[Date Prev][Date Next][Thread Prev][Thread Next] - [Date Index][Thread Index][Author Index]

Re: Weak Signal Banquet NF numbers..



What Pieter said is this:  The noise figure meter determines NF by
switching on and off a broadband noise source, and calculated the NF by 
comparing the converted noise signal (noise source ON) to the converter 
noise floor (noise source OFF).
(really a sensitivity measurement)

By allowing both sidebands through, the noise with the noise source on with 
be 3 dB greater than without it on-- making the unit seem more sensitive 
than it is, making the NF measurements seem better when in reality there is 
3dB more noise from the image present at the IF.

The Agilent NF Analyzer had a DSB measurement mode on it, but I think we 
didn't use that mode.. I'm not entirely sure.  Didn't comprehend the problem 
the day after I got the 3033.

The crux of this whole discussion, though, is properly modified (and I think 
that means an ISM filter in the second position and a 50 ohm strip or coax 
jumper in the first filter position is the proper modification) is that the 
Transystem is a good choice for a converter.

One thing I noticed though is that with the AO-40 spreadsheet calculator, if 
a 0.5 dB NF/10-20 dB gain preamp can be built to fit in the first filter 
position (I bet it could) on the Transystem-- it'd
put it into the league (actually slightly hotter than) of the SSB Electronic 
$450 converter.  The calcs show the 0.8 db system NF of the SSB would make 
it around 3dB more sensitive than a Transystem (minimum discrenable signal) 
-- I bet that the SSB unit either uses two PHEMT's in it, or one with a <2 
dbNF second stage GaAS FET or GaAS MMIC to obtain this.

Also one other argument I've heard is that it's simple to replace the ERA-3 
in the SSB converter as a 2M transmit fuse.  I think a mmic could be fitted 
in the IF of the 3033's easily if the IF amp is toasted.

I'm actually somewhat impressed by the simplicity of the design of the 3033.

Fred W0FMS



>From: Stan <stan@capeonramp.com>
>To: Jon Ogden <na9d@speakeasy.net>
>CC: Pieter Ibelings <elpieter@hotmail.com>,   "Frederick M. Spinner" 
><fspinner@hotmail.com>, amsat-bb@AMSAT.Org
>Subject: Re: [amsat-bb] Weak Signal Banquet NF numbers..
>Date: Fri, 25 May 2001 12:58:27 +0000
>
>Hello Jon and the net,
>
>What Peter is saying is that the noise figure indicated is about
>3 dB better (lower) if the actual test is for double sideband, than
>if you were expecting a single sideband test result.
>
>The double sideband response can be determined by injecting a clean
>RF carrier at the devices mixer product frequencies. If the device
>is DSB, ie no or poor image rejection, there will be a response at
>the IF port when the injected RF carrier is at LO + and - the IF freq.
>
>Example: LO = 10,223.0 MHz, IF = 145.0 MHZ, high beat (USB) would be
>	LO + IF = 10,368.0 MHz and the low beat (LSB) would be at
>	LO - IF = 10,078.0 MHz.
>
>The savvy manufacturers' spec the noise figure as DSB, double
>sideband,
>if no image rejection is used on the RF side of the mixer.
>
>DB6NT specs his 24 GHz transverter RX noise figure as 8 dB DSB.
>When you add a bandpass filter, the NF jumps up by 3 dB to about 11 dB
>plus the insertion loss of the band pass filter.
>
>I hope this sheds some light.
>
>Stan, WA1ECF
>
>
>Jon Ogden wrote:
> >
> > Peter,
> >
> > While I follow and agree with what you are saying, I am not sure I
> > understand your final conclusion.  I need some clarification.
> >
> > Are you saying that by measuring a system that converts both sidebands 
>with
> > no image rejection that the actual noise figure is 3 dB worse or 3 dB
> > better?
> >
> > I would think that by converting both sidebands, you are actually 
>generating
> > more noise and if you aren't taking that into account that your noise 
>figure
> > number will be worse than it really is.  Perhaps I have it backwards.
> >
> > Help?
> >
> > 73,
> >
> > Jon
> > NA9D
> >
> > on 5/23/01 7:56 PM, Pieter Ibelings at elpieter@hotmail.com wrote:
> >
> > > Hi Fred,
> > >
> > > When you measure the noise figure of a down converter that convertes 
>both
> > > sidebends to the IF frequency, you end up with a 3dB error. Let's say 
>you
> > > have a 15 dB ENR noise source like a HP346 that has really flat 
>response
> > > from DC to 18 GHz. As you power the source on and off, the converter
> > > converts both sidebands to the if frequency. This results in twice the 
>noise
> > > showing up at the IF. You get the noise from 2400 MHz and you get the 
>noise
> > > from the image. Both are almost the same signal level due to the flat 
>noise
> > > source.
> > > When you calibrate the noise meter, it turns the noise source on and 
>off and
> > > looks for the delta in noise between the two settings. It only looks 
>at the
> > > noise "once" not "twice" as in the double conversion explained above. 
>The
> > > noise figure meter when measuring noise figure looks for noise delta 
>between
> > > a resistor and a noise source. The bigger the delta the more sensitive 
>the
> > > preamp or converter is. Now, if you down convert both sidebands (twice 
>the
> > > noise) and the calibration is based on only one bandwidth of noise, 
>the
> > > meter thinks that you have a really good converter because it sees a 
>really
> > > big delta. This results in a error of around 3 dB. So when measuring a
> > > converter with no image rejection in a noise figure meter that expects 
>a
> > > single sideband, you need to add a correction factor of almost 3dB to 
>the
> > > measured result.
> > >
> > > I hope this makes sense. There are many different correction factors 
>that I
> > > have not accounted for the bulk is here.
> > >
> > > 73's
> > >
> > > Pieter N4IP
> > >
> > >
> > > ----- Original Message -----
> > > From: "Frederick M. Spinner" <fspinner@hotmail.com>
> > > To: <amsat-bb@AMSAT.Org>
> > > Sent: Wednesday, May 23, 2001 19:54
> > > Subject: [amsat-bb] Weak Signal Banquet NF numbers..
> > >
> > >
> > >> Here are the numbers that were measured at the Weak Signal Banquet at
> > >> Dayton:
> > >>
> > >> New DEMI 2.4 GHz Preamp (assembled by W1BFN):
> > >> 0.58 dBNF/16.95 dB gain
> > >> Old DEMI 2.4 GHz Preamp (assembled by W0FMS, measured at Winnipeg 
>CSVHF
> > >> meeting 2000):
> > >> 0.68 dbNF/11.78 dB gain
> > >>
> > >> Modified Drake 2880, lengthing bandpass filter strips and IF cap mods
> > > (W0FMS
> > >> modifications):
> > >> 6.00 dBNF/23.0 dB gain  @ 123 MHz IF, 2401 MHz center freq at RF
> > >>
> > >> Same Modified Drake 2880, with the OLD DEMI Preamp above in front:
> > >> 1.25 dBNF/37.5 dB gain  @ 123 MHz IF, 2401 MHz center freq at RF
> > >>
> > >> Transystem 3033 bought at the flea, unmodified.  Note the image freq
> > >> concerns, though..
> > >> 1.42 dBNF/28.8 dB gain  @ 123 MHz IF, 2401 MHz center freq at RF
> > >>
> > >> The NF debate about the Transystem applies-- but maybe someone can 
>explain
> > >> to me that when you pulse a broadband noise source such as the noise 
>head
> > > on
> > >> a new Agilent NF Analyzer, how the meter can distinguish one noise
> > > sideband
> > >> from another if both are getting converted to the IF?  So I 
>understand the
> > >> problem, but why doesn't it affect the NF readings?
> > >>
> > >> Anyway-- the T/S converter can operate in that range with careful 
>mods,
> > >> maybe an ISM filter, etc.... cheaply...
> > >>
> > >> Interestingly enough, when I massaged these numbers through the
> > >> AO-40 spreadsheet, the gentlemen on the board harping the SSB 
>converters
> > >> have a point.  The SSB converters with a 0.8 dB system noise figure 
>have
> > >> about a 3 dB lower minimum discernable signal (noise floor) than the 
>Drake
> > >> w/preamp and/or the Transsystem.  Now if the massive gain could 
>somehow be
> > >> handled, the Transystem with either DEMI preamps in front of it could 
>have
> > >> an edge on the SSB converter.  Also a Drake that had the 2 dBNF or so 
>mods
> > >> done to it that had a DEMI preamp in front of it could also 
>outperform the
> > >> SSB converter.
> > >>
> > >> But if not, for the $350 I've saved not getting overpriced SSB 
>electronic
> > >> equipment, I'm close enough with the Drake/DEMI preamp or the 
>Transystems
> > >> units with the 3'X2' BBQ's (hell, I spent it on a G-5500 Yaesu rotor 
>that
> > >> seems to be backordered from Japan at Dayton..)
> > >>
> > >> So, this in hand, maybe I'll come up with some suggestions in the 
>next few
> > >> weeks for cheap, simple Mode-S recvs and I also have some ideas for 
>mods
> > > to
> > >> the Transystems.
> > >>
> > >> Fred W0FMS
> > >>
> > >>
> > >> _________________________________________________________________
> > >> Get your FREE download of MSN Explorer at http://explorer.msn.com
> > >>
> > >> ----
> > >> Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
> > >> To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org
> > >>
> > > ----
> > > Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
> > > To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org
> > >
> >
> > -------------------------------------
> > Jon Ogden
> > NA9D (ex: KE9NA)
> >
> > Member:  ARRL, AMSAT, DXCC, NRA
> >
> > http://www.qsl.net/ke9na
> >
> > "A life lived in fear is a life half lived."
> >
> > ----
> > Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
> > To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org
>----
>Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
>To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org

_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com

----
Via the amsat-bb mailing list at AMSAT.ORG courtesy of AMSAT-NA.
To unsubscribe, send "unsubscribe amsat-bb" to Majordomo@amsat.org



AMSAT Top AMSAT Home