That’s what I said. See quote below:
Very true, fully agree.
In 2013 the 1090 Mhz LNAs were very costly, 100 euro plus. I decided to use a cheap satellite amplifier and purchased a Satellite+TV in line amp, 18 dB, 47 ~ 2400 Mhz. Instead of improving performance, it reduced it. I thought it is defective, so purchased another make from another supplier also 47 ~ 2500 Mhz. That also reduced performance. Finally I purchased a Satellite narrow band amplifier 950 ~ 2050 Mhz (Satellite L-band), and it greately enhanced performance.
Some of the eBay ADS-B LNAs appear to be generic amps, but with SAW filter/s added.
And for the few dollars difference, there is no reason not to go with other LNAs designed specifically for ADS-B.
I have other possible uses for the generic amp, but will try it on ADS-B just for for fun. The extra FA filter will be used as well.
Installed the generic eBay wideband amplifier board and the spare FA filter. After a quick gain adjustment, from 49.6 to 38.6, I’m happy to report that early results are satisfactory.
The reference station is composed of: FA antenna, RTL_SDR Blog triple filtered LNA unit, and an RTL-SDR Blog v2 dongle.
The test station is now composed of: Homebrew QuickSpider antenna, FA filter, generic eBay wideband amplifier, generic no-name dongle.
The two stations are installed in the garage, including antennas, 1 foot apart.
Before the changes, the test setup was always between 45-50% worse than the reference station, for both planes seen and messages. Now, there are times the test station is only 5-10% worse than the reference station in the number of planes seen, and better than the reference station in the number of messages.
Again, very early results, not enough observation time. More gain tweaking likely possible, but for a $8 amplifier board that was in the junk box waiting for some use, the results are encouraging.
Stick with the good stuff, but if you have one of these generic amps in the junk box, give it a try.
As always YMMV.
Update: Settled on 43.9 for gain on the test station. Incidentally, the total gain of both stations is practically the same now. The reference station is 48+27=75. The test station is 43.9+30= 73.9. The 1.1 difference plus 5 dB of the FA antenna, used by the reference station, likely accounts for the variation between the two stations. Very pleased with the results so far. Once again YMMV.
1 Like
just out of curiosity could you swap antennas? 
Found this one on eBay, and the price is right:
The same chip used in the newly released RTL-SDR Blog wideband LNA, minus the case and bias T circuit. The chip is the SPF5189Z with an NF=0.6dB.
Most of the other LNA/Amps have an NF=2dB.
It depends. You have to see the NF at the frequency you plan to use it. But yes, SPF5189Z is very good, a GaAs pHEMT that was impossible to get a few years ago.
I don’t know what mine has, the guy tried to erase it, maybe someone else has better eyes:
PS: Those are taken with my Samsung Galaxy S7 Edge and an LED flash light.
You allured me into this… To be honest I dont need these… but curiosity drove me to buy these.
I expect I will receive these by end of November.
CLICK ON IMAGE TO SEE LARGER SIZE
I ordered both as well, so, I put my money where my mouth is.
It’s a wonderful time for experimenters. Component level experimenting may be on the decline, but these eBay modules are fantastic for those of us with limited soldering and parts sourcing skills.
1 Like
Specs of the amplifier chip SPF5189Z
.
.
The Amplifier Chip SPF5189Z
.
The Finished Product
Download Data Sheet of Amplifier Chip
https://www.mouser.com/datasheet/2/412/pf5189z_data_sheet-781855.pdf
.
@ abcd567 I saw in my searches that they are made by different companies with slightly different characteristics. But yes, it’s about 0.8dB at 1090MHz. Plus the thermal noise of the passive components (you can’t get around that), I think it gets to about 1.8…2 dB in a real life product. The Chinese guys on the Internet give just the noise of the GaAs device, because it looks better, they don’t actually measure them.
A wideband amp has more thermal noise than a filtered, reduced band amp, that’s how equations are working. Also, 50 ohm has less thermal noise than 75 ohm. A hot device has more noise than a cooled device.
What is bad about those preamps is that they don’t have any ESD protection on inputs.
I am pissed that I cannot read my chips. I might buy just the SPF5189Z chip and replace them (or just the first one, that’s the important one).
2 Likes
That’s normally not a problem. Same with components that have mil specs versions, and ‘civilian’ versions.
The real problem with eBay purchased components, and boards as well, perhaps, is counterfeit components.
1 Like
Some manufacturers lie. Or their internal controls are so weak that don’t find when quality drops. Or even if they find issues, “it still works” so they sell that anyway.
When I was working in an electronics lab (couple of decades back), there were several brands that I would trust completely but the rest were “trust but verify”.
I know from experience that measuring the same device from two different manufacturers can yield different results, sometimes far from the catalog.
I find it hard to believe it happens with reputable companies. Are you sure you were not comparing A, B, and C suffixes?
If that becomes widespread, eBay purchases aside, it’s time to return to crochet as a past time.
Define “reputable” 
For me everything from US and Japan was OK. Europe - Philips (NXP now days).
After some research I found out what chips I actually have:
First stage is Qovo’s TQP3M9008. At 1000GHz it has a NF of 1.1dB and OIP3 of 37.6 dBm.
The TQP3M9008 is a cascadable, high linearity gain block amplifier in a low-cost surface-mount package. At 1.9 GHz, the amplifier typically provides 20.6 dB gain, +36 dBm OIP3, and 1.3 dB Noise Figure while only drawing 85 mA current.
The amplifier is internally matched using a high performance E-pHEMT process and only requires an external RF choke and blocking/bypass capacitors for operation from a single +5 V supply.
The TQP3M9008 covers the 0.05 – 4 GHz frequency band and is targeted for wireless infrastructure or other applications requiring high linearity and/or low noise figure.