Re: High Intercept Point Hemt MMIC LNA's (and available!)

[Km...@aol.com]

Greetings Gentlemen!

 

Has anyone on the group looked at the SPF-5122Z MMIC (RFMD) device?  This IC has been used for quite a while in Europe and is catching on here. I have been looking at it for our cubesat groundstation LNA because of RFI and low noise requirement at 437 MHz. It is designed to cover 50-4000 MHz.

 

It offers .65 dB NF (<49 K) at 1500 MHz at 14.5 dB gain, 36 dBm IP3 and +19 dBm output power !

 

It deserves a serious look!  Particularly if you are looking for an amp that won't "crunch" under nearby large out of band signals, state of the art in noise figure, broadband for several observing band operation, and low cost!

 

I bought several amps from Goran Popovic AD6IW, and found them very good, as advertised. The only drawback to the device for me is mounting them, they are damn tiny! (His single stage is about 60.00 each)

 

I lucked into a large quantity of these mounted, on microwave PCB, in a 3 amplifier config., (single LNA driving a balanced amp). I have been converting these by adding SMA connectors and removing the 3 dB hybrid on the front end that gave CP from the dual antenna feed.  They work great.  If anyone want one for themselves, get in touch with me directly. I can make them available in raw form with ddetails for conversion for 25.00 plus shipping if anyone is interested. They run on 5VDC at about 200 mA for the 3 amplifiers.  These have enough gain to work well followed by a filter like Marcus has. (I bought some from him years ago and they are slick!)

 

The gain and NF of my non-optimized amplifier is as follows:

 

Freq.                Gain (dB)            NF (dB)

144 MHz          40                       .86

216-222 Mhz    41                       .85        Radar Fence

408-437           42                        .58

610                 40                        .81

903                 35                        1.0

1296                30.3                    .92

1421                28.6                     .99 

 

By changing caps and bias  inductors, the sweet spot in gain/NF can be moved around.  I was very happy as the sweet spot occurred at my desired frequency of 437 MHz.  My noise figure meter goes directly to 1500 MHz and I was too lazy to put the converter in front for 2304 & 3456 MHz.  It is obvious that the quadrature couplers in the balanced section are playing a part in the performance, but still, not bad! And no tinkering except mounting connectors.

 

I will have some at SARA this year and may have a short (time-filler) talk about the chip.

Anyway, good discussion!

 

73

Jeff Kruth

WA3ZKR