Re: Added LNA and band pass filter - observations


Apr 28, 2014

 


----------------------------

#94743 Apr 28, 2014

I recently installed a roof-mounted 8-element coaxial cable collinear antenna. The antenna is up about 40 feet with a clear view of the sky in all directions. I am running a total of 125 feet of very high-quality quad-shielded RG6 to a "NooElec" R820T dongle. I realize that the cable length is very long, but I wanted to see how the system would perform. I was getting very good results, with roughly omnidirectional detection and range out to 150 NM and PlanePlotter Mode-S message rates at around 200/second at peak air traffic conditions. Note that I live in a very busy air corridor about 35 miles west of Chicago. I am equally distant from ORD, MDW and RFD airports. DUP is in the same city as my station location.

I wanted to try to optimize my antenna system by adding a simple band pass filter and LNA. I ended up with the following component lineup:



Components:             - 8-element Coaxial Colinear Antenna in PVC radome                                  - 100 feet of quad-shield RG6                                 -  "F" to "N" Adapter                                  - Polyphaser coaxial gas tube surge suppressor (grounded directly to cold water copper pipe)                                 - "N" to SMA Adapter                                 -  LNA4All LNA in shielded box                                  - Minicircuits VBFZ-1065+ Band Pass Filter                                  - SMA to "F" Adapter                                  - 25 additional feet of quad-shield RG6 to computer location                                 - "F" to MCX adapter                                 - "NooElec" R820T Dongle connected to USB hub

I initially tried placing the filter in front of the LNA. However, following the advice given by the designer in a forum posting, I found the best performance and frame rate was with the filter AFTER the LNA. This seems counter-intuitive, as it would seem better to block out-of-band signals from being amplified by the LNA. The post indicated that the best noise floor would be achieved with the filter AFTER the LNA, and this is certainly true. Frame rates increased by about 20% with the filter after the LNA. This is due to the capability of the LNA to operate at a very high bandwidth without overloading. Placing the LNA before the filter lowers the noise floor presented to the dongle considerably and improves the SNR. It also eliminates the 1.5 dB loss of the filter.

I found that I achieved approximately equal frame rates with both AGCs enabled and with the Tuner AGC disabled and set to optimal gain. The difference between the two settings is that with both AGCs enabled, my apparent signal strength percentages tripled. I also noticed the RTL1090 threshold went from -80dB with Tuner AGC disabled to -73dB with both AGCs enabled. I found it was better to operate with Tuner AGC disabled and the gain manually optimized. It is generally better to operate with as low a noise floor as possible. Turning on Tuner AGC adds 10 dB of gain and increases the apparent signal strength percentages, but with the LNA, the added gain is not really needed. It is better (higher message rates and range) to operate at a lower noise floor (THR).

My fine tuning procedure was as follows:

- Disable the RTL1090 Tuner AGC- Enable the RTL AGC- Increase the tuner gain, observing the PlanePlotter Mose-S message rate, RTL1090 message rate, and RTL1090 "THR" threshold.- Increase the tuner gain until message rates peak and begin to drop. The RTL1090 "THR" threshold will begin to rise at this point as well.- Reduce the gain one step for maximum message rate and minimum "THR" value, indicating the minimum noise floor. For me, the best setting was 48.0 dB, resulting in an RTL1090 THR of -82dB [6].

The improvement in frame rates with the LNA/filter and correct tuning was dramatic. I achieved an additional 50 NM of range (from 150 NM to 200 NM) and the decoded Mode-S frame rate as reported by Plane Plotter doubled from 200/sec to almost 400/sec under current traffic conditions. I expect it will go higher under higher air-traffic conditions. Very impressive, especially with the 100 foot coax run!

Pictures of my antenna are at:

projectmf.homelinux.com/station_pics/IMG_8169.JPG

Pictures of my surge protector/LNA/filter setup are at:

projectmf.homelinux.com/station_pics/IMG_8187.JPGprojectmf.homelinux.com/station_pics/IMG_8184.JPGprojectmf.homelinux.com/station_pics/IMG_8179.JPG

Regards,

Don



----------------------------

#94744 Apr 28, 2014

Interesting Don





Tom

GM4FDM













On 28/04/2014 13:49,donfroula@... wrote:

.I recently installed a roof-mounted 8-elementcoaxial cable collinear antenna. The antenna is up about40 feet with a clear view of the sky in all directions.I am running a total of 125 feet of very high-qualityquad-shielded RG6 to a "NooElec" R820T dongle. I realizethat the cable length is very long, but I wanted to seehow the system would perform. I was getting very goodresults, with roughly omnidirectional detection andrange out to 150 NM and PlanePlotter Mode-S messagerates at around 200/second at peak air trafficconditions. Note that I live in a very busy air corridorabout 35 miles west of Chicago. I am equally distantfrom ORD, MDW and RFD airports. DUP is in the same cityas my station location.



I wanted to try to optimize my antenna system byadding a simple band pass filter and LNA. I ended upwith the following component lineup:





Components:............ - 8-element CoaxialColinear Antenna in PVC radome

................................. - 100 feet ofquad-shield RG6

................................ - ."F" to "N" Adapter

.............................. .. - Polyphaser coaxialgas tube surge suppressor (grounded directly to coldwater copper pipe)

............................. .. - "N" to SMA Adapter

............................... .- .LNA4All LNA inshielded box

............................. . . - MinicircuitsVBFZ-1065+ Band Pass Filter

............................. ... - SMA to "F" Adapter

.............................. .. - 25 additional feetof quad-shield RG6 to computer location

.............................. . - "F" to MCX adapter

............................. .. - "NooElec" R820TDongle connected to USB hub



I initially tried placing the filter in front ofthe LNA. However, following the advice given by thedesigner in a forum posting, I found the bestperformance and frame rate was with the filter AFTER theLNA. This seems counter-intuitive, as it would seembetter to block out-of-band signals from being amplifiedby the LNA..The post indicated that the best noise floorwould be achieved with the filter AFTER the LNA, andthis is certainly true. Frame rates increased by about20% with the filter after the LNA. This is due to thecapability of the LNA to operate at a very highbandwidth without overloading. Placing the LNA beforethe filter lowers the noise floor presented to thedongle considerably and improves the SNR. It alsoeliminates the 1.5 dB loss of the filter.



I found that I achieved approximately equal framerates with both AGCs enabled and with the Tuner AGCdisabled and set to optimal gain. The difference betweenthe two settings is that with both AGCs enabled, myapparent signal strength percentages tripled. I alsonoticed the RTL1090 threshold went from -80dB with TunerAGC disabled to -73dB with both AGCs enabled. I found itwas better to operate with Tuner AGC disabled and thegain manually optimized. It is generally better tooperate with as low a noise floor as possible. Turningon Tuner AGC adds 10 dB of gain and increases theapparent signal strength percentages, but with theLNA,.the added gain.is not really needed. It is better(higher message rates and range).to operate at a lowernoise floor (THR).



My fine tuning procedure was as follows:



- Disable the RTL1090 Tuner AGC- Enable the RTL AGC- Increase the tuner gain, observing thePlanePlotter Mose-S message rate, RTL1090 message rate,and RTL1090 "THR" threshold.- Increase the tuner gain until message rates peakand begin to drop. The RTL1090 "THR" threshold willbegin to rise at this point as well.- Reduce the gain one step for maximum message rateand minimum "THR" value, indicating the minimum noisefloor. For me, the best setting was 48.0 dB, resultingin an RTL1090.THR of -82dB [6].



The improvement in frame rates with the LNA/filterand correct tuning.was dramatic. I achieved anadditional 50 NM of range (from 150 NM to 200 NM) andthe decoded Mode-S frame rate as reported by PlanePlotter doubled from 200/sec to almost 400/sec undercurrent traffic conditions. I expect it will go higherunder higher air-traffic conditions. Very impressive,especially with the 100 foot coax run!



Pictures of my antenna are at:



projectmf.homelinux.com/station_pics/IMG_8169.JPG



Pictures of my surge protector/LNA/filter setup areat:



projectmf.homelinux.com/station_pics/IMG_8187.JPGprojectmf.homelinux.com/station_pics/IMG_8184.JPGprojectmf.homelinux.com/station_pics/IMG_8179.JPG



Regards,



Don















This email is free from viruses and malware because avast! Antivirus protection is active.







----------------------------

#94745 Apr 28, 2014

Additional picture of LNA inside shielded box:

projectmf.homelinux.com/station_pics/IMG_8187.JPG



----------------------------

#94790 Apr 30, 2014

Don,



Looks fantastic! I've been on the fence about an LNA, now I'm gonna get

one for sure!

-Phil



On 4/28/2014 8:49 AM, donfroula@... wrote:

>

>

> I recently installed a roof-mounted 8-element coaxial cable collinear

> antenna. The antenna is up about 40 feet with a clear view of the sky

> in all directions. I am running a total of 125 feet of very

> high-quality quad-shielded RG6 to a "NooElec" R820T dongle. I realize

> that the cable length is very long, but I wanted to see how the system

> would perform. I was getting very good results, with roughly

> omnidirectional detection and range out to 150 NM and PlanePlotter

> Mode-S message rates at around 200/second at peak air traffic

> conditions. Note that I live in a very busy air corridor about 35

> miles west of Chicago. I am equally distant from ORD, MDW and RFD

> airports. DUP is in the same city as my station location.

>

>

> I wanted to try to optimize my antenna system by adding a simple band

> pass filter and LNA. I ended up with the following component lineup:

>

>

>

> Components: - 8-element Coaxial Colinear Antenna in PVC radome

> - 100 feet of quad-shield RG6

> - "F" to "N" Adapter

> - Polyphaser coaxial gas tube surge

> suppressor (grounded directly to cold water copper pipe)

> - "N" to SMA Adapter

> - LNA4All LNA in shielded box

> - Minicircuits VBFZ-1065+ Band Pass

> Filter

> - SMA to "F" Adapter

> - 25 additional feet of quad-shield

> RG6 to computer location

> - "F" to MCX adapter

> - "NooElec" R820T Dongle connected to

> USB hub

>

>

> I initially tried placing the filter in front of the LNA. However,

> following the advice given by the designer in a forum posting, I found

> the best performance and frame rate was with the filter AFTER the LNA.

> This seems counter-intuitive, as it would seem better to block

> out-of-band signals from being amplified by the LNA. The post

> indicated that the best noise floor would be achieved with the filter

> AFTER the LNA, and this is certainly true. Frame rates increased by

> about 20% with the filter after the LNA. This is due to the capability

> of the LNA to operate at a very high bandwidth without overloading.

> Placing the LNA before the filter lowers the noise floor presented to

> the dongle considerably and improves the SNR. It also eliminates the

> 1.5 dB loss of the filter.

>

>

> I found that I achieved approximately equal frame rates with both AGCs

> enabled and with the Tuner AGC disabled and set to optimal gain. The

> difference between the two settings is that with both AGCs enabled, my

> apparent signal strength percentages tripled. I also noticed the

> RTL1090 threshold went from -80dB with Tuner AGC disabled to -73dB

> with both AGCs enabled. I found it was better to operate with Tuner

> AGC disabled and the gain manually optimized. It is generally better

> to operate with as low a noise floor as possible. Turning on Tuner AGC

> adds 10 dB of gain and increases the apparent signal strength

> percentages, but with the LNA, the added gain is not really needed. It

> is better (higher message rates and range) to operate at a lower noise

> floor (THR).

>

>

> My fine tuning procedure was as follows:

>

>

> - Disable the RTL1090 Tuner AGC

>

> - Enable the RTL AGC

>

> - Increase the tuner gain, observing the PlanePlotter Mose-S message

> rate, RTL1090 message rate, and RTL1090 "THR" threshold.

>

> - Increase the tuner gain until message rates peak and begin to drop.

> The RTL1090 "THR" threshold will begin to rise at this point as well.

>

> - Reduce the gain one step for maximum message rate and minimum "THR"

> value, indicating the minimum noise floor. For me, the best setting

> was 48.0 dB, resulting in an RTL1090 THR of -82dB [6].

>

>

> The improvement in frame rates with the LNA/filter and correct

> tuning was dramatic. I achieved an additional 50 NM of range (from 150

> NM to 200 NM) and the decoded Mode-S frame rate as reported by Plane

> Plotter doubled from 200/sec to almost 400/sec under current traffic

> conditions. I expect it will go higher under higher air-traffic

> conditions. Very impressive, especially with the 100 foot coax run!

>

>

> Pictures of my antenna are at:

>

>

> projectmf.homelinux.com/station_pics/IMG_8169.JPG

>

>

> Pictures of my surge protector/LNA/filter setup are at:

>

>

> projectmf.homelinux.com/station_pics/IMG_8187.JPG

> projectmf.homelinux.com/station_pics/IMG_8169.JPG>

>

> projectmf.homelinux.com/station_pics/IMG_8184.JPG

>

> projectmf.homelinux.com/station_pics/IMG_8179.JPG

>

>

> Regards,

>

>

> Don

>

>

>

>







[Non-text portions of this message have been removed]







----------------------------

#94803 May 1, 2014

Hi Don, Just curious, do you have any measurement data (e.g. gain & NF) for your LNA? Have you try connect the LNA directly to your antenna without a long coax? I'm running my own mode-S receiver in Champaign. After moving my preamp on the roof (right below antenna) I got about 200% increase in data rate (now about 900 frames/second) and detection range went up to ~225 nm. Here's some information about my setup (from antenna to receiver)1/4 wavelength vertical antennaCustom made 1.09 GHz preamp with build-in SAW filter (13 dB gain, 2.8 dB noise figure)~100 feet LMR-400 coax cable with 8 dB lossCustom made 1.09 GHz receiver (detail can be found at hdl.handle.net/2142/47629) Regards,Dabin



----------------------------

#94818 May 2, 2014

Dabin,

Awesome thesis and design. I scanned it quickly.

I have not tried connecting the LNA directly to the antenna yet. I'm sure my results would improve. I'm adding about 8dB of noise to the setup through the 8dB of loss in the RG6 cable run to the LNA. I don't want to use a power injector because of the added loss, so I would likely run a separate coax to the ADS-B antenna/preamp for the DC voltage and tap off the additional coax to the roof I am using for my mast-mounted TV amp, using a DC "T" adapter.

I have not measured the LNA gain and noise at 1090 MHz, but the designer specs the unit (using a Minicircuits PSA4-5043+ E-PHEMT MMIC) at 18dB with = 1dB noise figure.

Are those PlanePlotter reported rates? I am maxing out at 600 fps on PlanePlotter, 2000 fps as reported by RTL1090.

I imagine the guys in the flight school at Champaign might be interested in your project.

Best regards,

DonWD9DMP


S
e
n
i
o
r
T
u
b
e
.
o
r
g