Category: Discovery Dish

Discovery Dish April Manufacturing Update

This was posted over on our Discovery Dish Crowd Supply updates page and we are cross-posting here too.

It’s been over a month since our last update so we thought we’d share some of the recent progress. As mentioned in our last update, during the month of February all manufacturing was shut down due to the Chinese New Year holiday. In early March, staff returned to the factories and began making progress on finishing Discovery Dish.

We have a snazzy new logo, and we are working on obtaining some stickers to include with the feeds.

 

Dish Manufacturing Progress

The molds for the dish stamping machine were successfully created in March and the test stamps have come out great. The manufacturer is still tweaking the secondary mold that stamps the mounting and connecting holes, but we expect that to be completed shortly. Once that is done and tested, we can begin stamping the dish segments en masse.

The Dish mounting system was also finalized and we added 45-degree markers to it, which can help with aligning skew.

We are still awaiting the results of the anodizing tests, but they should be completed by the end of this month. Anodizing the dish is important as the dish must be a dark non-reflective matte color, so that it does not focus hot sunlight onto the feed point plastic head.

Feed Manufacturing Progress

PCB Upgrades

While waiting for the manufacturers to finish up with the molds, we’ve been further refining the PCB feed. Our final version of the PCB has now moved to a PTFE substrate with significantly lower loss at higher frequencies. This has yielded an over 1 dB increase in SNR at the GOES 1.69 MHz frequency.

The change to PTFE was not without problems. An interesting RF engineering problem occurred with the move to PTFE that we wanted to share. When moving to PTFE the only changes to the board layout are PCB trace width changes to keep the impedances matched. Other than that, the boards and layout are essentially identical. However, we discovered that the dual LNA design started oscillating when we moved the PTFE substrate. Oscillations can occur with LNAs when RF essentially bounces back and forth between the two LNAs, which causes undefined behavior in the LNA, such as poor gain, multiple spikes in the spectrum, and unexpected current draw values.

We found this quite odd because oscillations were not occurring in the original FR4 PCB, and the QPL9547 LNA is advertised as ‘unconditionally stable’ which means that it should never oscillate. However, we found that unconditionally stable guarantees may not apply to two-stage designs. In the end, the fix was simple, we just needed to add a damping resistor to one of the inductors on the circuit which reduces its Q-value. It seems that the change from FR4 to PTFE effectively increased the Q-value of this inductor so much, which in turn induced an oscillation in the circuit.

Discovery Dish Feed Head Enclosure

We’ve also refined the entire feed assembly. The feed arm pipe now has a ruler laser etched onto it so that mounting it at the correct distance is easy. A skew angle guide has also been added around the neck. A thumbscrew locking mechanism has been added to the feed head neck too, so that skew can easily be adjusted without the need for a screw driver or Allen wrench to loosen the set screw.

The PCB enclosure has been slightly refined and the injection molding die is currently in production and due to be completed in mid-May. While waiting for the die to be made, we’ve been testing different plastic mixes for the head enclosure to make sure that they are UV stable. The plastic mix has certain strict requirements and choosing the correct mix is crucial. It has to be RF-transparent with a low relative permittivity value, it has to endure direct sun, UV damage, and freezing weather, as well as be water-proof too.

S-Band Feed

Previously as noted in prior updates we were testing an S-band feed with the FR4 substrate. But we found that there was too much loss and the SNR values we got were not great. The move to PTFE substrate means that our experimental S-band feed is now working very well. We will be releasing this in the near future as an additional feed product that can be used with Discovery Dish. This feed will have a frequency range of 2.2 GHz - 2.3 GHz. This covers the main S-band weather satellites, other satellites like Coriolios and JASON as well as the many dump-only S-band satellites that transmit signals only over certain regions.

As requested by most people interested in an S-band feed, the S-band feed will not include a downconverter, so to use it you will either need an SDR like the HackRF which is capable of tuning to the S-band, or a third-party downconverter product.

Discovery Dish Outdoor Metal Enclosure Progress

Our enclosure set is now complete, and the final packing has almost been completed. The user manual can be found here Discovery Dish Outdoor Enclosure User Manual.pdf.

The final set consists of:

  • 1x Metal Enclosure
  • 3x Custom metal cable glands
  • 1x Vent
  • 1x Electronics mounting board
  • 1x Pole mounting set (with hose clamps)
  • 1x Wall/DIN mounting set
  • 1x 10 mm x 10 mm x 8 mm thermal pad (to be placed under the electronics mounting board)
  • 1x 10 mm x 10 mm x 3 mm thermal pad (to be placed under electronics on top of the mounting board)
  • 1x Set of various screws and washers

(Note that there will be some minor changes from this image in sets going out of customers - the hose clamp will be shorter, and the mounting rails will be longer)

Meteor M2-4 Launch

We mentioned in our last update that a new weather satellite Meteor M2-4 was due to launch. The launch was successful and the satellite is now in orbit. The satellite was briefly turned on after launch, and we were able to receive HRPT images from it in the L-band. However, now it is currently in a testing phase so the transmitters are often turned off. We don’t know how much longer it will be in testing, but we assume it won’t be more than a few more months.

GOES-U / GOES-19 Launch Updates

We’ve been keeping an eye on the expected launch date for the next GOES satellite. Currently, it has been delayed from April 30, 2024, to the new date of June 25, 2024, when it will be launched on a Falcon Heavy from Kennedy Space Center, Florida.

Elektro-L4 Updates

In the last update, we mentioned that we were having some problems getting SatDump to receive Elektro L4 properly on computing devices that used ARM processors. After some investigation, we determined that this was a problem with buffer size settings in SatDump and we were able to suggest a fix in https://github.com/SatDump/SatDump/pull/616 which was implemented. New versions of SatDump have this problem fixed.

Driver Tweaks

We have been looking at the RTL-SDR drivers and have found a few tweaks that can improve performance at L-band frequencies. We’ve put a modified version of the librtlsdr/librtlsdr fork up our the rtlsdrblog GitHub at https://github.com/rtlsdrblog/librtlsdr. With this fork and the PTFE feed upgrades, we now get around 5-6 dB of SNR on GOES-18.

Antenna Rotator

The low-cost antenna rotator is finishing up with prototype testing, and we are now working on improving the design’s manufacturability.

Timeline

The ramping of progress from the Chinese New Year holidays to now has been a little slower than expected, but if everything goes perfectly to plan, we will be on time for shipping by the end of June. However, this is currently a best-case scenario. There are still a few manufacturing stages to get through like the final mass production, CE testing and sea freight shipping. Unfortunately, from prior manufacturing experience, there are always setbacks along the way that slow progress, so we are conservatively pushing our advertised timeline back by about 1-2 months. We apologize for any potential delays, but we are working hard to get the product out to you ASAP!

Customer Questions

We have had a few more customer questions over email which we’d like the answer publicly below:

Would it be practical to use this kit indoors to do hydrogen-line astronomy? I ask because I live in a flat in a block of retirement flats, and wouldn’t be able to place an aerial or dish outside, but could find space for a Discovery dish + rotation gear inside.

Unfortunately, indoor Hydrogen line astronomy is out of the question. The hydrogen line signal is just too weak to be seen indoors, and there would most likely be too much interference indoors as well.

What all do i need to purchase and get for the Discovery Dish for Radio Astronomy?

You’ll need these two components from the Discovery Dish Crowd Supply store:

  • Discovery Dish (Dish & Mount Only)
  • Hydrogen Line Discovery Feed

You will also need a software defined radio, such as an RTL-SDR Blog V3/V4 which can be purchased separately, a computer and somewhere to mount your dish (e.g. a mast/tripod etc).

Have you tested this for C-Band / Aero downlink reception?

Currently, we do not have a C-band feed, so we have not tested it for this purpose. Generally, a larger 1.2 m+ dish is required for C-band AERO, so the 70 cm Discovery Dish may not be suitable. We may test this with Discovery Dish in the near future just in case however.

I want to purchase a hydrogen antenna. Where should I buy it? Also, please tell me what is needed to observe seti radio waves.

The Discovery Dish (Dish & Mount) and Hydrogen Line Discovery Feed can be pre-ordered from the Discovery Dish Crowd Supply store. I’m not sure exactly what you are referring to with SETI radio waves. An alien signal could in theory be on any frequency, but the 1.42 GHz Hydrogen line frequency could be a good bet as it’s a universal frequency of interest that any technological civilization would be observing. Realistically the Discovery Dish would be too small to detect potential alien signals unless they were very strong.

I’ve seen a small home made 3D-printed dish called a ‘heliocone dish’ being used for HRPT. How is Discovery Dish different?

The heliocone 3D printer design going around is a great DIY solution for L-Band polar orbiting HRPT satellites. But it has some limitations as it cannot receive the weaker geostationary satellites due to its smaller size and use of a circular polarized feed. It also cannot receive satellites using the opposite circular polarization. It also cannot receive satellites on different frequencies or the Hydrogen Line without designing a new helical feed and using a different LNA+filter combo. Discovery Dish is a more of a general purpose ‘does-it-all’ and ‘ready to use’ out-of-the-box dish. With our dish and feeds you can receive the L-Band polar orbiting HRPT satellites as well as the geostationary satellites. You can quickly swap out the feed for a different feed that covers a different band as well.

 

SatDump V1.1.4 Released

SatDump is a popular program used to receive and decode images and other data from various weather satellites. SatDump works great RTL-SDR Blog dongles and with our Discovery Dish, an easy to use dish and feed for receiving L-band and other weather satellites. Recently the author of SatDump released version 1.1.4 which brings several new features including:

  • Autotrack/Scheduler improvements - multi mode has been added which allows all pipelines to run, even if more than one satellite is overhead at the same time. This is useful for NOAA and METEOR satellites at 137 MHz as there can often be more than one active satellite broadcasting images at different frequencies during a pass.
  • TUBIN Raw/Video Mode
  • ESA Cluster support
  • Additional Pipelines - Including Peregine X-Band TLM, the IM-1 Moon Mission and PRETTY S-band dump.
  • Themes - Choose between Dark, Light, Phosphor and Win98 themes.
  • Android Improvements - Blog V4 support added on Android, OpenCL support added.
  • Added composites - Various composite image products added to various satellites. Including Fog, panchromatic, ice detecting and more.
  • Other Features - support for M1 Mac builds added and various other fixes.
  • Bug Fixes - Memory leaks fixed and various bug fixes including a bug that caused problems with RTL-SDR devices on low power hardware like Raspberry Pi's and Orange Pi's.

If you enjoy SatDump consider donating to the main author at ko-fi.com/aang23.

SatDump Multi-Mode Feature. Receiving data from multiple 137 MHz satellites at the same time.

Multiple Comprehensive Tutorials on Weather Satellite Decoding

Over on his website "Jacopo's Lair" IU1QPR (@original_lego11) who is also a developer for SatDump has written up many tutorials about weather satellite decoding that involve the use of SatDump. SatDump is a popular piece of software often used with RTL-SDRs and other low cost SDRs for decoding weather satellite images.

With a small satellite dish, feed, RTL-SDR and LNA+filter and the SatDump software it's possible and download beautiful images of the earth from many geostationary and polar orbiting weather satellites. We note that we are currently taking pre-orders on Crowd Supply for our Discovery Dish system, which is low cost hardware designed to help users get started with weather satellite reception.

Over on Reddit IU1QPR has created a listed summary of all the tutorials he's written. These are currently the most up to date and comprehensive tutorials that we have found on this topic. The tutorials cover everything from what satellites are available, what dish sizes you need, what SDRs can be used, what LNA+filter and other hardware you need, and how to use the SatDump software.

Satellite reception and decoding

Automated stations

SatDump usage

All have been moved to SatDump's documentation page

Satellite data processing and usage

From the HRPT tutorial: What various HRPT signals look like on the spectrum.
From IU1QPR's HRPT tutorial: What various HRPT signals look like on the spectrum.

Discovery Dish Updates and Some Cool Hi-Res Images

We have recently posted an update on our Discovery Dish crowd funding campaign over on Crowd Supply. Check it out on the update page, or on the repost down below.

Discovery Dish Teardown Session Livestream Recording

Thank you to Helen Leigh and Crowd Supply for featuring us on one of their Teardown Sessions live streams a few days ago. If you missed it, feel free to watch the recording below. On the livestream we discussed the Discovery Dish and talked a bit about the journey we took to get to the final product design.

Teardown Session 38: Discovery Dish

Enclosure Glands and Vents

We have decided to include a few cable glands and vents with the Discovery Dish Enclosure, as these will ensure that the bottom of the enclosure is protected against water jets and any splash back from the ground, as well as allowing the electronics inside to breathe a bit. Allowing waterproof enclosures to breathe is important in many environments to avoid condensation build up inside.

The glands and vents will be metal to ensure that RF tightness of the enclosure is maintained as much as possible.

The electronics inside can be passively cooled via thermal pads that sink all generated heat to the metal enclosure which acts as a large thermal mass and heatsink.

In the image below you can also see the mounting board. We are still planning to reduce the hole spacings on the board.

Rotator Timelapse

We’ve been testing an early prototype design of our upcoming antenna rotator for the Discovery Dish, and have created a quick preview timelapse of it running overnight. With mechanical designs like this it’s important to do some long-term testing, so we’re going to be running prototypes non-stop for several months while tracking many more satellites than would be typical.

DD Rotator Preview

 

Example Weather Satellite Images Downloaded

Some people have asked for high resolution examples of what can be received from satellites with the Discovery Dish. Below are a few samples.

GOES 18 Full Disk

Discovery Dish GOES 18 Full Disk Blend

 

GOES 18 Mesoscale

GOES 18 EMWINN

GOES 18 NWS

Metop AVHRR (Advanced Very High Resolution Radiometer)

Metop IASI (Infrared Atmospheric Sounding Interferometer)

GK-2A Full Disk

FengYun 4A Full Disk

Meteor MSU-MR (Multispectral Scanner Unit - Medium Resolution)

NOAA AVHRR (Advanced Very High Resolution Radiometer)

Crowdfunding Goals

We just wanted to clarify a point regarding how crowdfunding works. If the goal isn’t reached then everyone who ordered won’t be charged. We have had a few concerns from potential customers wondering if we will keep the money if the goal isn’t reached, but this is certainly not the case! In fact, credit cards will only be charged if we hit our funding goal. You can learn more in the Crowd Supply Guide.

The goal is set relatively high as this product requires a number of molds to be created for the dish and the various plastic parts, and molds typically have a high fixed initial cost. There is also a high minimum order quantity that we need to commit to in order to do a production run.

But the campaign is currently over 70% to its funding goal and we are expecting some large reseller orders to come in during the last few days of the campaign, so please don’t worry as the goal will almost certainly be reached with the help of just a few more individual supporters. If you have been on the edge, please consider supporting us to get this product started!

Customer Questions

In my environment temperatures get down to -20 to -30 degrees C. Will the electronics in the feed hold up?

The components used in the feed all have ratings down to at least -40 degrees C. In very cold environments, the one thing we would suggest considering is if a dish heater is required. These are heating strips that can be placed on the dish and can help melt snow/ice buildup.

What is the hole pattern on the dish?

The hole pattern on the dish has no specific function, the holes are simply used for reducing wind loading and weight. The manufacture of the prototype dish requires that the holes be cut by laser cutter, but the laser cutter we have available was not large enough to do the entire dish at once. So it was manually rotated around, and this caused an uneven pattern.

The production version of the dish will split into three petals, and each petal will be manufactured via a stamping process. Stamping is when a sheet of metal is placed under a heavy molded block of metal, and then that block of metal is pressed down on the sheet metal to create a desired shape. With this stamping process we will have perfectly neat hole patterns.

I suggest that the S-band version of the feed not use a downconverter, and just use an SDR that can receive S-band instead.

We currently have a similar opinion.

To explain this customer question/comment, we note that as mentioned in the previous update, we are planning to soon test an S-band version of the feed which should be able to receive S-band satellites.

However, the typical software defined radio used is an RTL-SDR, which cannot reach S-band frequencies like 2.2 GHz where most S-band satellites transmit. To get around this, we could add downconversion circuitry to the S-band feed, which would increase complexity and cost. This would convert the 2.2 GHz frequencies down to a frequency that the RTL-SDR can receive (below 1.766 GHz). Alternatively, we could simply recommend that customers interested in S-band reception instead use another SDR such as the HackRF, PlutoSDR, or LimeSDR Mini 2.0.

Once we have tested the S-band version of the feed, we will make a decision on if we should add a downconverter or just recommend the use of other SDRs that can reach the S-band.

Can any of the feeds be used for 1296 MHz EME (earth-moon-earth bounce communications)?

Sorry no, the feeds will not be suitable for EME, as that requires transmission which our feeds do not support.

I would like to use the dish on an astronomical mount. What is the expected weight of the dish and feed?

The dish itself weighs less than 1 kg (2.2 lbs). Together with the feed and mount we expect it to weigh a total of less than 1.5 kg. This is significantly lighter than a Wi-Fi dish which is already 1.6 - 2 kg (depending on the brand) for just the dish by itself.

Are weather satellites encrypted?

No, most weather satellites like this are not encrypted. Although these satellites come from various countries’ governmental space and/or military agencies, weather satellite data is generally considered public science. If it’s not necessary, adding encryption is undesired as it adds complexity to the system and increases the amount of data that needs to be transferred.

Obviously high-end military and commercial satellites are encrypted and we cannot receive data from those. It’s possible that future weather satellites could be encrypted, but given the current trend of new weather satellites being unencrypted this seems unlikely.

 

The Latest Progress on Discovery Dish

Over on Crowd Supply we are currently crowd funding for the Discovery Dish, a system that aims to help make satellite dish based radio projects more accessible for use with low cost software defined radios like the RTL-SDR. We've recently posted an update which we pasted below.

Discovery Dish: Simplified system for weather satellite reception and hydrogen line radio astronomy

The Latest Progress on Discovery Dish

First, we want to thank everyone who has purchased a Discovery Dish! We are about two weeks into the campaign now and we’ve reached over 30% of our goal. Please help us get there by sharing the campaign with anyone you think might be interested!

Progress Report

Here are the latest updates:

  • We’ve been working on getting manufacturing of the molds and electronics ready to go once we receive funding. We’re finalizing our CAD files and double checking everything so we’ll be ready to go once the campaign ends. We put six months as our target before shipping, but we’re hoping to actually get the product out sooner than that. The main delays in the timeline will be the Chinese New Year holidays early next year and the time it will take to sea freight our bulk production runs.
     
  • For the enclosure, we’ve begun getting samples of the general electronics mounting board. It will be made out of a conductive metal which is important for grounding RF noisy electronics to the enclosure, and will also allow heat to transfer out of the enclosure via a thermal pad underneath the board. Once we get our prototypes we will share more images.
 
  • We've also began considering how we might implement a 2.2 GHz S-Band feed for the Discovery Dish. The return-loss characteristics of the feed were designed to be good at 2.2 GHz, so we probably don't need to change much of the core feed design. The main question will be if it's feasible to implement a downconverter for use with RTL-SDRs (which have a maximum frequency limit of 1.766 GHz), or if it's better to just use a HackRF for this band. Updates on our investigations will be provided as we test further.
  • We are also actively working on our rotator prototype which we hope to release next year as a companion product to the Discovery Dish in order to make reception of polar orbiting satellites easier and more accessible. We don't want to release too much information on the rotator at the moment as things could still change a lot, but currently we are ordering samples of some custom parts that we need to test a production version. We are also developing the microcontroller firmware so that it will be compatible with the EasyComm II rotctl protocol.

Customer Questions

Finally, we’ve received a few questions from customers which we’ll answer publicly below:

Inmarsat is circularly polarized. Is the Inmarsat feed circularly polarized?

Our feeds are all linearly polarized. But this actually does not matter much for Inmarsat because our dish is more than large enough for Inmarsat, and Inmarsat signals are relatively strong. Using a linear feed on a circularly polarized signal results in a 3dB loss which is relatively insignificant in this case. With a small patch antenna such a difference is significant, but not so with a larger dish.

Will this work with a SatNOGs rotator?

Yes, the Discovery Dish comes with a standard pole mount which can be used to mount it on the SatNOGS rotator arms.

What is the amplifier/filter architecture like in the feed?

The signal chain is as follows: Feed -> QPL9547 LNA -> SAW -> QPL9547 -> SAW -> SMA Output. So our feeds are dual-amplified and dual filtered.

Can the coax on the feed be swapped out for longer and lower loss coax?

Yes, the feed uses an SMA connector so you can swap out the coax cable if you like. Thicker cables may require different sized strain-relief at the end of the feed arm, though.