Phase-coherence experiments with RTLSDR dongles

[Marcus D. Leech]

I did some experiments last night with a high-quality (from a Marconi
2955B test-set) master-clock of 28.8MHz to a pair of
RTLSDR dongles, based on the R820T, that had been lightly modified to
accept an external, common, clock.

I then used an external CW signal of 144MHz, and observed the two
resulting carriers and looked at the phase offsets.

What I found was that the phase-offset varied all over the place, over
time-scales of under 1 second. That means that these devices are
wholly unsuitable for phase-coherent applications such as interferometry.

I repeated the experiment with a USRP B100 with a TVRX2 card in it--that
card has two "TV-type" tuners in it (TDA18272), both of which are
connected to
a common 16MHz clock source coming from the B100 motherboard, which
itself uses a 2.5PPM TCXO master-clock. Under the same test
conditions (carrier fed to both sides through a splitter), the
B100+TVRX2 displayed *excellent* long-term and short-term phase-coherence
between the two receive channels. This was an interesting test,
because the TDA18272 tuner is targetted at exactly the same consumer-level
markets as the R820T tuner in these RTLSDR dongles. I was surprised
that there could be such massive inter-channel-phase-noise difference
between the two devices.

I don't have any *quantitative* numbers, but, qualitatively, it seems
like these cheap dongles won't work well for interferometry.


--
Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

[Rolando Paz]

Interesting results Marcus.

You took some pictures of your tests?

I would like to see the B100 + TVRX2 for interferometry with two antennas.

Best Regards

Rolando Paz

2013/9/10 Marcus D. Leech <patchv...@gmail.com>

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
--- You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Marcus D. Leech]

> Interesting results Marcus.
>
> You took some pictures of your tests?
>
> I would like to see the B100 + TVRX2 for interferometry with two antennas.
>
> Best Regards
>
> Rolando Paz
>
>

No, I didn't take any pictures. Perhaps I shall replicate at some point.

I'm busy preparing to go travelling for 10 days, spending time with Ken
Tapping and Hugh Pett, visiting the DRAO observatory, and doing some
back-yard tinkering with Ken and Hugh.

[Wolfgang]

Marcus,

thanks for sharing this with us.
This is unfortunate. Do you think it would be worthwhile trying a different
dongle with a different tuner such as the E4000?

Regards
Wolfgang


-----Ursprüngliche Nachricht-----
Von: sara...@googlegroups.com [mailto:sara...@googlegroups.com] Im
Auftrag von Marcus D. Leech
Gesendet: Dienstag, 10. September 2013 19:06
An: sara...@googlegroups.com
Betreff: [SARA] Phase-coherence experiments with RTLSDR dongles

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to

sara-list-...@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups
"Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an

email to sara-list+...@googlegroups.com.

[Marko Cebokli]

The problem is not in the dongle, but in the way the data transfers are
started. To start the data transfer, a separate system call is needed for each
dongle, and you never know how much time the operating system will put between
the two. If you get swapped to disk after the first call, it could be in the
order of seconds. Also, the buffering done in the dongle, in the USB interface
hardware, and by the kernel driver, is rather impredictable.

Marko Cebokli

[Marcus Leech]

Marko:

I`ll have to think about this.  When I did a similar experiment using a USRP-B100, which is also a USB device, I got excellent phase coherence.

Granted, given that the two USB devices aren`t *started* at the same exact instant, so there`ll be some unknown, random, phase-offset.  But once both devices are streaming at a

  constant rate, buffering-jitter shouldn`t make any difference, as long as samples aren`t dropped.  The applications view is two sample-streams running at a fixed rate.  As long as

  there aren`t sample-drops, then there should be phase-coherence.

--

[Wolfgang]

Thanks for your explanation, that makes a lot of sense to me.
I was contemplating to eventually build a "poor man's pulsar spectrometer"
where you use a larger number of dongles each tuned to a slightly different
frequency, so that you cover a larger bandwidth in total.
Then you could dedisperse by applying proper delay between the time series
recorded for each dongle.
According to your thoughts, the timing jitter however may just be too large
to allow this. The timing jitter in this application would have to be small,
at least better than 10% of the pulse width which again is typically a few
percent of the pulse period. So we are talking about a msec or less
allowable jitter for a moderate pulsar. This may be a too stringent
requirement for a multi-dongle setup.

Regards
Wolfgang



-----Ursprüngliche Nachricht-----
Von: sara...@googlegroups.com [mailto:sara...@googlegroups.com] Im

Auftrag von Marko Cebokli
Gesendet: Donnerstag, 12. September 2013 16:00
An: sara...@googlegroups.com
Betreff: Re: AW: [SARA] Phase-coherence experiments with RTLSDR dongles

[Marko Cebokli]

Hello Marcus,

 

with the USRP you can have both (or four with the original USRP) inputs on the same USB device. The samples from both chans are interleaved in the fpga, and are transported from there on as a single stream, so that their relative timing can't change in the following pipeline - all delays affect both channels exactly the same.

 

With two independent streams, you have a random initial delay between the channels. Indeed, if they are sampled with the same clock as in your case, this relative delay won't change during a single run, provided no samples are lost in either pipeline.

 

One might try to determine the delay by injecting a reference RF signal (pseudorandom noise with repeat time > max delay, for example) into both channels, which would then need to be filtered out (or time-gated out) before the final correlation. To avoid a cable mess, the reference could be broadcast over the air, with a small TX antenna.

 

Or else, if you are not interested in absolute phase, and just want some fringes from a strong source like the Sun, you might simply do a blind search for max correlation.

[Marcus Leech]

Yeah, OK, we're on the same page now.

One has to assume, for two independent, but common-clocked PLL synthesizers, that the relative "starting" phase will be random.  Add to that, whatever phase-offset is caused by timing delay between starting the two streams.  But everything, after that, should "march together".

It is this "marching together" that I failed to see in my experiments, despite having a high-quality common clock source (derived from an OCXO from a lab instrument).

From that experiment, it was clear that the channel-to-channel phase noise was unacceptably high--relative phase was changing on a sub-second basis.

Repeated experiment with a B100 + TVRX2 (trying to replicate some of the overall experiment) yielded much better results.  Night and day results.  There were tiny, high-frequency,

  relative excursions between the two channels, but not enough to disturb an integration of even as little as a second or two.  The B100+TVRX2 might make a good interferometry

  receiver, although it has built-in hardware AGC that can't be turned off, with good up-front filtering, so that the AGC is never tripped by some RFI, it should work out well.

[Marko Cebokli]

Hm, the only reason for such behavior I can see, is that the samples get silently dropped somewhere. Do you see sudden discrete jumps, or is there a continous drift of phase?

 

I have only two RTL dongles, but of different makes (one E4000 and one R820), so I can't try to replicate your experiment. But I'll try to buy more dongles, then I might try.

 

As for the reference signal, simple RF pulses injected in both antennas, one to two us long, and with an amplitude well above the noise would probably work. They would be easy to distinguish from the astro noise by amplitude alone, and cutting them out in the time domain would not sacrifice too much of the desired signal.

[Marko Cebokli]

Another guy doing such experiments:

 

http://kaira.sgo.fi/2013/09/16-dual-channel-coherent-digital.html

 

Marko Cebokli

[Marcus D. Leech]

On 09/29/2013 02:11 PM, Marko Cebokli wrote:

Another guy doing such experiments:

 

http://kaira.sgo.fi/2013/09/16-dual-channel-coherent-digital.html

 

Marko Cebokli

 

 

Yup.  Juha Vierinen.

He's getting much better results than I am.  I puzzled about the reasons.  Still experimenting.

-- 
Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

[Marcus D. Leech]

On 09/29/2013 02:11 PM, Marko Cebokli wrote:

Another guy doing such experiments:

 

http://kaira.sgo.fi/2013/09/16-dual-channel-coherent-digital.html

 

Marko Cebokli

 

One of the puzzlements is that if I tune to somewhere near one of the harmonics of the clock, I get perfect coherence, with the output of the
  correlator strongly proportional to the peak power of the two input signals.

But any signals coming from "outside" aren't coherent in any way.  The correlation tends towards zero, even when I'm feeding from a precision
  signal generator, at various "reasonable" power levels (from -100dBm to -50dBm).

[Alexander J]

http://www.radio-sky.ru/forums/index.php?showtopic=65&view=findpost&p=324 -  This project is a amateur radio interferometer. The author (Alex Plaha) has created true dual-channel coherent receiver.  http://www.radio-sky.ru/forums/index.php?act=attach&type=post&id=228 He joined the output of the second tuner to the first chip.

> Von: sara...@googlegroups.com [mailto:sara-list@googlegroups.com] Im

sara-list-unsubscribe@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to

sara-list-unsubscribe@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to

sara-list-unsubscribe@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+unsubscribe@googlegroups.com.

[Marcus D. Leech]

On 10/08/2013 11:10 AM, Alexander J wrote:

http://www.radio-sky.ru/forums/index.php?showtopic=65&view=findpost&p=324 -  This project is a amateur radio interferometer. The author (Alex Plaha) has created true dual-channel coherent receiver.  http://www.radio-sky.ru/forums/index.php?act=attach&type=post&id=228 He joined the output of the second tuner to the first chip.

??

Can anyone translate the russian?

-- 
Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

[Myamiphil]

Wow what a find!

I have Google translate as a link in my Firefox browser...Its easy to install.  It works ok. Seems they are discovering the power of these RTL dongles . Cascading them.

Its a slow process but take each post and translate it... GT can only translate a max of 8000 characters at a time. But its enought to do one paragraph at a time. You can't translate whats in the pictures though.

They are doing what folks have been talking about here on the SARA list.

" The chart below is a block diagram receiver E4000 shows that the output of its internal crystal oscillator, a divider N buffer stage and launched into the findings 24,23, CKOUTP and CKOUTN respectively. By default, the divider ratio N is equal to 1. If so, the quartz of the dongle No-0, we can not unsolder and powered dongle No-one from the internal oscillator dongle No-0. To do this, connect the output of the generator to the dongle No-0 (pin 23 or 24 on the E4000) to the point where he was installed on the quartz dongle No-one (closest to the pin receiver E4000). Such an option rework further simplify design. So you can poprobovat.Ya at once noticed that the E4000 has Buffered output of the generator, so I had to make a separate crystal oscillator 28.8Mgts that konstruktsiyu.Nizhe little complicated in the last photo are all modifications to the dongle, after which they turn into hardware part of the RTI-2. Then it's up to the software. We put the drivers, run RTI-2.exe and observe the universe in a huge range of frequencies. But a wide range of frequencies implies a corresponding antenna and LNA. Therefore, antennas and LNA for RTI-2, you must devote a separate topik.Posle first light RTI-2 to weather the sun waiting for the precise antenna pointing to other radio sources. The accumulation of observations LebedyaA, Cassiopeia, Virgo and Taurus on the RTI-2 will compare it to the RTI-1 by different criteria."






These are the first drawings and schems I've seen of the board and chips.
Going to the top of the article stream you can get a better idea of how the discussion progresses.

There's also links at the top for software.  It seems to be all the usual stuff. The link after  WIKI

Software radio astronomer
   
      Tools radio astronomer
         
      - Radio-SkyPipe II - On-line recorder. Collection of data on your computer using a sound card or ADC. [Site of the manufacturer]
   
   
      
      Utilities
   
       - Astroplanner - Planning observations nebulous objects (Win., shareware / freeware, only English.). Site of the manufacturer]
       
      - CNebulaX - Scheduler observations to control the telescope (Win., freeware). Site of the manufacturer]
         
      - DeepSky2000 - A powerful scheduler observations (shareware / freeware). Site of the manufacturer]
       
      - DishPerformance - Calculation of minimum detectable signal level (Win., freeware, only English.). [Site of the manufacturer]
     
        - Observation Manager - Manager of observations for amateurs and observatories (Win., freeware). [Site of the manufacturer]
     
        - ProGuider - Patriotic design, best-in-class (Win., eng., Freeware). [Site of the manufacturer]
     
       - Radio Eyes - A tool for planning and implementation of radio. (Win., only English.). [Site of the manufacturer]
   
      - Parabola calculator - A program for calculating the parabola of any diameter and depth. (Win., only English.). [Forum Thread]
      
      Educational Software
     
      - C2A - Powerful class planetarium CartesduCiel (freeware). [Site of the manufacturer]
       
      - Cartes Du Ciel - Planetarium for printing identification cards and faint objects (Win., eng., Freeware). [Site of the manufacturer]
     
      - HNSky - Fast and powerful planetarium (Win., eng., Freeware). [Site of the manufacturer]
        
      - StarCalc - Best planetarium for everyday astronomical problems of Russia (Win., eng., Freeware). [Site of the manufacturer]
      
      - Stellarium - Great planetarium (freeware). [Site of the manufacturer]
   
      - Redshift 5.1 - Legendary interactive planetarium!. [Site of the manufacturer]

Best Regards

Phil

Lat: 40.8367633  Long: -74.1768412

 

---

From: Marcus D. Leech <patchv...@gmail.com>
To: sara...@googlegroups.com
Sent: Tuesday, October 8, 2013 9:01 PM
Subject: Re: AW: [SARA] Phase-coherence experiments with RTLSDR dongles

> Von: sara...@googlegroups.com [mailto:sara-list@ googlegroups.com] Im

sara-list-...@ googlegroups.com

For more options, visit this group at
http://groups.google.com/ group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@ googlegroups.com.

For more options, visit https://groups.google.com/ groups/opt_out.

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to

sara-list-...@ googlegroups.com

For more options, visit this group at
http://groups.google.com/ group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@ googlegroups.com.

For more options, visit https://groups.google.com/ groups/opt_out.

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

-- 
Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

[wezelball]

Wow, this looks very interesting!  I've been wondering if this could be done, this would be an awesome RA project!.  The hardware mods look fairly straightforward, but I'm not sure of the software.

Is anyone in the group going to take a crack at this?

Dave

[Marcus D. Leech]

I've been trying to reproduce Juha's results with phase-coherence and
the dongles I have here. No luck so far.

[Marcus D. Leech]

On 10/09/2013 05:45 PM, wezelball wrote:

What he's done is to take the "I" output of one tuner, and substitute it
for the "I" input on the RTL2832U chip, so that while the RTL chip
thinks its
sampling I/Q from a single tuner, it's sampling "I" from one, and "Q"
from another. Since the two tuners won't have zero phase offset, I"m not
sure why he's doing this, by using *both* dongles independently, with
a common-clock, he can derive the complex visibility. Ah well.

My own experiments haven't yet produced any fruit. I'm feeding in a
common clock, and the output data are completely uncorrelated.

Juha took a look at my data the other day, and, indeed, they're
uncorrelated. Can't figure out why. We're both using the same hardware
and software.

[Raydel Abreu (CM2ESP)]

Marcus,

Good, he took a time to see your data. By my side Juha has no reply none of my e-mails...

Raydel

2013/10/10 Marcus D. Leech <patchv...@gmail.com>

Juha took a look at my data the other day, and, indeed, they're uncorrelated.  Can't figure out why.  We're both using the same hardware
  and software.




--
Marcus Leech
Principal Investigator
Shirleys Bay Radio Astronomy Consortium
http://www.sbrac.org

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-unsubscribe@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
--- You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+unsubscribe@googlegroups.com.

[Marcus D. Leech]

On 10/11/2013 08:45 AM, Raydel Abreu (CM2ESP) wrote:
> Marcus,
>
> Good, he took a time to see your data. By my side Juha has no reply
> none of my e-mails...
>
> Raydel

He's not generally that responsive, I've found.

[Raydel Abreu (CM2ESP)]

Hi Marcus,

Two things....

Can I get a copy of the software or script you are using for phase coherence experiments? I want to try with my dongles, but I don't have the software required.

By the way, I yesterday tried your great program Simple-RA, in an attempt for solar metric observation, but I got no luck, poor antenna gain and high noisy environment. I hope to try again during weekend with better antennas.

I tried with a 4 el yagi and a 30dB pre-amp 1 MegaSamples/sec and 550K for the detector. 12 seconds integration and 2 minutes of spectral averaging during a solar meridian transit. DC gain was 1000 and frequency was 135.5, a small clean slot I found here in Havana. Are this parameters correct, or do you recommend a diferent configuration?

One last thing, how can I open for post processing the resulting data?

Thanks,

Raydel

2013/10/11 Marcus D. Leech <patchv...@gmail.com>

[Marcus Leech]

There's a README file that comes with simple_ra that gives information about post-processing.

You may find that a 4-el Yagi just doesn't have enough gain to allow you to see the Sun at 135MHz.

I'll send you a script perhaps later this evening for the coherence experiments.

sara-list-...@googlegroups.com

For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.

To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

[Raydel Abreu (CM2ESP)]

Great, thanks for the script, I will be waiting for it.

4el is too small, I am working in setting up a L band system, at GHz there is less noise and a dish will provide nice gain, but it is a slow process

2013/10/11 Marcus Leech <patchv...@gmail.com>

[Marcus D. Leech]

On 10/11/2013 11:26 AM, Raydel Abreu (CM2ESP) wrote:
> Great, thanks for the script, I will be waiting for it.
>
> 4el is too small, I am working in setting up a L band system, at GHz
> there is less noise and a dish will provide nice gain, but it is a
> slow process

> \
Here's the .grc script for my coherence tests.

[Alexander J]

When using Google Translate, make sure that there is a space between sentences.
Example in quote: "So you can poprobovat.Ya at once noticed that the....". In the original "So you can try. I immediately noticed that...."

среда, 9 октября 2013 г., 8:02:20 UTC+3 пользователь myamiphil написал:

[Malcolm Mallette]

 

This mail list has recently had more messages about software than radio astronomy.  What is happening?

 

The main reason for software problems appears to be a lack of business knowledge by software developers.  There are two rules that must be followed in the choice of an operating system to develop your software for:

 

1. If you develop software for your own use, develop it for the operating system that you prefer and that you use.

If you use Linux as your favorite operating system, develop the software for Linux.  If you use Apple OS as your favorite operating system develop the software for Apple OS.  If your favorite operating system is Windows, develop your software for Windows.

 

2. If you are developing software for others to use, either planning to sell it to them or to give it away for free, develop the software for the operating system or systems most people use.

Windows is the operating system most people use.  Software developed for others to use should be first written for Windows.  It does not matter what operating system you think is best, Windows is used by most people.  Then, if you wish, develop it for the Apple OS and then for Linux.

 

Windows is the  choice of most computer users.  My firm had to use Windows as that is what most of our clients used. I did not want to shift operating systems from the office to home.  Most people are in a similar position.

 

Again, if you write software for your own use, write it for your favorite operating system.  If you are writing it for others to use, write it for Windows, and, if you wish, create versions for other operating systems. When writing software for others, forget your geek self and think like a business man.

 

The two rules also apply to making devices to be controlled by software.  You sell more if most people can run the software.

 

Malcolm

http://radioastronomy.uindy.edu

 

 

 

 

 

 

 

[Alexander J]

Computers are cheap.  For example http://cubieboard.org/ - incredible thing 10 years ago. Today it is a reality. This computer can be installed close to the antenna and the observer receive data via the Internet. No need for Windows. Only hardware and free software - Linux, Android.

суббота, 19 октября 2013 г., 17:11:00 UTC+3 пользователь Malcolm Mallette написал:

[Marcus D. Leech]

On 10/19/2013 10:11 AM, Malcolm Mallette wrote:

 

This mail list has recently had more messages about software than radio astronomy.  What is happening?

 

The main reason for software problems appears to be a lack of business knowledge by software developers.  There are two rules that must be followed in the choice of an operating system to develop your software for:

 

1. If you develop software for your own use, develop it for the operating system that you prefer and that you use.

If you use Linux as your favorite operating system, develop the software for Linux.  If you use Apple OS as your favorite operating system develop the software for Apple OS.  If your favorite operating system is Windows, develop your software for Windows.

 

2. If you are developing software for others to use, either planning to sell it to them or to give it away for free, develop the software for the operating system or systems most people use.

Windows is the operating system most people use.  Software developed for others to use should be first written for Windows.  It does not matter what operating system you think is best, Windows is used by most people.  Then, if you wish, develop it for the Apple OS and then for Linux.

While for purely-commercial applications, this may be reasonable, I have to strongly disagree for free/open-source software.  People giving their
  software away are, for the most part, developing it for *themselves* first, and being altruistic in giving it away.   If a person is developing software
  on their own time, for free, they're unlikely to develop for an operating system environment that they either dislike, or are not comfortable with.

 

Windows is the  choice of most computer users.  My firm had to use Windows as that is what most of our clients used. I did not want to shift operating systems from the office to home.  Most people are in a similar position.

I work for a software company that develops for:

    o Windows
    o Linux
    o MacOS
    o Android
    o iOS

The software landscape is actually *much* more diverse than you might think.

Furthermore, in *science*, a *huge* fraction of the systems people use are Linux or *BSD-based.  

 

Again, if you write software for your own use, write it for your favorite operating system.  If you are writing it for others to use, write it for Windows, and, if you wish, create versions for other operating systems. When writing software for others, forget your geek self and think like a business man.

Giving software away isn't a business.  It's supposed to be a pleasure.  Most people who write free software are also professional software developers,
  where they have to "think like a business man" most of the day.  Why should they have to carry that bull-crap over into their *recreational*
  programming activities?

 

[Paul Oxley]

Malcolm

 

I agree with your conclusions. We are using this philosophy on the development of software for RASDR. The target platforms include Windows, Linux and MAC-OS. We together with a vendor who is building the hardware are using software tools that can be compiled into the three platforms with only minor changes in code. I think you will like the result which should be available soon.

 

Paul Oxley

[Malcolm Mallette]

Paul,

 

Great!!!

 

It  is very important that Mac and Windows users can just install the program in the manner other programs are installed.  If they have to “build” or compile anything most will give up. 

 

I did software that works with our files to adjust for gain changes due to temp, combine observations etc.  That is on our webpage as a Python Script.  That way Mac and Windows and other OS users can download Python for their OS and then use the script to run the program that works with our files. All our observation files since 2009 are on the web.

 

Malcolm

[Christopher Hart]

Hi All,

I've been a lookie-loo on this list so far, but I had to chime in on this one.

If SARA members aspire to be like professional astronomers, they should have zero issues compiling software packages for UNIX-based OS'es.

Marcus' post is dead right when it comes to scientific environments.
The only place I disagree with him is that I never think like a business-man, it's the only way I'm able to be creative. :)

Respectfully,

Chris Hart, MBCS
Senior Software Engineer,
ZeroLag Communications, Inc.
http://www.zerolag.com

[Wolfgang]

From: Malcolm Mallette <mall...@comcast.net>
To: sara...@googlegroups.com
Sent: Saturday, October 19, 2013 10:11 AM
Subject: [SARA] The Business of Software Development

  

.....

1. If you develop software for your own use, develop it for the operating system that you prefer and that you use.

If you use Linux as your favorite operating system, develop the software for Linux.  If you use Apple OS as your favorite operating system develop the software for Apple OS.  If your favorite operating system is Windows, develop your software for Windows.

......

 

In my observation, this is what happens in almost all cases in radio astronomy (and science in general). People in the scientific and amateur community develop software primarily for their own purposes. If other people find it useful, they make it available "as is". Sometimes the software is well documented and easy to install, in other cases documentation may be minimal and/or installation can be painful.

In the science community nowadays almost all computing is based on Linux type OSes. Therefore the majority of software is based on Linux.

 

Things seem to look different in the amateur radio world. I am probably not the right person to judge this but it seems that this community does more development under the Windows environment.

 

Only in rare cases software has been migrated to all major platforms because the software has not been intended to be "marketable" to a broad audience. As a result, you may have to live with what is available from other authors and their choice of the operating system. So your personal choice will be based on what you want to use from other authors, what you need to develop by your own and what you feel comfortable with. But you need to make this choice and live with the consequences unless you want to support two or more OS at the same time for your radio astronomy work.

 

At our Stockert radio telescope, the choice has been to go for a almost pure Linux environment for the following reasons (in sequence of importance):

- Development skills and OS knowledge were much more readily available among the people in our group

- Lots of software was available for the Linux environment where it was hard to find a Windows equivalent

- Some specific software which was indispensible for our work (in particular pulsar recording and processing software) was only available for Linux and could never ever be migrated to Windows

- Licensing cost

 

Sometimes the discussion Linux vs Windows vs Mac takes an almost religious dimension and gets emotional ;-)  I believe this is not necessary, just make your choice and let other people live with their choice.

 

Regards and happy computing!

Wolfgang

 

 

 

 

 

[Paul Oxley]

Malcolm & Wolfgang

 

I am not sure who said what in this message, but let me respond.

 

My own preference is to avoid Linux at all cost. This is not an emotional choice, but rather one based on experience. Although I am personally capable of dealing with most any platform. I don't like the overhead associated with Linux. You are always looking for some missing piece of code and don't have time to focus on what you want to accomplish. If you hobby is to make Linux work, great. It isn't mine.

 

However, I agree that software should be able to be used on any platform of the user's choice. The use should be as simple as possible to allow the user to use the software, not experiment with things such as compiling code, finding the missing link, etc.

 

When I pointed out that our objective developing RASDR is to have the software (and hardware) work across the big three platforms, this does not mean that we would expect the user to do the compiling. We intend to release precompiled versions of RASDRviewer C/C++ Code  (our main software) that can be used out of the box on a suitable platform. However, for the user who wants to write code, we will also release the source code.

 

The other big variable in user's machines is the size and speed of their systems. To address this variability, we intend to have user tunable parameters that can optimize the performance of the software on most machines. For the high end user with a big fast machine that has a USB3 connection, the options would include the full capability of the hardware to collect data at 28 M Samples / Second with a corresponding bandwidth of 28 MHz. For the slower smaller machines perhaps without USB3, the performance parameters can limit the capabilities, but allow the use of the software without encountering congestion or system failures. The big/fast machines will also be able to use a very large FFT with 16,384 data points in real time. Since the FFT operates in real time, it can also be used to perform functions such as digital band pass filtering. The smaller machines may need to limit the size of the real time FFT.

 

For users that want to do further post processing, we plan to have the capability of spooling the source binary data to a disk or solid state drive. The size and speed of the spooling will also be tunable to match the user's machine and drives. We also are considering spooling the data to a Ethernet connection in a packetized format to allow the user to attach multiple PCs to a network and distribute the processing across the different machines.

 

Outputs are also planned to allow the user to interface with the FFT output data, sample data and Power over Time data. These outputs would likely be in a ASCII format with the option of a .csv file for importing into Excel. The plan also includes the capability of producing time stamps that are based on a GPS input.

 

Since we are still in the development stage, we would appreciate any potential user input on what would work best in their environment. If you have a format for the data that would work best for your application, let us know the format. For example, if you want to be able to use the standard math processors, or Python coded programs, let us know the format needed.

 

I would also like to address some of the list of reasons for using Linux. First, the skill level to write code is nearly identical across the platforms. If you look you can find suitable tools for all three. You need to avoid the proprietary (even in LINUX) tools that do not work across platforms. The choice of language that will work across platforms include C/C++, Python and even Fortran. Using capabilities of the languages such as preprocessor switches can allow the selection of certain processes to match the platform. In RASDR, at this time the preprocessor selection is limited to the selection of the drivers necessary for the USB3 interface. For windows, the driver is written by Cypress who makes the FX3 USB3 chip. For other platforms, the driver uses the USB library functions contained in Linux.

 

We also have chosen to use GUI functions (wxWidgets and Open GL) that can be used across platforms. It was necessary to avoid some of the Windows or Linux only GUI interfaces. As a result, the look and feel of the software should be the same on all platforms. The downside is that if you want to use the supplied source code for your own modification, you will need to setup the tools to use wxWidgets/OpenGL on your choice of platform. From experience, this is not a trivial task.

 

Our intent is to provide the capability of getting data into post processing programs for functions such as Pulsar analysis in the best way possible.

 

We don't have any Licensing Costs. We use the Open Source Licensing approach.

 

Paul Oxley

RASDR Team

[WILLIAM & MELINDA LORD]

Paul,

We appreciate your approach to allow the RASDR software to be used across the big three platforms without requiring users to have to write code.  Most of our members work full time and do radio astronomy as a hobby.  They want to be able to get to the observing and analyzing data rather than spend weeks on computer code.

Thanks for your diligent work on this project.

Bill
 

Bill & Melinda Lord
www.tnSkyNet.com
www.radio-astronomy.org

Light travels faster than sound. This is why some people appear bright until you hear them speak. :-)

---

From: Paul Oxley <oxl...@att.net>
To: "sara...@googlegroups.com" <sara...@googlegroups.com>
Sent: Sunday, October 20, 2013 10:29 AM
Subject: Re: AW: [SARA] The Business of Software Development

[Wolfgang]

Paul,

maybe I was not very clear in what i was trying to say, so in a nutshell:

 

I support Malcom's view that if you develop software for your own you choose the environment that suits you best.

Malcom' suggestion to develop software for all major OSes, however, may be very desirable but will not materialize especially in the scientific community as people are just trying to get things done and then move on to the next task. This is probably why much software is available for either Linux of Windows only. Therefore you are forced to decide for yourself.

 

I observe your great work with the RASDR and it good to see that you will support all major OS. This is a great contribution to the community as this puts everybody in a position to use your work in his environment, no matter what his choice might have been.

 

There are good and valid reasons why you (and others) have chosen to use Windows as your prime environment. Other people (like ourselves) come to the conclusion that Linux is the better choice for them. This was my last point I wanted to make: There are always good reasons why somebody chooses to go one way or the other. I am amazed that in some cases people get into an emotional discussion about pro's and con's of Windows vs. Linux while it should be just a decision based on individual circumstances.

 

Having said all this, we certainly will not forget (thanks for Bill pointing this out) that our prime goal is to do observations rather than writing code ;-)

 

Regards

Wolfgang

 

 

 

[Marcus D. Leech]

From: Malcolm Mallette <mall...@comcast.net>
To: sara...@googlegroups.com
Sent: Saturday, October 19, 2013 10:11 AM
Subject: [SARA] The Business of Software Development

  

.....

1. If you develop software for your own use, develop it for the operating system that you prefer and that you use.

If you use Linux as your favorite operating system, develop the software for Linux.  If you use Apple OS as your favorite operating system develop the software for Apple OS.  If your favorite operating system is Windows, develop your software for Windows.

......

 

In my observation, this is what happens in almost all cases in radio astronomy (and science in general). People in the scientific and amateur community develop software primarily for their own purposes. If other people find it useful, they make it available "as is". Sometimes the software is well documented and easy to install, in other cases documentation may be minimal and/or installation can be painful.

In the science community nowadays almost all computing is based on Linux type OSes. Therefore the majority of software is based on Linux.

 

Indeed.  I find the position of "you should focus on Windows first, and then maybe deal with the others if you have time" to be a
  profound misunderstanding of what most open-source/free software is all about.

Let me use a couple of analogies.    There's a guy who is passionate about making harps in his spare time.  So passionate, in fact, that he gives
  them away to anyone who wants a harp or wants to learn to play the harp.   But you criticize him because, well, "most people play guitar or violin
  or piano, you should be making those instead and giving them away.  What a profound disservice to the musical community to be giving away harps."

You are regularly invited to your friends house, who put on a fantastic dinner party, usually with lots of wholesome, organic food, often from unusual
  cultural heritage.  You tell them "most people eat hotdogs and hamburgers.  You should think more like a restaurant, and serve hotdogs and
  hamburgers.  You'll never get anywhere giving away this hippy-trippy organic stuff."

[David Fields]

Well said, Bill -- Thanks, Paul, for all you're doing.

Paul, I wish we'd included those thoughts of yours in the latest RASDR update for the SARA Journal.

My viewpoint on the RASDR software is that we release first for Windows. Tweaking for the other platforms will require more work/time than we might think, but there will be a lot of people willing to help when they have high-capability hardware in their hands.

David Fields
RASDR Team

[Steve Olney - VK2XV]

G'day Marcus,

While I think that the use of non-Windows OSes is perfectly justifiable for the reasons given, for me there is another side to the coin.   For those of us (or maybe just me who is an amateur programmer...:-) who program under Windows, as it does the job adequately, there can be a degree of mild derision for our efforts from some quarters.  While not in any way suggesting any is coming from your direction, please allow me to use the two analogies to convey the nuance of that attitude...

    Analogy #1: There's a guy who is passionate about making harps in his spare time.  So passionate, in fact, that he gives

  them away to anyone who wants a harp or wants to learn to play the harp.   But you criticize him because, well, "most people play guitar or violin
  or piano, you should be making those instead and giving them away.  What a profound disservice to the musical community to be giving away harps."

    

    Flipside #1: The passionate harp guy infers that people who don't play a harp are not real musicians and so any music played by them is of inferior quality. 

Analogy #2: You are regularly invited to your friends house, who put on a fantastic dinner party, usually with lots of wholesome, organic food, often from unusual  cultural heritage.  You tell them "most people eat hotdogs and hamburgers.  You should think more like a restaurant, and serve hotdogs and

  hamburgers.  You'll never get anywhere giving away this hippy-trippy organic stuff."

   Flipside #2:   The friends think that those who eat hotdogs and hamburgers are not gastronomes and so should be ignored. 

While I don't cry myself to sleep over these devaluations, it can be annoying - much the same as when you rightly get annoyed when people don't acknowledge your previous good work as you mentioned recently.

Cheers

Steve, VK2XV 

[Marko Cebokli]

HM, "start to thnk like a businessman"

 

Anybody acting according to the above maxim, will NEVER go and write radio astronomy software.

 

There just are "bilyuns and bilyuns" of better ways to make money. If you think like a businessman, radio astronomy will certainly be one of the very last things, that will come to your mind.

Certainly there ARE sciences that can make you rich, like pharmacy, but radio astronomy is definitely for the "severely business challenged".

 

Marko Cebokli

 

 

[Marcus D. Leech]

True.

But I'll point out that with the SKA project looming large, getting a non-trivial fraction of that capital-infrastructure "business" wouldn't
   likely be non-lucrative :)

[Don Latham]

Darnit Marcus, harp music makes me think of death
and hippy food gives me heartburn...
:-) :-)
Don

Marcus D. Leech

> --
> --
> You received this message because you are subscribed to the Google
> Groups "Society of Amateur Radio Astronomers" group.
> To post to this group, send email to sara...@googlegroups.com
> To unsubscribe from this group, send email to
> sara-list-...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/sara-list?hl=en
> ---
> You received this message because you are subscribed to the Google
> Groups "Society of Amateur Radio Astronomers" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to sara-list+...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>


--

�The power of accurate observation is commonly called cynicism by those
who have not got it.�
-George Bernard Shaw



Dr. Don Latham AJ7LL
Six Mile Systems LLC
17850 Six Mile Road
POB 134
Huson, MT, 59846
VOX 406-626-4304
Skype: buffler2
www.lightningforensics.com
www.sixmilesystems.com

[Marko Cebokli]

On Sunday, October 20, 2013 04:58:38 PM Marcus D. Leech wrote:

HM, "start to thnk like a businessman"

 

Anybody acting according to the above maxim, will NEVER go and write radio astronomy software.

 

There just are "bilyuns and bilyuns" of better ways to make money. If you think like a businessman, radio astronomy will certainly be one of the very last things, that will come to your mind.

Certainly there ARE sciences that can make you rich, like pharmacy, but radio astronomy is definitely for the "severely business challenged".

 

Marko Cebokli

True.

But I'll point out that with the SKA project looming large, getting a non-trivial fraction of that capital-infrastructure "business" wouldn't
   likely be non-lucrative :)

Sure, but very little of that money will go to the radio astronomers, most will end in the hands of constuction contractors and "the people in between"

-- Marcus LeechPrincipal InvestigatorShirleys Bay Radio Astronomy Consortiumhttp://www.sbrac.org

[Wolfgang]

.......

Sure, but very little of that money will go to the radio astronomers, most will end in the hands of constuction contractors and "the people in between"

........

I heard an interesting remark recently when we had a group of master students from the University of Bonn for hands on training at our telescope: They were asking about the likelihood of finding a job in radio astronomy, and the professor responded: "The SKA will probably need about 500 radio astronomers, and there are only around 500 professional radio astronomers at the moment worldwide".

If this is true, at least quite a few young astronomers will be able to earn a living from radio astronomy which certainly is good news.

Regards
Wolfgang

  

[Mario Cannistrà]

Hi Marcus, Marko, all.

I know I'm replying to an old thread and hopefully I will not write something already known.

I see this thread shifted at a certain point to the software topic abandoning the main conversation about the russian mod.

I tried to read that forum and I think there's an important consideration in post #22  ( http://translate.googleusercontent.com/translate_c?depth=1&hl=en&rurl=translate.google.com&sandbox=0&sl=ru&tl=en&u=http://www.radio-sky.ru/forums/index.php%3Fshowtopic%3D65%26view%3Dfindpost%26p%3D546&usg=ALkJrhgFcJ7Q9qw9sWGRkuNbwdawSC79ng )

Reading the google translation is really difficult but, my understanding is that he has seen the exact issue you described about misaligned outputs from 2 USBs and so he decided to "merge the signals in one dongle" just to get only one data stream from only one USB....and that apparently solved the issue.

Did you already consider repeating your experiment with one clock source to two R820T dongles this time connected in the "cascaded" way ?

The russian blog roughly describes the E4000 approach, does anyone here have a clear "diagram" or steps to approach this with the R820T ?

Regards,

Mario

On Thursday, October 10, 2013 7:16:42 PM UTC+2, Marcus wrote:

On 10/09/2013 05:45 PM, wezelball wrote:
> Wow, this looks very interesting!  I've been wondering if this could
> be done, this would be an awesome RA project!.  The hardware mods look
> fairly straightforward, but I'm not sure of the software.
>
> Is anyone in the group going to take a crack at this?
>
> Dave
>
What he's done is to take the "I" output of one tuner, and substitute it
for the "I" input on the RTL2832U chip, so that while the RTL chip
thinks its
   sampling I/Q from a single tuner, it's sampling "I" from one, and "Q"
from another.  Since the two tuners won't have zero phase offset, I"m not
   sure why he's doing this, by using *both* dongles independently, with
a common-clock, he can derive the complex visibility.  Ah well.

My own experiments haven't yet produced any fruit.  I'm feeding in a
common clock, and the output data are completely uncorrelated.

Juha took a look at my data the other day, and, indeed, they're
uncorrelated.  Can't figure out why.  We're both using the same hardware
   and software.

[Alexander J]

Translation with PROMT:

#1

 ...in a mode of a two-channel ADC.

 Now we can transfer to the computer only a stream of 8 bit quadratures at frequency of discretization nearby 3 MSPS.

Besides receiving 8 bit quadratures there is a true possibility of digital processing of a signal, is not worse than at DSP or FPGA. 

 For example, in case of strong hindrances around frequency of observation it is possible to realise the digital filter with very narrow band pass.

 

 #5

 Software part it is realised thanking completely open project where all source codes for work with DVB dongles are laid out. 

 For it special gratitude to developers from Osmocom.

 

 #6

 From receiver E4000 block diagramme it is visible that the out of its internal quartz generator, through divider N and the buffer cascade is deduced on pins 24,23, CKOUTP and CKOUTN accordingly. By default the ratio of divider N is equal 1.

 If it so quartz from dongle No-0 does not need to be unsoldered. 

 And to switch dongle No-1 to the internal generator dongle No-0. 

 For this purpose it is necessary to connect a generator out on dongle No-0 (a pins 23 or 24 on Е4000) to a point where quartz on dongle No-1 has been soldered earlier in(contact near  receiver E4000 ).

 Such variant will even more simplify a design.

 

 #7

On a photo, are shown as real changes look in more details.

On the first photo tap of quadratures from Е4000 dongle No-1 is shown. 

On the second photo input of quadratures from Е4000 dongle No-1 on rtl2832u dongle No-0 is shown. 

On the third photo - as all looks together.

#17

From first two drawings it is visible that the amplitude of interference petals at supervision М1 is more at RTI-2, than at RTI-1.

Probably, the reason in big sensitivity of a new radio telescope, RTI-2. The further tests should confirm it.

On noise RTI-2, too it has appeared better RTI-1 (RTI-2  project with USB RTL,  RTI-1 -  first project of radio telescope).

On the diagrams received with RTI-2, the green curve corresponds to the 8-bit correlator, and red 1-bit. 

At visual comparison it is possible to draw a conclusion that for powerful radiation sources, correlation diagrams (1 and 8 bit), are almost identical. Difference in that that the 8-bit correlator has peaks, emissions, and at 1-bit that are absent. 

The 1-bit correlator means carries out also function a Low-pass filter that is extremely necessary at observation in a zone of hindrances. 

#19

Having burnt the second dongle, has decided not to experiment any more without protective diodes on an input.

...

1. Has installed in a path cut, in front of the protective diode the condenser 100 pF. Thus a LNA power on the central wire of a coaxial will not burn the protective diode a dongle.

2. To cut paths with quadratures did not become, has simply removed condensers through which quadratures went on rtl2832u, and on their place has soldered wires for carrying over of quadratures on main dongle. 

#21 

Lokkie: it is probably intelligible and clear to explain what for carrying over of quadratures from one receiver to another and how works with this software is carried out? I can not understand what for it it is made? Thanks!

#22

The matter is that having made the general clock generator for two dongles we receive coherent receiving of a signal on both receivers E4000.

Quadrature outs from these receivers we digitize  too synchronously through rtl2832u.

I.e. the coherence of signals from two aerials passing through different dongles remains.

...

For our connected radio telescope-interferometre RTI-2 I have found a simple way out, have got both signals from both receivers only on one ADC rtl2832 by carrying over of quadratures with one dongle, on an ADC of another and have started up them through one USB port. I.e. in such scheme the data from both aerials arrives only through one USB port synchronously, and thus the coherence of the digitized signal remains.

USB the port is used by the second simply for a power of the second a dongle and in date transmission does not participate. 

After such alteration at once it was possible to find out correlation of signals from different aerials.

But the reasons of loss of coherence of the digitized signal by its transfer through different USB ports remain for me secret. 

[Marko Cebokli]

The problem with these dongles is, that there are no good datasheets for either the tuner chips or the demodulator/usb interface chip.

(I could never really understand what a chip manufacturer can gain by hiding the datasheets of his chips???)

 

If I understand right what Alex did:

the demodulator chip has more inputs than needed for a single tuner - maybe it has an option for I/Q baseband, but the tuners work in a low-IF real samples mode - anyway, it is possible to connect two tuner chps to one demodulator chip. This removes the problem of two independent USB streams, so when the chips are running from the same clock, coherent sampling is possible.

 

I do not plan to try interferometry with RTLSDR dongles, as SIDI works well, with ten times the bandwidth, and gives me enough opportunuty to play for the next few years :-)

I also hate doing reverse engineerng, not out of respect for IP, but because people keep bringing me dead gadgets and expect me to repair them without dcumentation :-)

 

Marko Cebokli

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Mario Cannistrà]

Thank you Alexander for the full translation. I will try to grasp some info from that.

Regards,

Mario

[Mario Cannistrà]

Hi Marko.

Thank you for your kind reply.

Some time ago I've found 4 PDFs at this URL:   http://www.avionics-bangalore.com/viewtopic.php?f=14&t=108

Probably you've already seen them since you're referring to "good" datasheets and the 2832 one is actually...a portion of the original I think.

I will try to experiment using the info available until now.

I browsed your pages describing SIDI and I understand you already have a complete and actually wonderful receiver. And good antennas as well.

That kind of HW build is unfortunately out of my time availability.  I would really prefer to start from available building blocks, possibly cheap to allow me a better start investing time on antennas, sw and obviously the whole thing setup selecting the best frequency(es) for my environment and space availability. The more I read the more I need to stabilize on the best selection to avoid analysis-paralysis  :-)

At the moment I prefer to stick to dongles experimentation that I hope will also help me closer understand most of the involved needs. I still have tons of books and papers to read... :-)

Kind regards,

Mario

[Mario Cannistrà]

I also have R820T dongles only and I also have a couple of TCXO that I will try use on 2 of them.

About Alex: I actually wrote him yesterday early afternoon through that forum "send message form".

I hope to get an answer soon since I'm really curious about the actual implementation.

Yes, lower frequencies could be interesting and I plan to run again a wider spectrum survey to see my RFI "landscape".

I just have a 16 x 4 meters space that I could use for not so spaced antennas or for one antenna array (or 2).  

I will have to play with frequencies and antenna sizes to leverage this space...

Will see...

Regards,

Mario

On Mon, May 12, 2014 at 11:04 PM, Marko Cebokli  wrote:

Hello Mario,

 

thanks for the link! The datasheet for the e4000 is quite good, but this chip is out of prduction, and I only have R820 based dongles. The R820 datasheet is the same I had before, does not describe the registers.

 

As far as I can see from the pdf's about the RTL chip, it has a single low-IF input, so I can not see how you can get it to read two independent channels.

But then, E4000 has I/Q baseband output, so the pdf's about the RTL chip may not be correct? As I said, hard to work in such confusion. We'll have to wait for some more "chipyleaks"!

 

One advantage of the dongles over SIDI, if they are capable of interferometry, is the coverage of frequencies below 950MHz. On the lower frequencies, their lower bandwidth might be less important, as you are unlikely to find many MHZ of clear frequency space there anyway.

 

But the skill and time needed to modify the dongles for interferometry probably wont be much less than for the modification od DVB-S tuners for SIDI?

 

I won't try dongle interferometry in the near future, because I would like to prepare a talk about SIDI for the just announced European Conference on Amateur Radio Astronomy, and want to accumulate some more results until then.

 

You might try to contact Alex Plaha directly, to discuss the dongle interferometry experiments he has done. I never contacted him, so I do not know whether he speaks English.

 

Marko Cebokli

[Marko Cebokli]

Here is a site with LOADS of info about LNB hacking:

 

http://www.g4hjw.metahusky.net/

 

Marko Cebokli

 

[Marcus D. Leech]

Speaking of which, I just took delivery of a pair of PLL321S-2 Ku-band LNBFs today.  After I went out and bought a T9 Torx driver, I opened one
  up, and indeed it uses the RDA3860M Ku-band front-end chip, which uses a 27MHz Xtal as the reference.  It should be very straightforward
  to phase-lock a pair of these with an external 26-27MHz clock.

[Wolfgang]

Mario, all,

I just came across a blog entry about a multi-dongle coherent receiver which is very interesting: http://yo3iiu.ro/blog/?p=1450

Mario, since you seem to have posted a comment on the blog: Have you tried to reproduce the results?

Regards

Wolfgang

 

 

 

---

Von: sara...@googlegroups.com [mailto:sara...@googlegroups.com] Im Auftrag von Mario Cannistrà
Gesendet: Freitag, 9. Mai 2014 17:30
An: sara...@googlegroups.com
Betreff: Re: AW: [SARA] Phase-coherence experiments with RTLSDR dongles

[Mario Cannistrà]

Hi Wolfgang, all.

I'm still very busy on the job and will be until the end of the worldcup (I work in the sports digital media industry).

I did not try reproducing the results in that blog, not only because of my lack of free time, but also for some additional considerations I made about the hw price and other info I've got from the web since mid June.

Will try to summarize:

- that clock generation/distribution board is really interesting but the price is about $ 150 : I would like to keep the budget low, always trying to follow the "do more with less" idea

- I was modding an RTL SDR with a TCXO but I paused in the middle of the mod

- I've found on twitter what seems to be a cheaper clock generator (programmable via Arduino). Tweet with picture here : https://twitter.com/microbuilder/status/476708044374802433  

- it seems Adafruit will start selling that after the internal testing phase

- I wrote Adafruit today kindly asking for news and details (i will post here the reply)

- I also found sw and hw info for a breakout using the same IC (Si5351) by NT7S   http://nt7s.com/2014/06/si5351-libraries-and-breakout-board/

- just asked him about using TCXO instead of crystal and specific outputs features  (i will post here the reply)

- his version has transformer-isolated outputs (Adafruit one seems to have no isolated outputs)

- If this chip can be programmed for same frequency on 3 outputs, I think it could be a really cheaper and interesting alternative, even adding a small arduino for easier frequency management.   i would then save lots of money for other dongles, antennas, etc 

- if anyone here has some time to go through the specs, here is the pdf link:   https://www.silabs.com/Support%20Documents/TechnicalDocs/Si5351.pdf

I'm sorry for me no longer posting in the SARA forum ... hope to restart my activities on RA around mid July :-)

Ciao all,

Mario