Hi all,
I (finally) got the differential hydrogen-line interferometer I've been working on running in the field. So far, technical difficulties have prevented me from getting any reliable data as of yet, but even so I'm not particularly confident about the data I have gotten; it looks to me like there's some kind of pattern, but I'm worried I'm just trying to find patterns in random noise.
I'll leave an update on this post when I have actual data to report, but regardless I thought it would be worthwhile to describe my setup and see if anyone sees anything wrong with it:
I'm using the differential interferometer (don't know if that's the right term for it) design on the CCERA website, so I have cable runs from my two dishes that plug into two LNAs, physically nearby one another to keep their temperatures the same. The output of the LNAs are split and combined in two different ways; one SDR receives the two signals with the same phase delay, and the other receives one signal 180 degrees out of phase from the other.
- The cables connecting the dishes to the combiner are both 50-foot RG11 coax with F-type connectors, but I have to use adapters to convert them to SMA on both ends.
- The LNAs are both Nooelec SAWbird H1+ Barebones, and I'm powering them using their microUSB port. They're both in a plastic bag and are rubber-banded together to keep their temperatures the same.
- The SDRs are both Nooelec NESDR SMArt XTRs; one of them receives the output of both LNAs through ~15cm SMA cables, since using only connectors put everything under too much stress, and the other is connected to one LNA by a male-to-male SMA connector and the other by a SMA cable cut to half a wavelength (accounting for the velocity factor of the cable) plus the length of the connector inside the hexagonal fasteners.
- I'm using a Raspberry Pi to run the GNU Radio code to read both SDRs and output the data to a file. The flow-graph does the following in order:
- Reads from both SDRs at a sample rate of 2MHz with the correct target frequency, with gain set to manual and IQ offset set to automatic
- Converts the outputs from complex to magnitude squared
- Subtracts the two squared magnitudes
- Moving average over 500k samples (0.25 seconds of integration time)
- Keeps 1 in 500k samples (this makes the moving average mostly pointless; maybe this should be a different number?)
- Outputs the data as floats to a raw binary file using an unbuffered file sink
I know this is a lot of information, but I was hoping that a few people who know what they're talking about could confirm whether or not this setup has flaws in it. It wouldn't surprise me in the least if there's something I'm doing that makes the entire thing non-functional (honestly, it would kind of explain a lot), so if anyone has time to read through this and give some feedback I would really appreciate it.
Thanks,Aidan
--
--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/56a432f7-ecfb-4e9c-b35e-07de77f9ada3n%40googlegroups.com.
I totally understand that having the amplifiers before any cables is much better, and while it would be kind of hard to change my setup to work like that I definitely would do that if I was aware it was an option. My understanding was that this type of interferometer only works if the LNAs are at the same exact ambient temperature. My dishes (which are 1.5m homemade mesh prime-focus dishes) are around 80-ish feet and while they are probably pretty close temperature-wise I was under the impression that the small differences in ambient temperature would drown out any real signal. Is it just a tradeoff, where the signal is either noisy or has more severe gain fluctuations, or is it not as much of a problem as I think?
Also, after looking through some of the incomplete data from last afternoon (GNU Radio programs seem to crash sometimes on a Pi because of memory access issues: I'm currently running the Python program from a Bash script that restarts the script when it crashes), I definitely agree that the integration time is too short. I tried averaging the data in Python after the fact, and the data looks a lot better. I'll definitely try increasing the integration time for tomorrow's observation.
As for writing my own Python block, that sounds like a good idea. I've been editing the top_block.py file instead of modifying the .grc file since the editor is slow and I'm more familiar with Python anyway, so I think that would make this way easier. I'll have to see how much I need to downsample for the Pi to keep up with it, though!
Thanks as always,Aidan
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/55b3a848-14e9-477b-9b6b-060399029730n%40googlegroups.com.
For an indication of performance vs noise, I suggest reading this
( and see attached chart )
For RG-11, 50 ft length, loss at 1.4GHz = 2.7 to 3.3 dB ( depending on reference ) + adapters : that has a significant impact on your system performance .
Alex KK4VB
- The cables connecting the dishes to the combiner are both 50-foot RG11 coax with F-type connectors, but I have to use adapters to convert them to SMA on both ends.
- The LNAs are both Nooelec SAWbird H1+ Barebones, and I'm powering them using their microUSB port. They're both in a plastic bag and are rubber-banded together to keep their temperatures the same.
- The SDRs are both Nooelec NESDR SMArt XTRs; one of them receives the output of both LNAs through ~15cm SMA cables, since using only connectors put everything under too much stress, and the other is connected to one LNA by a male-to-male SMA connector and the other by a SMA cable cut to half a wavelength (accounting for the velocity factor of the cable) plus the length of the connector inside the hexagonal fasteners.
--
--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/245514b1-4899-478e-a8b9-35acca6a7290n%40googlegroups.com.
Hi Aidan,
Look at this info on the Friis Equation
Noise figure (NF) and noise factor (F) are measures of degradation of the signal-to-noise ratio (SNR), caused by components in a signal chain.It is a number by which the performance of an amplifier or a radio receiver can be specified, with lower values indicating better performance.Linear passive devices have noise figure equal to their loss.
Expressed in dB, the NF is equal to -S21(dB).
Something with one dB loss has one dB noise figure
So, if the 50 ft of RG11 has 3 dB atten and each of the 4 connectors ( F and F<>SMA adapters ) have 0.25 dB loss, you have 4 dB of loss =4dB of Noise 'temperature' added to the system due to having the Cable before the LNA.
What can I actually do to reduce the system temperature? Move the LNA to the antenna feed_point before the 50ft of cable
Self powering the SAWBird also puts 5V on the Output . DC Blocks will not hurt ... depends on the SDR if that is an issue or not.
Alex KK4VB
On Thursday, May 12, 2022, 08:27:29 AM EDT, Apeirohedron <apeirohe...@gmail.com> wrote:
Yeah, 0.18K doesn't sound like much. Possibly a dumb question, but what can I actually do to reduce the system temperature? Is it just 'have better components'? Is that calculation for my current setup with the amplifiers before the cable run, or will that not make a difference?
Also, related question: I'm trying to put the amplifiers before the cable runs, but I've found that they're producing a lot of DC voltage which I hadn't considered previously (I've checked and the SDRs and LNAs are all undamaged as far as I can tell). From other posts, it seems like I'm supposed to have a DC block after each LNA. Is that the correct solution? I don't want to kill any of my equipment.
Thanks,Aidan
--
--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/459834154.1391485.1652359476814%40mail.yahoo.com.
And, although the equation shows that noise in the later stages is reduced by the gain of the device and preceding stages,it is easier to 'visualize' that the initial noise sources are Amplified by the Later Stages.
" The First Noise is the Worst Noise "
AP
--
--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/413491035.1391906.1652361201485%40mail.yahoo.com.
I'm unaware of any SDRs that DO NOT have a DC-block on the input
--
--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/27a31a3cf9fce0553b2ed19bf30946837181d718.camel%40googlemail.com.
Yeah, 0.18K doesn't sound like much. Possibly a dumb question, but what can I actually do to reduce the system temperature? Is it just 'have better components'? Is that calculation for my current setup with the amplifiers before the cable run, or will that not make a difference?
Also, related question: I'm trying to put the amplifiers before the cable runs, but I've found that they're producing a lot of DC voltage which I hadn't considered previously (I've checked and the SDRs and LNAs are all undamaged as far as I can tell). From other posts, it seems like I'm supposed to have a DC block after each LNA. Is that the correct solution? I don't want to kill any of my equipment.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/ad5f5322-55f5-4a24-8cc7-c3494a80f817n%40googlegroups.com.
Quick question then: by 'good hygiene', do you mean that I would be taking a risk by not adding DC blocks or that it's general good practice but not necessarily an issue in this situation? I definitely do want to do things correctly as much as possible, but I'm under time pressure and don't want to lose observation time waiting for ordered parts to arrive if not using them is safe.
- Aidan
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/b0c0141f-aefc-4ae2-b571-1b365e2f3d1en%40googlegroups.com.