1420MHz VLBI

[Rob Barter]

David,

I see in your project notes you mentioned VLBI and GPS disciplined oscillators.  Have you taken this any further at all for 1420MHz interferometry?  Very interested if it's viable for an amateur setup.  Computing power nowadays is not an issue.

Cheers

Rob

[David Lonard]

Hello Rob,

I did some VLBI, which I would define as using two physically separate clock sources, over short baselines of about 100 meters, and at a low frequency (50 MHz) a few years ago. I didn't try moving my radios further apart to different locations, so I can't say for sure if it can be done. I would look into the KiwiSDR which might be able to do long baseline interferometry at HF frequencies. Ionospheric delay at HF and low VHF is pretty large, likely dominating in terms of signal delay over any clock issues.

For all my high frequency work, I now use a single common clock with long coax runs and I think at 1420 MHz and beyond, VLBI will be much harder. With a common clock, I can get good correlation even at 12 GHz. I've purchased a XTRX Octopack which has an integrated GPS and can do timestamping, so it might be possible to try VLBI with another user who also has a XTRX. It would work a lot like the KiwiSDR but at higher frequencies with much higher bandwidth, but it would be best to save quadrature IQ data on a SSD that would then shared between each station.  A XTRX will fill up a 1 Tb SSD in about 10 minutes. 

David

--
You received this message because you are subscribed to the Google Groups "Amateur radio interferometry" group.
To unsubscribe from this group and stop receiving emails from it, send an email to amateur-radio-interf...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Rob Barter]

Thanks David,

It does sound like the challenge is to get accurate enough gps disciplined clocks with very low phase noise to support a 21cm capture.  The cost might be prohibitive.

I noticed you mentioned the XTRX before.  I found a price mentioned, around $2900, but can’t find anyone actually selling it?

Rob

[Leif Johansson]

Skickat från min iPhone

15 jan. 2019 kl. 23:49 skrev 'Rob Barter' via Amateur radio interferometry <amateur-radio-...@googlegroups.com>:

Thanks David,

It does sound like the challenge is to get accurate enough gps disciplined clocks with very low phase noise to support a 21cm capture.  The cost might be prohibitive.

Are rubidium oscillators good enough? - they can usually be found @ around 2-300 USD on ebay

I noticed you mentioned the XTRX before.  I found a price mentioned, around $2900, but can’t find anyone actually selling it?

Crowdsupply

[David Lonard]

See this link for some code that I wrote to successfully do VLBI.

https://groups.google.com/forum/#!topic/amateur-radio-interferometry/k5rmMLVmsPI

It requires aligning IQ data streams from two receivers using the first two programs, the third program is used to do cross correlations.

Rubidium oscillators are not good enough for any VLBI in my experience, even at HF. I bought two rubidium oscillators and tuned them on my oscilloscope and I couldn't keep the clocks aligned no matter how carefully I tuned them. GPSDOs are much better and will do the job at HF and VHF frequencies and can be found on Ebay at reasonable prices. At higher frequencies, VLBI poses a big challenge that I never attempted to solve and this is that the fringe rate gets really high. The pros use clock delays to do 'fringe stopping' so that they can achieve sufficient integration of signal over time. 

I would try to get non-VLBI interferometry up and running before moving to VLBI. Still, I think VLBI is something that could be possible at an amateur level at HF/VHF frequencies with a SDR like the base model XTRX. It has two coherent channels that could provide a single baseline. However, its real strength is that it has its own GPS reference and can produce timestamped IQ data and if coordinated reception is performed at multiple sites across the country/world, participants could upload their IQ data to a common repository so that it could be cross-correlated later. RFI is a huge problem at HF/VHF frequencies and I know that I can not get any fringes in a suburban/urban environment, but can get fringes in a rural area. Also, data will build up at a very fast rate and analysis will be a problem. Computers with SSDs to record Tb/hour level data and some strategies to lower the overall data volume are needed, say by doing FFTs of 16 bit data, then reducing the bit depth down to 1 or 2 bits before uploading to a repository for common access.

[Leif Johansson]

Slightly off topic... 

I recently learned about this: https://www.ohwr.org/projects/white-rabbit - open hardware (relatively inexpensive). Folks I know in the time and frequency "biz" say it can do nanosecond-scale sync with very little phase noise over standard fibre gig ethernet. 

[David Lonard]

The LOFAR group uses White Rabbit.

https://indico.narit.or.th/event/80/attachments/70/174/2-Jan_David_Mol_LOFAR_A_decade_of_correlating_in_a_software_telescope.pdf

[Rob Barter]

Am I right in thinking this works on a dedicated fibre link (up to 10km)?  They mention sharing the clock to 24 stations which I think it was referring to the local stations?  Good for the case where your further apart than a coax cable can bring the RF signal back to a common point but closer than the 10km.  

Wonder if it would still work with laser communications thru air rather than a fibre optic cable, assuming laying a cable isn't possible.  I'm guessing the total latency is the limiting factor for White Rabbit.  Too much latency or variability and it's a problem.

Rob Barter

[Leif Johansson]

On 2019-02-04 23:35, rob.b...@bedrocksystems.co.uk wrote:
> Am I right in thinking this works on a dedicated fibre link (up to
> 10km)?  They mention sharing the clock to 24 stations which I think it
> was referring to the local stations?  Good for the case where your
> further apart than a coax cable can bring the RF signal back to a common
> point but closer than the 10km.  
> Wonder if it would still work with laser communications thru air rather
> than a fibre optic cable, assuming laying a cable isn't possible.  I'm
> guessing the total latency is the limiting factor for White Rabbit.  Too
> much latency or variability and it's a problem.

I think you're relying on the optical framing of gig-E so doing this
over laser might be a bit of a project.

Cheers Leif

[Leif Johansson]

On 2019-02-04 22:27, David Lonard wrote:
> The LOFAR group uses White Rabbit.
>
> https://indico.narit.or.th/event/80/attachments/70/174/2-Jan_David_Mol_LOFAR_A_decade_of_correlating_in_a_software_telescope.pdf
>
>

Yeah and CERN etc.

What struck me was that the cost of the nodes is pretty low - a single
point-to-point link might not set you back more than a couple of high-
end SDRs

> <mailto:amateur-radio-...@googlegroups.com>>:

>
>> Thanks David,
>>
>> It does sound like the challenge is to get accurate enough
>> gps disciplined clocks with very low phase noise to
>> support a 21cm capture.  The cost might be prohibitive.
>
> Are rubidium oscillators good enough? - they can usually be
> found @ around 2-300 USD on ebay
>
>>
>> I noticed you mentioned the XTRX before.  I found a price
>> mentioned, around $2900, but can’t find anyone actually
>> selling it?
>>
>
> Crowdsupply
>
>> Rob
>>
>> On 15 Jan 2019, at 18:38, David Lonard <dmlo...@gmail.com

>>> <mailto:amateur-radio-...@googlegroups.com>>

>>> wrote:
>>>
>>> David,
>>>
>>> I see in your project notes

>>> <https://redirect.viglink.com/?format=go&jsonp=vglnk_154756192695713&key=e6d017f13f6a6e8cc619287ca1b92ca3&libId=jqxu91sf01013trd000DA2kckzh7q&loc=https%3A%2F%2Fstargazerslounge.com%2Ftopic%2F328609-radio-intensity-interferometry-at-low-ish-cost%2F&v=1&out=https%3A%2F%2Fsites.google.com%2Fsite%2Famateurradiointerferometry%2Fproject-updates&ref=https%3A%2F%2Fstargazerslounge.com%2Fforum%2F75-radio-astronomy-and-spectroscopy%2F&title=Radio%20Intensity%20Interferometry%20at%20low-ish%20cost%20-%20Radio%20Astronomy%20and%20Spectroscopy%20-%20Stargazers%20Lounge&txt=VLBI%20Oscillators>

>>> you mentioned VLBI and GPS disciplined oscillators. 
>>> Have you taken this any further at all for 1420MHz
>>> interferometry?  Very interested if it's viable for
>>> an amateur setup.  Computing power nowadays is not an
>>> issue.
>>>
>>> Cheers
>>> Rob
>>>
>>> --
>>> You received this message because you are subscribed
>>> to the Google Groups "Amateur radio interferometry"
>>> group.
>>> To unsubscribe from this group and stop receiving
>>> emails from it, send an email to
>>> amateur-radio-interf...@googlegroups.com

>>> <mailto:amateur-radio-interf...@googlegroups.com>.

>>> For more options, visit
>>> https://groups.google.com/d/optout.
>>>
>>> --
>>> You received this message because you are subscribed to
>>> the Google Groups "Amateur radio interferometry" group.
>>> To unsubscribe from this group and stop receiving emails
>>> from it, send an email to
>>> amateur-radio-interf...@googlegroups.com

>>> <mailto:amateur-radio-interf...@googlegroups.com>.

>>> For more options, visit https://groups.google.com/d/optout.
>>
>> --
>> You received this message because you are subscribed to
>> the Google Groups "Amateur radio interferometry" group.
>> To unsubscribe from this group and stop receiving emails
>> from it, send an email to
>> amateur-radio-interf...@googlegroups.com

>> <mailto:amateur-radio-interf...@googlegroups.com>.

>> For more options, visit https://groups.google.com/d/optout.
>
> --
> You received this message because you are subscribed to the
> Google Groups "Amateur radio interferometry" group.
> To unsubscribe from this group and stop receiving emails
> from it, send an email to
> amateur-radio-interf...@googlegroups.com

> <mailto:amateur-radio-interf...@googlegroups.com>.

> For more options, visit https://groups.google.com/d/optout.
>
> --
> You received this message because you are subscribed to the
> Google Groups "Amateur radio interferometry" group.
> To unsubscribe from this group and stop receiving emails from
> it, send an email to
> amateur-radio-interf...@googlegroups.com

> <mailto:amateur-radio-interf...@googlegroups.com>.

> For more options, visit https://groups.google.com/d/optout.
>
> --
> You received this message because you are subscribed to the Google
> Groups "Amateur radio interferometry" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to
> amateur-radio-interf...@googlegroups.com

> <mailto:amateur-radio-interf...@googlegroups.com>.

> For more options, visit https://groups.google.com/d/optout.
>
> --
> You received this message because you are subscribed to the Google
> Groups "Amateur radio interferometry" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to amateur-radio-interf...@googlegroups.com

> <mailto:amateur-radio-interf...@googlegroups.com>.