[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[linrad] Out of band QRM. Worst case ?

Hi Stan,

Well, looking at the specs for the preamps of LNA Technology, one would
gather that something really spectacular has been acomplished. LNA Technology
would not reveal all the circuit details, but a number of things can be figured out
from the pictures. The claim is for .2db noise figure and a 1 db output compression
of +20dbm. The device used is the ATF33143. My test indicate that even when
designed so for best power output consistant with a .2db nf, about +16dbm is about
the maximum you can get. This is with an unstable design. For good stability with
common solutions, less output power is obtained. Another thing that is obvious
is that he uses a tapped input on his cavity. The device gate is at the high impedance
end of the cavity.  The ATF33143 wants to see a 150 ohm source impedance for
optimal noise figure. This pretty well fixes the selectivity curve. Simulations here
(Serenade SV) showing the selectivity curve for a tapped input cavity show that at 100 Mhz, 
the gain is only 12 db down from peak. There is virtually no selectivity in the 150 Mhz
range. The majority of the selectivity shown in his curves obviously (obvious to me),
comes from his dual section output filters. Larger cavity diameters are evident in
the photo (2"). The larger cavity in this case only serves to reduce cavity loss and 
has no impact on selectivity unless the cavity were made really small.
Basically with strong FM BCB signals and throw in a few paging transmitters
for good measure and this preamp would stand a good chance of being overloaded.
LNA Technology.... Good, yes, perhaps not as good as he says. It certainly would
not be as good as the selectivity curves show.

73, Jim Shaffer, WB9UWA.

On 14 Mar 2004 at 2:36, Stan wrote:

> Hello the Net:
> Preamplifier problems I have had in the past were due to :
> 1. the 1/4 KW pages on 151 Mhz and the nearby FAA station at 136 MHz
> 2. low noise preamp in the 2M band but also a very low noise figure,
> with low + gain, in the 850 to 950 MHz cellular/garbage band.
>     like the SSB model LNA 145.
> I really like the LANDWEHR mast mounted preamplifiers , that I use for
> terrestrial and satellite work decent NF and gain,  tightly around the
> 2M band, NF ~1db so not spectacular for eme. Many preamplifiers are
> broadband with poor performance in a dense electromagnetic
> environment.
> for 2M EME and terrestrial I am considering the cavity input product
> offering of:
> http://www.lnatechnology.com
> Stan, WA1ECF
> wb9uwa@xxxxxxxxxxx wrote:
> >HI Leif,
> >
> >As you know, I have been wrestling for some time with the front end
> >problem for 2M EME here. My plan is to drive an RX2500 direct with
> >the front end. 
> >
> >I think you overestimate what is commercially available and well
> >suited to Linrad implementation. The second stage you mention that
> >has only 10db gain and a 2db noise figure probably does not even
> >exist even though it is easy to homebrew with 3x J310's. The first
> >stage with only 15 db gain does not exist either. Most commercial
> >preamps will have between 23 and 28 db gain at 2M with the smaller
> >gain number ususally being the result of loss in the output circuit. 
> >There will be a great temptation to use some of the commercially made
> >preamps by itself to drive the RX144 or worse yet, many will try to
> >cascade two of these fairly high gain preamps, perhaps with a
> >commercial bandpass filter between them. This would be a disaster for
> >IP3  make mediocre a potentially great system. I suggest as a minimum
> >your RX144 should have a system noise figure of 2db. 1db nf or less
> >would even be better. In that case some EMEers in some very harsh
> >environments would likely use no external preamp at all.
> >
> >The first stage preamps that would seem to be called for in a Linrad
> >system would be in the ATF33143 and MGF1801 class. Neither of these
> >are known for being unconditionally stable. When you throw in the
> >requirement for only 15 db gain in the first stage, then lossless
> >feedback seems to be called for. As far as I know there are NO
> >commercial manufacturers successfully using enough lossless feedback
> >to reduce the gain down to anywhere near 15db. An oscillating preamp
> >is often the result. Perhaps someone with the knowhow and proper
> >design simulation software will step up to the challenge. One could
> >use a  normal commercially available preamp with an attenuator after
> >it, but the result would be poor output power from that stage and due
> >to a lack of input selectivity, the "damage" may already be done.
> >
> >In the US, there will be no signals of any consequence below 144 Mhz
> >until the FM BCB. Starting at only 150 Mhz there are many two-way FM
> >radio signals. 250 watt paging transmitters are often present at 152
> >and 157Mhz. These paging signals here are -35dbm. There can be two of
> >them at once at 152Mhz. I am in a town of only 40,000 people, but
> >adjacent to a town of 60,000 people. Certainly large cities will
> >present a bigger challenge. Commercial preamps generally lack
> >substantial selectivity AHEAD of the active device. Two Meter
> >repeaters are present from in the low 145 Mhz range to 148Mhz.
> >
> >73, Jim Shaffer, WB9UWA.
> >
> >
> >
> >
> >On 13 Mar 2004 at 1:10, Leif Åsbrink wrote:
> >
> >>Hi All,
> >>
> >>I am going to make a front end for 144 MHz that will fit
> >>the WSE converters and Linrad.
> >>
> >>There will be two RF amplifiers, one preamplifier for
> >>mounting in the tower and I will probably not make them
> >>since such amplifiers are available from many sources.
> >>
> >>The noise figure of RX144 will be about 11 dB and there 
> >>is a need for a second RF amplifier with high IP3.
> >>
> >>The idea is to have a tower mounted preamp with a 
> >>gain (including cable losses) of about 15 dB followed 
> >>by a high level amplifier with a noise figure of 
> >>about 2 dB that will add about 10 dB gain.
> >>
> >>A good preamplifier may have an output IP3 of +25dBm.
> >>The input IP3 of the RX144 is about +20 dBm but if
> >>both signals fall within the passband of the RX70, the
> >>input IP3 is only +15 dBm.
> >>
> >>After adding the second RF amplifier, the IP3 mis-match
> >>between a +25 dB IP3 preamplifier and the Linrad system
> >>amounts to 20 dB for in-band signals.
> >>
> >>The in-band input IP3 (at the antenna) will be -10 dBm
> >>while compression will start at something like -20 dBm. 
> >>
> >>The second RF amplifier could incorporate a filter that
> >>converts in-band interference to out-of-band interference.
> >>Are there any signals within 10 MHz from the 2 m band that
> >>could produce more than -30 dBm so they would be near blocking
> >>or is there any risk for a signal pair that could produce 
> >>third order intermodulation on 144 MHz ? Two equally strong
> >>in-band signals at -50 dBm would give IM3 at -130 dBm which 
> >>would be well audible.
> >>
> >>As far as I know there is nothing here in SM but I do know 
> >>there are pagers just below 144 MHz in the UK. It will be 
> >>fairly easy to design filters that make performance limited
> >>by the mast mounted preamplifier for signals that are more
> >>than 4 MHz outside the band edges if one wants full sensitivity from
> >>144 to 146 MHz. Likewise it would be fairly easy to make the system
> >>limited by the preamplifier for signals below 142.5 or above 146 MHz
> >>if one wants full sensitivity from 144 to 144.5 MHz only.
> >>
> >>My question is: Is there a real need for better than this?
> >>
> >>
> >>73
> >>
> >>Leif / SM5BSZ
> >>
> >>
> >>
> >>
> >>_______________________________________________
> >>Moon-Net posting and subscription instructions are at 
> >>http://www.nlsa.com/nets/moon-net-help.html
> >>
> >
> >
> >
> >