NF.
Here I agree ,I made some VSWR measurement on my 4X17M2 arry on 144 while
turning my antenna and noticed VSWR was not constant so the gain error while
optimising preamp "in situ" is worse than using the bench method.
Don't forget what DJ9BV showed:when suffering from get error and trying to
optimise NF ,you will tune it to maximum gain error condition i.e the PANFI
indicates minimum NF but in fact you detuned the preamp!
6/ If by adding cable or adaptors changes the NF of the preamp on a
noise measurement system then look at the measurement system - the preamp is not
at fault only the test method.
It may even be a way to model your gain error!
On 4 February 2000 K3PGP wrote:
I haven't been following this too closely, so I don't know
if anyone suggested using an FM receiver to optimize NF or not.
What I've been doing for many years is to crank my receiver over to the FM mode,
then radiate a weak signal into the antenna. Just enough to start quieting
the receiver. Although I have used a coupler into the coax to do this I
prefer to actually radiate a very weak signal into the antenna. Once you
have the signal level set so it's just starting to quiet the receiver, tune
everything for maximum quieting. One can get pretty close just using their
ears or you can hook an AC voltmeter across the speaker terminals and tune for
lowest reading.
I have found this method extremely sensitive to changes in NF of the front end.
I actually prefer it over the NF meter as very small changes on the order of
0.25 dB in NF can easily change the quieting of the FM receiver by several dB.
The secret is to keep the signal weak enough to just start quieting the
receiver. The only problem of course is you don't know what your absolute
NF is only that you have tuned it to the minimum possible which in my book is
all that matters.
One big advantage of this system is it is relatively immune to the inevitable
gain changes that will occur during the tuning process.
On 4 February 2000 G4FUF wrote:
K3PGP - John wrote:
> I haven't been following this too closely, so I don't know if anyone
suggested using an FM receiver to optimize NF or not.
>
> What I've been doing for many years is to crank my receiver over to the FM
mode, then radiate a weak signal into the antenna. Just enough to start
quieting the receiver. Although I have used a coupler into the coax to do
this I prefer to actually radiate a very weak signal into the antenna.
Once you have the signal level set so it's just starting to quiet the receiver,
tune everything for maximum quieting. One can get pretty close just using
their ears or you can hook an AC voltmeter across the speaker terminals and tune
for lowest reading.
Whilst in principal this is correct it can still lead to a system performing
badly because you may have failed to quantify many of the unknown factors which
can creep in. Maximum S/N may have been arrived at for the installation in
question. It tells us nothing about the "quality" of the installation.
> I have found this method extremely sensitive to changes in NF of the
front end. I actually prefer it over the NF meter as very small changes on
the order of 0.25 dB in NF can easily change the quieting of the FM receiver by
several dB. The secret is to keep the signal weak enough to just start
quieting the receiver. The only problem of course is you don't know what
your absolute NF is only that you have tuned it to the minimum possible which in
my book is all that matters.
Well this would still be true if we "tuned" it up with a hidden 3dB
pad in front............or only 13 out of 16 antennas are connected........ So
its clear we need to know some other basic parameters about the installation ?
> One big advantage of this system is it is relatively immune to the
inevitable gain changes that will occur during the tuning process.
Thats the whole crux of the matter. The tuning process needs to be based on a
relative measurement. So when you first put the preamp in-line you need to have
SOME prior knowledge about its performance. Likewise if you cannot quantify
system losses between potential discontinuities (i.e interfaces, coax to coax
etc) then the scope for having a really poor system NF is real.
On 4 February 2000 OZ5IQ wrote:
Oh boy - tremendous of arguing on the subject. I?ll
resume the major things and conclude:
1) Basically - the Preamp must be adjusted ON SITE, IF the RTL NOT is
better than ie. 10dB. (The more the better).
2) NF optimization may be done easily, in FM mode IF :
1) is well isolated from the generator(read ant)
SM5BSZ method
2) this has nothing to do with coax loss, which acts as a
pad.
3) The FM method optimization, is old knowledge as well, but
it seems
that it has been forgotten !
3) The NF min point is typically NOT equal to max gain (old knowledge but
certainly truth) Typically the trim cap. is moved a bit !
Thanks for all the replyes, many direct as well as here, I am quite aware, that
many of us have had great pleassure and much old stuff have been recalled.
SO, guys GOOD NF optimization and let us keep up the good work, as well as
the spirit and finally I am still pleased of the great spirit the contributors
have brought here.
On 5 February 2000 K3PGP wrote:
>Whilst in principal this is correct it can still lead
to a system performing badly because you may have failed to quantify many of the
unknown factors which can creep in. Maximum S/N may have been arrived at for the
installation in question. It tells us nothing about the "quality" of
the installation.
Precisely! This method is only used to make the final optimization.
It is assumed the preamp has already been checked on the bench.
>Well this would still be true if we "tuned" it up with a hidden
3dB pad in front............or only 13 out of 16 antennas are connected........
So its clear we need to know some other basic parameters about the installation
? Thats the whole crux of the matter. The tuning process needs to be based on a
relative measurement. So when you first put the preamp in-line you need to have
SOME prior knowledge about its performance.Likewise if you cannot quantify
system losses between potential discontinuities (i.e interfaces, coax to coax
etc) then the scope for having a really poor system NF is real.
Although all of the above are very valid points, I was NOT suggesting that
anyone do the first tuning of an unknown preamp on an unknown antenna.
I think you missed the point. Or maybe I did :-^) !!! And if I did
you have my apologies!
I thought we were discussing how to squeak the last bit of performance out of a
WORKING system.
The FM quieting method is to be used with a known good preamp that has already
been tested on the bench. It's also assumed that your antenna is working
and all of them are connected. The FM method is then used to squeeze out
that last bit of performance by tweaking the preamp to the antenna, not to tune
up an untested preamp or untested antenna for the first time!
I can usually squeak a little more SNR out of the system by using the FM method
and doing the final tuning on the antenna. However your mileage may vary
depending on how careful you are and exactly what you are trying to achieve.
Not every idea is appropriate for all situations.
On 7 February 2000 F6DRO wrote:
1) Basically - the Preamp must be adjusted ON SITE,
IF the RTL NOT is better than ie. 10dB. (The more the better).
THIS IS NOT TRUE!NOISE MATCHING DEPENDANCE TO SOURCE MATCH IS LOW PREVIOUS
SOURCE SWR IS 1.3/1 OR BELOW.YOU WILL DETUNE YOUR PREAMP BY TRYING TO OPTIMISE
IN SITU IF IT HAS BEEN TUNE ON THE BENCH WITH A GOOD SETUP
On 7 February 2000 G4FUF wrote:
For those that wish to understand the implications of the recent discussions on
here regarding NF and improve their knowledge I would suggest a usefull and
concise document can be viewed or downloaded in *.PDF format from:
www.hp.woodshot.com/hprfhelp/4_downld/lit/other/primer2.pdf
In the application notes section look for "Primer 2", its 61kB
You might like to consider reading some of the other related articles also to avoid mistuning your preamps at the antenna........
(Note: I'm not showing any E-Mail
address here in order to avoid them from being collected by SpamBots. You can
possibly find the E-Mail addresses of the above OM at QRZ.COM.)
This discussion is still not closed. If you have any opinions or additional
related information you would like to be published here, just send
me an E-Mail
Read
other Moon-Net discussions
Home page
Site map
Radio Sherlock search engine
Terms of service
Privacy
Cookies