My SDR is as sensitive as a dead scanner

[Lasse Radio]

I got my SDR dongle today, its a DVB-T with RTL2832 and E4000 bought on Ebay får £11. Installed the software and tried with several antennas. I was able to recieve a broadcast on 107 MHz, a very strong local transmitter. But nothing else. I have an old Bearcat scanner as a reference and tuned in a few strong data transmissions on 400 MHz band without an antenna, but the SDR dongle was deaf as dead duck with all tested antennas. I receive nothing but the very strong broadcast. If this is the normal sensitivity  of this dongle it's nothing but an expensive piece of crap.  Has anyone similar experience? Is my dongle defective somehow?

Regards

Lasse

[Lasse Radio]

Just as a referens if someone can verify that something is wrong with my dongle....  the screenshot below is a radio broadcast on 107 MHz, it's a local transmitter on 1 kW situated about 2 km from where i live in line of sight.  My temporary indoor antenna is about 0.5 meter, so it's not the default tiny one. The signal is down at -90 dB and noise floor at -115 dB. To me it seems abit too little for a strong local transmission....

http://www.radioactives.se/mixmegapol_107MHz_1kW.jpg 

regards

Lasse

[Matt Dodge]

Lasse,

Have you tested with the dongle on a USB extension?  I have mine on a 15' extension.  Found that noise from the PC wipes out most of near by signals near my qth.

Cheers

--
You received this message because you are subscribed to the Google Groups "Ultra Cheap SDR" group.
To view this discussion on the web visit https://groups.google.com/d/msg/ultra-cheap-sdr/-/hDgmefXk-zwJ.
To post to this group, send email to ultra-c...@googlegroups.com.
To unsubscribe from this group, send email to ultra-cheap-s...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/ultra-cheap-sdr?hl=en.

[Lasse Radio]

No I don't have that long extension, only 4' or so. Funny thing, we tested the dongle outside (with a laptop), to get out of the house. My scanner received several local data transmissions without an antenna mounted, the dongle did not, not even a small tiny peak with an antenna designed for that particulat band! It seems like it is allmost deaf, as if the gain slider was at minimum. But it's not, it's at maximum.....

Regards

Lasse

[Paulino Kenji Sato]

Hi,
Verify gain control settings.
Is possible to gain control in manual, and in maximum attenuation.

Disconnecting and Connecting antenna make difference?

Another possibility is a very strong signal in any frequency
obfuscating low signals. The scanner have a filter preventing unwanted
out of band signal to disturbing in band reception.

You is not first to complain about low sensitive of the dongle.

73s
Paulino

[Paulino Kenji Sato]

Hi,

On Mon, Aug 20, 2012 at 6:29 PM, Lasse Radio
<radioactiv...@gmail.com> wrote:
> No I don't have that long extension, only 4' or so. Funny thing, we tested
> the dongle outside (with a laptop), to get out of the house. My scanner
> received several local data transmissions without an antenna mounted, the
> dongle did not, not even a small tiny peak with an antenna designed for that
> particulat band! It seems like it is allmost deaf, as if the gain slider was
> at minimum. But it's not, it's at maximum.....
>
> Regards
>
> Lasse
>

Definitely you dongle is deaf.
Tuner Internal damage or,
Solder Problem or, (try to solve heating entire dongle on mini electric oven)
Short Circuit on antenna path or,
Some parts missing on antenna path.
Visual inspection (take a hires photo) can identify these last two
issues, or a continuity test.






Paulino

[Lasse Radio]

Thanks Paulino, that was what I hoped to hear :).  I have contacted the seller and awaits his answer.  

[Robert Nickels]

On 8/20/2012 4:46 PM, Lasse Radio wrote:
> If all the dongles is as bad as mine, no one would even bother buying
> them and most people would complaint


That's true, and I've measured both of mine to have less than 1 uV
sensitivity from 64-990 mHz. While that's not a lab quality number,
the sensitivity is very good (when properly working of course).

The E4000 tuner RF gain settign can reduce gain by as much as 49dB if
I'm not mistaken, but you still ought to hear plenty of signals even at
the minimum gain setting.

73, Bob W9RAN

[Lasse Radio]

Here are two highres images of the dongle pcb:

http://www.radioactives.se/dongle_2832u_e4000_1.jpg 

http://www.radioactives.se/dongle_2832u_e4000_2.jpg 

There is a strange solder joint where I marked "bad solder". The antenna center pin is definetly in connection with the smt beside it. I'm not sure if they should be connected at all.... 

Regards

Lasse

[jdow]

Odd are it has no protection diodes and has a blown LNA now.

{^_^} Joanne, W6MKU

On 2012/08/20 08:00, Lasse Radio wrote:
> I got my SDR dongle today, its a DVB-T with RTL2832 and E4000 bought on Ebay f�r
> �11. Installed the software and tried with several antennas. I was able to

[jdow]

Ah, if it is THAT close you may find nothing else will make it through.
The dongles have perhaps a 50 dB to 60 dB dynamic range. If the other
stations you want to hear are too far away and too low power the local
station will effectively suppress them into invisibility.

On HF I have effectively the same sort of problem - a VERY strong station
on 1590 (or so) that would blot out most anything else on HF given the
wide open nature of the conversion performed to bypass the E4000 chip.

{^_^} Joanne, W6MKU

[jdow]

Considering that it might not have a protection diode pair on the front
end it's REALLY easy to blow it out with static discharge or even very
strong signals.

{^_^}

> --
> You received this message because you are subscribed to the Google Groups "Ultra
> Cheap SDR" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/ultra-cheap-sdr/-/LCFprYuOWDcJ.

[jdow]

Yup - no diodes. It's blown.

{^_^}

> --
> You received this message because you are subscribed to the Google Groups "Ultra
> Cheap SDR" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/ultra-cheap-sdr/-/d5BXd6N6HukJ.

[jdow]

Look at the other side from what you marked as bad solder. See the single
tab going down with the two tabs below it on either side? That is where
a two diode pack of ultra-high speed diodes should go to protect the
receiver front end. Any static discharge, as with attaching a large antenna,
can blow your front end.

The really cheap units seem to have skipped the diodes. The more expensive
ones do have the diodes. (All three of mine do. But one of them has the
E4000 chip mounted twisted about 5 degrees. The tuner does not work at all.
The vendor replaced that one. Yours is visually OK. So getting it replaced
is likely to be iffy.

(Sadly the RF connector is a bad one for static. I don't think it's shell
contacts before the center conductor. And even then the coax can be charged
leading to a sharp transient even with capacitor isolation into the E4000
chip.)

{^_^}

On 2012/08/20 15:41, Lasse Radio wrote:
> It looks like that is the only way for the antenna center pin to enter the
> circuit. None of the two smt on the belly are connected. So it is probably
> correct.
>
>
> http://www.radioactives.se/dongle_2832u_e4000_3.jpg
> <http://www.radioactives.se/dongle_2832u_e4000_1.jpg>

>
> regards
>
> Lasse
>
> --
> You received this message because you are subscribed to the Google Groups "Ultra
> Cheap SDR" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/ultra-cheap-sdr/-/YgHsSuLQQAsJ.

[Pierre F5OPV]

That's not a bad solder: it's the RF tuner entrance !!!!

regard

---

Date: Mon, 20 Aug 2012 15:28:08 -0700
From: radioactiv...@gmail.com
To: ultra-c...@googlegroups.com
Subject: Re: [ultra-cheap-sdr] Re: My SDR is as sensitive as a dead scanner

--
You received this message because you are subscribed to the Google Groups "Ultra Cheap SDR" group.

To view this discussion on the web visit https://groups.google.com/d/msg/ultra-cheap-sdr/-/d5BXd6N6HukJ.

[Lasse Radio]

Thanks for all your info and comments. It seems like this particular "brand" is nothing worth buying, so question is, which one to buy? Anyone have a suggestion on which to get...  there are several so it's like shooting in the dark not knowing how they are constructed or look inside.

Regards

Lasse 

[Lasse Radio]

Just a quick note, A minute ago I got a message from the seller and they will send a new one without any cost. That's great. Lets hopa that one is better. In any case I'll buy another one, different brand/type...

Regards

Lasse

[jdow]

The one I have two of is a little black square with USB on one end and
antenna (MCX) connector on the other. I had to beat on the vendor a little.
Then he sent me a good one. (First level support wasn't so good. Second
level rectified things.) The bad one had the E4000 chip mounted wrong.

At least it had the diodes. It runs warm due to its small size. And it
does not seem quite as sensitive as the other one I have from a US vendor.
Both were via E-Bay.

{^_^}

> --
> You received this message because you are subscribed to the Google Groups "Ultra
> Cheap SDR" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/ultra-cheap-sdr/-/8Qnt-zVVg_cJ.

[jdow]

They're cheap enough having more than one is handy. (I've actually run
two at once just to prove I could.)

I'm planning on the less sensitive unit becoming a dedicated receiver for
an AcuRite temperature/humidity sensor that transmits on 433.955 MHz give
or take a couple of kHz. I just finished cracking this one's basic code.
Now I get to see if some of the "never changing" bits are really calibration
numbers.

{^_-} Joanne, W6MKU

> --
> You received this message because you are subscribed to the Google Groups "Ultra
> Cheap SDR" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/ultra-cheap-sdr/-/vCtgLf4UbgkJ.

[Lasse Radio]

Thanks Joanne.  As you pointed out, they are so cheap that I'll probably buy a few. My son also want's one or two so it's most likely that we buy a few in case one or two is bad.

Regards

Lasse

[Leif Asbrink]

Hello Lasse,

It could be a software issue. If one sets a low gain the
sensitivity is really poor. Presumably you are under Windows
and use an ExtIO dll. The one I have tried does not
give acceptable sensitivity, but there are other dlls around.

Have a look here:
http://www.sm5bsz.com/linuxdsp/hware/rtlsdr/rtlsdr.htm

By controlling the dongle under Linux where one has easy access
to the source code I can set the gain distribution through
the dongle in an optimum way. Then the sensitivity is quite
good. (NF=7dB) The dynamic range is near the theoretical
limit for the number of bits.

Regards

Leif / SM5BSZ


> Reading posts and reviews of these dongles you get the impression that it
> is acceptable to quite good although with low sensitivity compared to a
> scanner. Most people do recieve transmissions/signals even with moderate
> antennas. I recieve one transmission only. I have gone through the whole
> band 64-1700 MHz and there are no strong signals anywhere, except the local
> brodcast that just barely reach up out of the noise. I do have computer
> generated noise, but thats very low, just a few dB above the noise floor.

> If all the dongles is as bad as mine, no one would even bother buying them

> and most people would complaint.... I don't see much of that compared to
> all the success stories.
>
> Regards
>
> Lasse
>

[Lasse Radio]

I have no means of using Linux at the moment, but I'll check that website.  SDR needs a fast computer and my laptops and netbooks are just below what is recommended. So no point installing Linux on those. The only one avalable for Linux install at the moment is an Acer Aspire one on 1 Gb 1.6 GHz and that probably to weak to handle the flow of information even at low sampling rate. Has anyone tried using this netbook with HDSDR?. 

Regards

Lasse

[Lasse Radio]

I bought two new ones of another type that accordning to the tech description is ESD protected with BAV99. Let's hope they work.. :)

Regards

Lasse

[Leif Asbrink]

Hello Joanne,

The dongle has a dynamic range of 80 dB in 500 Hz bandwidth.
(The standard in ARRLs testing.) The NF is then 10 dB.

With higher gain one can get NF=7 dB with 74 dB dynamic
range. A better idea would be to use a low noise amplifier
at the antenna....

http://www.sm5bsz.com/linuxdsp/hware/rtlsdr/rtlsdr.htm
Table 1.

73

Leif / SM5BSZ

> --
> You received this message because you are subscribed to the Google Groups "Ultra Cheap SDR" group.

[jdow]

Modulo the dongle in question I am forced to question what you
claim the ARRL testing showed.

It's 8 bits. It's dynamic range from a single bit changing to full
dynamic range is 256/1. It's NOT going to give 80 dB. 256/1 is 48dB.
You get another few dB (about 2) from the signal processing. So that's
about a 50dB dynamic range from minimum signal to maximum signal. With
very strong signals look for spurs from that signal to be maybe 50 dB
down. And signals less than 50dBbelow the strong signal are just not
going to be receivable without some exotic noise injection work and
long term averaging. This is the inverse situation of the noise from a
direct digital frequency synthesizer. (And I've over simplified.)

Either you are misreading what the ARRL published or their tests are
heavily involved with flooby dust.

The noise injection technique essentially sums a well known noise
source with the incoming signal and then processes that noise out in
the signal processing. The noise can be a sine wave signal or a PRN
sequence generator. For an example of the first case the AN/USC-28
DISCS satcom transceiver required an oscillator injection somewhat
off frequency when used to its direct sequence spread spectrum signal
for back to back lab testing. Otherwise reception was pretty bad.

{^_^} Joanne, W6MKU

[Leif Asbrink]

Hello Joanne,

The dynamic range in a 2 MHz passband is 50 dB as you say.

It seems you did not care to read any of the information
on the page I linked to.

When the bandwidth is reduced to 500 Hz there is a processing
gain of 2000000/500 = 36 dB so the theoretical limit is 86 dB.

ARRL has not published anything on the e4000 dongle as far as
I know.

The rtl-sdr can receive a CW signal on 144 MHz which is 80
dB weaker than another signal that is present simultaneously
if the frequency separation is 150 kHz or more. At close
separation reciprocal mixing adds noise, but the dongle
can receive a CW signal that is 70 dB below a signal that is
as close in frequency as 10 kHz (provided that the strong
signal is clean enough.) Conventional transceivers are about
25 dB better.

By "can receive" I mean that S/N is 0 dB in 500 Hz bandwidth.
Such a signal is weak, but easy to copy. The equivalent
bandwidth of the human ear is 50 Hz or a little lower so
the S/N at the "decoder" is 10 dB which is usable.
Good EME operators copy signals at much worse S/N.

What you write about "exotic noise injection" is just not true.

It is obvious however that the LO of the rtl-sdr must be
well separated from the signal of interest to avoid the
sideband noise of the LO itself which produces audio noise
by detection in the mixer. That is a well known thing in
amateur SDR. We do not place our interesting signals
on the center spur....

73

Leif / SM5BSZ

[Leif Asbrink]

Hello Lasse,

The rtl-sdr runs fine on my old Pentium 3 (650 MHz) under
Linux. A laptop would have to be very old to not do the job;-)

Leif / SM5BSZ

> --
> You received this message because you are subscribed to the Google Groups "Ultra Cheap SDR" group.

> To view this discussion on the web visit https://groups.google.com/d/msg/ultra-cheap-sdr/-/nKK3Q7ZPX-UJ.

[Lasse Radio]

Thanks Leif for the tip, that sounds very promising!  What version of Linux are you using?  

Regards

Lasse

On Wednesday, August 22, 2012 3:21:51 PM UTC+2, sm5bsz wrote:

Hello Lasse,

[Leif Asbrink]

Hi Lasse,

I tested Debian squeeze on the Pentium 3, but any distribution
should work fine.

Leif / SM5BSZ

[Robert Nickels]

On 8/22/2012 2:49 PM, Leif Asbrink wrote:
> I tested Debian squeeze on the Pentium 3, but any distribution should work fine.

An option, especially if you want to run a fancier DSP like SDR# could
be to separate the tasks. I'm listening to Chicago area air traffic
being streamed via rtl_tcp by a Raspberry Pi via ethernet - which still
only takes 20% CPU utilization ( 1 Ms/s) of a $35 board. This
approach puts a bunch of I/Q samples into a TCP packet, which might
allow a slower PC or netbook to better handle the DSP chores.

Leif, thanks for your excellent performance measurement and analysis of
the E4000-based dongles. I was hoping someone would do this with
proper equipment, and you already have! The performance is truly
amazing for what it is, and it's nice to have actual data to back up
anectdotal "ear measurement". I don't have the proper equipment to make
lab grade measurements but I didn't quite believe it was so good ;-)
I'll be studying your data in more detail...

Could your AGC/noise figure code enhancements be added and recompiled to
incorporate them into rtl_tcp?

73, Bob W9RAN

[Jago Pearce]

Is there a way to tell for sure if the chip is blown? This could be a useful reference for people.

Also, what would be an example of something enough to blow it? Cellphone next to standard ariel? VHF handheld next to standard ariel? Or does it require a big ariel of some kind?

[jdow]

The simplest and safest test is the receiver going numb. In "theory" you
could use a 10k resistor in series with a variable voltage source in
place of the antenna and run the source from 0 to about .6 volt. Measure
the dongle side of the 10k resistor for voltage. It should show a diode
characteristic beginning at about .3 volts and should never get above
about .5 volts. If the measured voltage tracks the dongle side of the
10k resistor it's not protected and you probably have not blown the
E4000. YMMV and I do not warrant the results. This has to be done with
extreme care as simple static electricity can blow the poor babies out.
A cellphone placed beside the antenna might be able to. A ham transmitter
antenna near the dongle's antenna will blow it out. (That configuration
CAN blow out many ham transceiver inputs as a matter of fact.)

{^_^}

> --
> You received this message because you are subscribed to the Google Groups "Ultra
> Cheap SDR" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/ultra-cheap-sdr/-/mIG3ge-w8TwJ.

[darky]

There is a very easy test for that - when you receive your dongle - try to measure the signal strength of FM station which you know using the small antenna which coem with the dongle. Later on you can always re-measure to be sure that it is still working.

 

My dongle droped down in sensitivity with about 20 dB. I damaged it only by touching the antenna connector, ataching home made antennas, etc. It's not that I have strong signals nearby, it was simply statick electricity ...

 

Good luck!