Trial Run Results for Three Types of Whip Antennas

wiedehopf

New Member
Make sure you don't get the polarities mixed up. (pretty sure you didn't because it would be worse if you had ;) )

Also does that power inserter allow power to the dongle?
You need one that doesn't.
The bias-t we always showed in the drawings blocks DC towards the dongle.
This thing you are using doesn't look like it blocks DC from going to the dongle.

Especially 12 V will likely kill the dongle and maybe the Pi soon after.
 

AnirbanSen

New Member
While stopping dump1090-fa, I noticed that running dump1090-fa from the console churned out a lot of data. Tried running Virtuar Radar Server on the PC and accessing http://pi2:30005, I was seeing many aircrafts. (8 on the regular Pi, 25 on this Pi2). This connection is not stable and works intermittently. At the same time, the Pi2 continues to remain inaccessible (some brief windows ssh does work). There might me multiple issues mixed in this.

The blue light above (12V DC_DC buck) is now supplying about 13.8V (increased it a bit) to the amp. The power goes thru the inductor straight to the amplifier on the roof. After the amplifier, the other Bias-T on the roof works as a DC block (as the amplifier has a DC pass through). About 1m of wire to the antenna.

There is no potential difference across the dongle side of the Bias-T. Am positive both Bias-Ts are working fine. The design is the same one as @ab cd recommended. (Inductor 2.2 mH, Ceramic Capacitor 200pF). What I feel unsure is whether the switching power supply is supplying some leakage current time to time to the ground (⏚) terminal. Continuity tester shows AC Neutral and DC Ground, power inserter body and RPi (⏚) are all the same. I have felt some tiny shocks on light touches of the arm, but the multimeter picks up nothing - AC or DC.

I don't know if I have to invest in rtl sdr v3 hardware.
 
Last edited:

AnirbanSen

New Member
IMG_20190802_141810.jpg

The antenna on the roof. The base of the antenna is covered using copper shielding tape. The red electrical insulation tape covers the F to F connectors joint. The copper wires are coated with silicone paste.

IMG_20190802_141820.jpg

The blue lunch box now houses the inline amplifier and the Bias T (working as a DC block). The red cylinder outside is a Lightning Protector. (Not sure if it works - it has no connection to the ground ⏚ ). The wire to the left is about 32-36m to the Bias T inside my apartment.

IMG_20190802_143446.jpg

Those building are the highest the 5-6Km proximity on the South side.

Is the iron frame on which the antenna is mounted likely to affect the signal in any way?
 
Last edited:

AnirbanSen

New Member
UPDATE:
Pi2 seems to be running now on its own. I haven't made any changes to the system. SSH seems more available. However, it still locks up from time to time. During a lock up, feeding to flightaware would stop, SSH window would become inoperative, and VirtualRadar would not receive any signal. I have been unable to pinpoint to the cause.
The range is much better than my existing setup on the 6th floor:
New antenna: https://flightaware.com/adsb/stats/user/anirbansen#stats-108363
Old antenna: https://flightaware.com/adsb/stats/user/anirbansen#stats-102898

Site 102898 would be using the new antenna after a successful swap.

Best gain appears to be AGC. Keep losing aircrafts with values between 20 and 49.6. Currently no filters are being used at all.

I am wondering if Pi2 is requiring more power from the adapter. The 0 Pi Wireless has low power requirements and is running off a 230V-5V 1A DC charger. Will switch to a new charger later and update.

PiAware and Graphs1090 are not running on the Pi2 as there is something wrong with lighttpd. Pi2 is being used solely to test out Buster. The plan was to switch the new antenna on the main Pi, but that has led to the Pi itself having undiagnosed issues.
 
Last edited:

foxhunter

New Member
And i am still on my way with my very simple indoor DIY antenna with impressive results
I've set the gain manually to 49.6 after the setting "max" delivered too much messages > 3db

As it does not hurt, i am too lazy to improve it. But this will change :D
I am using the blue FA pro plus stick and i am surprised about the performance overall.

1564921018928.png
 

wiedehopf

New Member
I have been unable to pinpoint to the cause.
Connect the negative/ground terminals of the power supplies for the RPi and bias-t.
If there is a potential difference, then it doesn't go via the dongle and should help with your problems.

Edit: I'm a little confused, that thing displaying 13.2 V is simultaneously supplying 5V via USB?
Also don't put the cables going to the bias-t so close to the dongle.
(Maybe just put aluminium foil around the dongle, maybe that will help)
Have you checked which tuner that dongle has?
It should be somewhere in the output of this script of mine:
Code:
sudo bash -c "$(wget -q -O - https://raw.githubusercontent.com/wiedehopf/adsb-wiki/master/rtl_test.sh)"

If DC on the center conductor to the stick is indeed blocked by the capacitor, then this shouldn't be a problem.

You can also fit both power supplies with extra smoothing if you want.
1564944730818.png


The coil/inductor goes in series with the positive line of the power supply.
The capacitor goes across + and - on the side of the load (pi or bias-t).

So the positive cable coming from the power supply to the RPi for example is interrupted, you insert the coil.
Do the same with the supply for the bias-t.
Then the capacitor is in correct polarity connected across + and - on the side of the load, not towards the power supply.

This is an LC filter to reduce ripple.
 
Last edited:

AnirbanSen

New Member
Connect the negative/ground terminals of the power supplies for the RPi and bias-t.
If there is a potential difference, then it doesn't go via the dongle and should help with your problems.

Edit: I'm a little confused, that thing displaying 13.2 V is simultaneously supplying 5V via USB?
Normally, that should happen as the circuit shares a common ground. But perhaps I'll need a short just in case.
The display shows the input voltage of the buck converter (4.5 - 40 V input, 5V output). (See https://robu.in/product/dc-dc-4-5-40v-5v-2a-usb-charger-step-converter-voltmeter-modul/)

Also don't put the cables going to the bias-t so close to the dongle.
(Maybe just put aluminium foil around the dongle, maybe that will help)
Noted. But the dongle also gets very hot. Will the foil trap the heat to help dissipate it. Will foil it and keep it away.

Have you checked which tuner that dongle has?
It should be somewhere in the output of this script of mine:
RTRealtek, RTL2838UHIDIR, SN: 00000001
(iv) rtl_test -t showed that both dongles are otherwise fine when tested on a freshly booted system.
Using device 0: Generic RTL2832U OEM
Detached kernel driver
Found Rafael Micro R820T tuner
Supported gain values (29): 0.0 0.9 1.4 2.7 3.7 7.7 8.7 12.5 14.4 15.7 16.6 19.7 20.7 22.9 25.4 28.0 29.7 32.8 33.8 36.4 37.2 38.6 40.2 42.1 43.4 43.9 44.5 48.0 49.6
[R82XX] PLL not locked!
Sampling at 2048000 S/s.
No E4000 tuner found, aborting.
Reattached kernel driver
If DC on the center conductor to the stick is indeed blocked by the capacitor, then this shouldn't be a problem.

You can also fit both power supplies with extra smoothing if you want.
View attachment 2912

The coil/inductor goes in series with the positive line of the power supply.
The capacitor goes across + and - on the side of the load (pi or bias-t).

So the positive cable coming from the power supply to the RPi for example is interrupted, you insert the coil.
Do the same with the supply for the bias-t.
Then the capacitor is in correct polarity connected across + and - on the side of the load, not towards the power supply.

This is an LC filter to reduce ripple.
Can a single 2.2mH inductor work for both the BiasT and the RPi? They share the main step down source.
+12V DC ------[INDUCTOR]---[1000mF Electrolytic Capacitor]----+[13.2V DC to 12-13V DC]----[BIAS TEE]------


Annotated the picture in case there is any confusion:
Revised IMG_20190804_015336.jpg
 

wiedehopf

New Member
Noted. But the dongle also gets very hot. Will the foil trap the heat to help dissipate it. Will foil it and keep it away.
Yeah heat is a problem.
If you have a metal box with holes, you can place it in that :p
Also put the RPi Zero a bit further from the dongle if possible.
Distance is your friend when interference from other electronics is concerned.

Anyway it's mainly for testing to find out if it's a problem.

They should already share a common ground, i do agree.

Can you run this:
Code:
sudo bash -c "$(wget -q -O - https://raw.githubusercontent.com/wiedehopf/adsb-wiki/master/rtl_test.sh)"
It should also show undervoltage on the RPi as well as the dongle losing samples.

Unscrew the left conductor of the bias-t you built and measure the center conductor against the RPi ground and bias-t case ground.
Should both read 0 V.
Maybe you just have a bug in your bias-t and it's putting voltage on the center conductor.
Also when disconnected you can measure potential difference beween both grounds.
 

AnirbanSen

New Member
Code Result:

12945 /usr/bin/dump1090-fa --device-index 0 --modeac --gain -10 --ppm 0 --net-bo-port 30005 --max-range 360 --net --net-heartbeat 60 --net-ro-size 1000 --net-ro-interval 1 --net-ri-port 0 --net-ro-port 30002 --net-sbs-port 30003 --net-bi-port 30004,30104 --net-bo-port 30005 --json-location-accuracy 2
--write-json /run/dump1090-fa --quiet
-----
Lost samples in the first 2 seconds after starting the test are common and not a problem!
Starting 30 second rtl_test, standby!
-----
Found 1 device(s):
0: Realtek, RTL2838UHIDIR, SN: 00000001

Using device 0: Generic RTL2832U OEM
Found Rafael Micro R820T tuner
Supported gain values (29): 0.0 0.9 1.4 2.7 3.7 7.7 8.7 12.5 14.4 15.7 16.6 19.7 20.7 22.9 25.4 28.0 29.7 32.8 33.8 36.4 37.2 38.6 40.2 42.1 43.4 43.9 44.5 48.0 49.6
[R82XX] PLL not locked!
Sampling at 2400000 S/s.

Info: This tool will continuously read from the device, and report if
samples get lost. If you observe no further output, everything is fine.

Reading samples in async mode...
Allocating 15 zero-copy buffers
lost at least 140 bytes
Signal caught, exiting!

User cancel, exiting...
Samples per million lost (minimum): 0
-------
Test finished!
More than 2 lost samples per million or other errors probably mean the receiver isn't working correctly.
Try another power supply before condemning the receiver though!
-------
-------
No undervoltage detected, looking fine!
If the dongle is not directly plugged into the Raspberry Pi, lack of power/voltage could still be an issue.
Even without detected undervoltage a better power supply can often improve reception!
For optimum performance i would recommend the Official Raspberry Pi power supply.

Multimeter Values:
1) Bias-T Centre and Ground: 0.00V
2) Power Supply 0V Ground and Buck Converter Ground (12-5V Stepdown): 0.03 - 0.05V [This is probably how the switched power supply works - the output ground is not the same as the input ground]
3) DVB-T Ground and RPi Ground: 0.07 - 0.08V [Again, the tuner stick does not share a common ground perhaps - is this true?]
4) Power Supply 0V Ground and DVB-T Ground: 0.16 - 0.19 V [This is a high value could explain why the RPi locks up especially when the DVB-T is connected to the Bias-T as this shorts the ground.

While try to shutdown the system, it again experienced a lockdown.
Other than lighttpd, system had been working normally till the above test. Sometimes in between, I put the RPi on a regular charger. For the test, I reverted to the buck converter. Now it is sluggish. I shall try to power the system with a regular 1A cell phone charger. Perhaps the problem lies in the [230V-12V] and [12V-5V] setup.
 
Last edited:

AnirbanSen

New Member
Update 2:
Somehow the connecting to the DVB-T to the Bias-T affects the DVB-T for sometime. I shall change the cable to a pigtail.
i) The RPi is detached from everything and started on a simple separate power supply. -- SSH works normally.
ii) The DVB-T without any antenna is plugged in. The DVB-T locks up the RPi
iii) Upon restart of the RPi with the DVB-T (without antenna) plugged in, the SSH response starts slow. Becomes better after a few mins.
iv) Before connecting the DVB-T and Bias-T, the voltage between DVB-T Ground and Bias-T Ground was about 0.3 - 0.9V (fluctuating constantly). Note that they are on a different AC power source.
v) Upon connecting, Virtual Server Radar started picking up data.
vi) New RTL Test showed no undervoltage or any loss.

Overall, still very temperamental :rolleyes: :)
 
Last edited:

wiedehopf

New Member
3) DVB-T Ground and RPi Ground: 0.07 - 0.08V [Again, the tuner stick does not share a common ground perhaps - is this true?]
That's interesting.

Just connect all the grounds together, this hopefully helps ;)

Maybe the wireless isn't happy with the interference created by the DVB-T stick.

Also you can switch off MLAT for fr24feed, that way you save some processing power.
Generally doing MLAT on the Pi Zero is not recommended.
 

AnirbanSen

New Member
Just switched antennas. Roof antenna is now on main Pi board. Seems to be performing normally. Will post Graphs data in an hour plus.
As for MLAT, I am pushing the poor Pi a lot. Not just MLAT, look at the other stuff running. I need better cooling options for the RPi.

CPU[|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||97.1%] Tasks: 88, 101 thr; 3 running
Mem[||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| 180M/464M] Load average: 2.75 2.28 2.22
Swp[|||||||||| 9.25M/100.0M] Uptime: 23:39:55
Mem[||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| 180M/464M]
 
Last edited:

AnirbanSen

New Member
Another curious fact I want to report is that when it rains, or the air is very humid, the range is around 220-240nm and messages >= 3dBFS hovers between 3 and 9% approx. Any idea if RF refraction increases with humidity or cloudy weather?

A small external USB fan crudely fanning the heat sink has lowered temperature to 40C from 55C as the Pi 0 CPU is between 90 - 96%.


Today the other LNA+filter arrived. I am traveling this week so perhaps next week I'll get to test it.
IMG_20190813_171650-1200x1600.jpg
 
Top