Архив за 05.09.2016

Review: FlightAware Pro Stick

Sept 5, 2016

Update Nov 8, 2016: Misinformation section removed as FlightAware updated relevant pages, so observations are not longer valid.

Most folks think of the Pro Stick (commonly written ProStick) as an ADS-B receiver only.

Wrong: much more than an a "High Performance USB SDR ADS-B Receiver", essentially a regular RTL-SDR dongle with built-in low noise amplifier (LNA).
Priced around $17-18 dollars, wildly popular in the ADS-B community, so hard to get one, especially in Europe (waited three weeks to be able to order this one). As of writing, unavailable on Amazon UK, but available on Amazon USA.
Background information and testing notes in the Manifesto, underlined text are links, bring you to a new page in a new window. Click / tap images for full-screen glory.

Pinch of salt

This post is exploring the feasibility of the ProStick as a general RTL-SDR receiver, and tests performance on ADS-B.
The Pro Stick has been designed and marketed as an ADS-B receiver, so observations and comments below should be strictly viewed with that in mind.

In the envelope

Static envelope with thank-you card, receiver with SMA connector, large orange case, new negative benchmark for plastic quality. It's woeful to touch.
No antenna or magnetic mount - unusable out of the box.
Size comparison from left to right: Nooelec Nano 2 and SMArt, rtl-sdr.com v.2, Prostick, generic chinese:

Case is so wide that clearly blocks adjacent USB port on Raspberry Pi microcomputers / laptops / side-by-side hubs, but if placed in top right USB port in a Pi, flat bottom enables bottom two ports to be used.

Case and Internals

Comes apart by pressing the bottom part inwards, retaining clips break off with prying and too much force.

Instant surprise: is this a previous generation rtl-sdr.com dongle with TCXO? Printed circuit board (PCB) looks and feels the same at first glance, different dimensions only evident in a side-by-side comparison:

The two PCBs are very similar, especially the tuner part, instant giveaway is LED on the ProStick, which is handy for visual feedback that the dongle works.
Advantage is that the Prostick fits into rtl-sdr.com's metal case ($5).

Bit loose, but a perfect match with no rattling when both end plates are on. Read the modding part later if thinking of getting a metal case.


Win 7 with i3 2.2 GHz 8GB laptop, SSD drive.
Requires Zadig driver installation for the actual port in use just like any RTL-SDR dongle, worked first time, no hassle.
All software identified the ProStick as "Generic RTL2832U".
SDRConsole v2.3 Build 2381: no problems.

Latest SDR# 1467 gave "no device selected", 1444 was reliable.
SDRUno worked, but with very limited gain adjustment range.

As a TV Tuner

DVB-T is not available in the USA, but common in Europe.
After reinstalling Realtec driver for TV use, SichboPVR immediately found the ProStick, and I had TV stations after 2 minutes of scan. On a discone with 15 feet / 5 meters of coax cable.

Drift and heat

No drift what-so-ever. 0 ppm cold and hot. Looks like a TCXO dongle, acts like one.
Heat very similar to other plastic-cased dongles by hand.

Noise profile

Ran rtl_power to get an idea when the noise floor starts to rise, 15 minute runs:

For comparison, same command with rtl-sdr.com's previous generation v.2 dongle:

ProStick's graph suggest that anything above 20 dB gain will incrementally increase the noise floor. In practice, signals above 20-25 dB overloaded the dongle (check waterfall in next image), and more than 10-15 dB was rarely necessary.

Amplification and gain

Biggest hurdle with the Pro Stick is finding the sweet spot, where benefits of a lower system noise figure outweigh required gain; this differs from location to location, and in a city centre, I found 8-12 dB less gain versus a non-amped dongle to give the same SNR/waterfall results in SDRSharp.

HF, VHF and UHF audio performance

Similar behavior to any R820T2 dongle, and nearly indistinguishable by ear to a previous generation rtl-sdr.com stick.
Gain settings and antenna is crucial; on a discone, easily overloads in the presence of strong local signals due to amplification.
Easier and harder on a daily basis, because morning radio is instantly on without fiddling with sliders, but faint signals are harder to hear due to large difference between two adjacent gain settings.
Shortwave listening with an upconverter is hit and miss, 20 ft / 6.5 m random wire used for the image below, minimal gain for midday strong international broadcast station. Middle of the night browsing brought in the usual suspects, hint: click on gain slider and use keyboard arrows.

Really weak signal reception on VHF/UHF is excellent with a frequency-specific antenna (tested on airband), only noise or noise with distant and faint voice in the background versus non-amped dongles. Both scenarios will be beyond what I call enjoyable, but for data reception, e.g. weather satellites and baloons, digging out a signal from the noise floor is the most important requirement, and the reason people turn to amplification.
Audio quality: not as good for extended listening sessions with constant voice as either a SMArt or any generation Silver dongle.

Direct sampling

Read previous post here if unfamiliar with the concept; essentially, signals below 30 MHz can be heard by inserting a wire into a hole on the printed circuit board.

Direct sampling hole is there, and same hole works on older rtl-sdr.com dongle, so why not?

Try it, then find your jaw. I'm still looking for it, direct sampling works really well.

But it's an ADS-B receiver

Yes, and better than any other standalone / non-amped dongle due to amplification.
Testing setup variance max 2%, large Flightaware antenna with FA 1090 MHz filter on first-floor windowsill. Data from FlightAware, plots from Planefinder. No fiddling with gain, standard PiAware image on identical Pi 3s in ModMyPi Starter Kit.
After seeing results, tested back-end setup again with two v.3 dongles on large FA antenna, and variance was less than 1 percent over 16 hours and 12,000+ position reports.
Proves that 1) new Silver dongle has excellent quality control, generic chinese dongles had 3-4 percent variance, and 2) sleep well, results are reliable.

Versus unamped v.2 rtl-sdr.com dongle

At this stage I gotta say your results will vary, but in my location, I've spent countless hours devising and testing methods to max out ADS-B, yielding a maximum of 15 % improvement, and right now, making inroads with water cooling, but:
61 percent more position reports? Astonished.

Is it filtering? Let's remove the FlightAware filter:

Whether you need a filter is up to your location, but every component in a receiver system has an insertion loss; in my case, results were nearly identical and within variance, can be due to the following reasons:
1) insertion loss is the same as improvement due to filtering, or
2) the filter doesn't work.
Hard data is 59.71% more total position reports.

Cheap no-frills antenna setup

Dismiss the $45 Flightaware antenna and $20 filter, because one of the ModMyPi Starter Kits can be bought for that money; let's use a common telescopic supplied with RTL-SDR dongles.
Extended to half-wave, SMArt's mount and cable, metal ground plane (ok it's a chocolate tin), same location/height, splitter, same rtl-sdr.com v.2 dongle as earlier:

Sidenote: Due to local flight pattens, I normally see maximum range at 250 degrees, and whereas day-to-day variation means results are uncomparable, for me it's hard to justify the $65 price tag for the large FlightAware antenna and filter. I'll run a comparison with four SMArts and three other antennas, but that's a different post.

Versus external amplification

Low-Noise Amplifiers with higher noise figures are available, cost around 25 dollars. Janilab's preamp (2db NF, review here) and the LNA4ALL (1db NF, review here) are two I have experience with, both have strengths and weaknesses. Always place external LNAs at the antenna.
An LNA is recommended over a ProStick IF, and it's a big IF:
- Long lengths of coax cable must be used to antenna - to overcome coax losses,
- required connectors are already on hand, and/or
- cost is not an issue.
I wouldn't bother with external LNAs if required antenna cable length is below 30 feet / 10 meters, but definitely advise their use over the ProStick if coax length exceeds this limit. The ProStick, like any LNA, amplify electrical noise picked up by the antenna cable - longer cable means more noise pickup, resulting in less performance.
If you have an LNA and powering it at the antenna is not a problem, adding a ProStick and turning down gain will bring down overall system noise figure. For example, an LNA4ALL provides 18dB gain at 1090 MHz with 1dB NF, then signal reaches ProStick's LNA adding further amplification. You're around 35 dB gain, and from Noise Floor testing we already know that we're good with a further 20 dB gain before the noise floor starts to dramatically rise. Even with 15 dB gain for the tuner, overall system gain will be 50 dB with a NF around 1.5 dB, which will be significantly lower than any RTL-SDR dongle at 50 dB.

ProStick ADS-B testing sidenotes

0. Results reflect data acquired at my location.
1. Any RTL-SDR dongle will enable you to receive aircraft position signals, and with a metal tin underneath a magnetic mount and a stock telescopic antenna, rtl-sdr.com's previous generation yielded more than 130 nautical mile range in this test. I've had similar results with $8 chinese generic dongles, so if you just want to know what a spotted aircraft is doing, don't spend money on specialised gear. A full FlightAware kit (antenna+filter+ProStick) is around $80 dollars, which is instant get for ADS-B fanatics, but absolutely unnecessary for normal people or for plane spotting.
2. Better range and coverage with a no-obstruction and elevated location would be possible, but I can't position the antenna elsewhere, nor invite planes to fly over for more data.
3. Amplification works, so if you already own an amplifier, use it. Gain can be reduced in PiAware, and should be experimented with.
4. Maximum range primarily depends on height above ground and line of sight. Antennas, preamps, filtering, or murmuring prayers won't change the laws of physics, they only help to get the maximum out of a particular location. In my case, I can't reliably get over 165 nm, no matter what I put on my windowsill.
5. Range claims, from receiver or antenna manufacturers, should be viewed with caution, because range is just one element in the jigsaw and very easy to manipulate for or against a receiver. I find polar plots and total number of aircraft a better indication, hence provided above.

Modding - improve performance

These steps apply to all RTL-SDR dongles, not just to ProStick.
Cooling is the easiest and cheapest way of increasing a dongle's performance.
Stage One: holes in top to let heat out, above black squares on board.
Mark approximate positions as per image, remove cover, use knife tip and lighter. Or drill. Or hot needles. Choose your weapon.

Don't be an idiot and don't do the above with cover attached, you'll ruin the dongle.

Stage Two: add metal case from rtl-sdr.com for a fiver AND thermal pads (see previous post for images).
Sticky material (often blue, order one pad not little squares, easier to cut to size and apply) will transfer heat to case, enclosing the dongle in a coffin where it can't breathe will be pointless otherwise. If you can't or don't want to apply thermal pads, leave the antenna connector side panel off, and orient the dongle vertical (up-and-down), so heat can escape.

Stage Three: add heatsinks to metal case. Won't work as well without thermal pads, I use small self-adhesive Raspberry heatsinks as in image.

Stage Four, Five, Six, N-th: Anything beyond this point will be a significant amount of pain and/or cash to do properly: oil and/or forced air and/or water cooling, massive heatsinks, Peltiers, etc will eke out the last ounces of performance, but it's up to you whether you can pay for and live with them, and if your partner won't bat an eyelid.


The Pro Stick is, in many ways, seems to be an amplified previous generation rtl-sdr.com dongle, with benefits and drawbacks built in: better signals with the right antenna, head-scratching and noise with wrong settings.
Works as an RTL-SDR dongle, weak signal reception is equal or better than generic or premium dongles. Audio quality is not as good.
ADS-B performance is great. It is an ADS-B receiver first and foremost after all.
Beginners? No. Not for radio use. Proper gain setting is crucial as dongle overloads easily, but yields good results on frequency-specific antennas and / or with filtering. No antennas, mount or cable supplied, but it's stuff you can get for a few extra dollars, or build yourself for pennies.
Ultimate value for weak signals? Yes. Amplification is king for weak signal reception, but a preamp alone costs more than this dongle, then gotta deal with connectors and power.
17 dollars, done and dusted, for this gem is a steal.

If you enjoyed this post or want to improve your reception performance, please order my RTL-SDR
Guide Book on Amazon Kindle.
Close to 200 pages of knowledge, information and diagrams for all levels of expertise.
Сентябрь 2016
Пн Вт Ср Чт Пт Сб Вс
« Авг   Окт »