Future SDR Project Survey

[Steve Haynal]

Hi Group,

If you haven't already, please take the future SDR project survey. It covers these project ideas. Only 29 people have responded so far. That is a 2.9% response rate for a group with 1000 members. There are some interesting patterns emerging, and I will share results this weekend.

I have posted the survey to gauge interest in these project ideas:

https://forms.gle/HsxKB4aWdpTuRFxP7

Please read over the project ideas:

https://github.com/softerhardware/Hermes-Lite2/wiki/Future-Projects

73,

Steve

kf7o

[N7IHQ]

I am really excited about the Hermes-Nano project!

Jim

N7IHQ

[Steve Haynal]

Hi Group,

I have closed the survey and here are the results:

https://docs.google.com/forms/d/1j-JWiFoBHh8PEnfcuqmwTnAKb5IunV6kwLtyRgDcB2s/viewanalytics

I included comments. Four or five people identified themselves but their comments didn't seem of a private nature to me so I just published all comments. Thanks for the comments in this survey and in the group threads. I've read and considered every comment. If and when I start serious work on any of these project, I will post more details, questions and discussion.

The clearest result in my opinion is that people would like to see a Hermes-Lite 2.1 with updated FPGA. The larger and similarly priced Efinix FPGA is also interesting to me as it opens up many possibilities like more advanced DSP, more skimming receivers, better buffering for wifi, a small RISC-V soft processor for control, etc. One common requirement for all the proposed projects is RTL clean-up and porting to another FPGA. I'll probably work on this along with the CW keyer first.

Another common theme beyond what I proposed was the desire for more bands (VHF/UHF), more power and better peripheral integration (transverters, PAs, ALC). To be honest, I am a die-hard QRP digital mode operator who wants to be able to use a radio remotely. The HL2 already meets all my desires but has grown larger than my initial expectations. It is very unlikely that I will have the motivation, interest or expertise to work on VHF/UHF, PAs or peripheral integration. What I am willing to do is the following:

• If I revisit the HL2 layout, make minor changes to better facilitate others working in these areas. Minor changes means roughly less than 5% PCB rework. I am not ripping up and redoing large portions of the layout. Possible small rework includes making i2c more accessible for mixers, running some specialized IOs in the new inner layers of the HL2.1 PCB, adding a silicon switch for RX instead of RX2 and RX3 in parallel, etc. I would also try to accommodate similar interfacing wishes in any new project.
• Make small updates to the gateware again to facilitate others. This includes improved i2c pass through, i2c broadcasting of command words, uart pass through, control of a few extra IO signals, etc.
• Pay for out-of-pocket prototyping expenses for HL2 companion card or enhancement projects. For example, a run of 10 boards to test a new design. We have enough money in the R&D and repair funds to cover this.
• Help prepare the project for production and sale by Makerfabs. I have established connections with Makerfabs, know what they like, and can help expedite and simplify this process. 

All of this assumes following of the contributor guidelines, using github (managing contributions from several people is a nightmare without a revision control system. Git is a good standard. Learn how to use it.), keeping everything open source, and not forcing unnecessary forks of the project. A good example of this is the N2ADR filter board (use this link if you want to build of that project as it links to production code). I also hope that efforts by Bill and/or others to build a small RP2040-based board to offload IO results in something similar.

If people are interested in a more major redesign of the HL2 or new SDR, maybe Art who has already created his own versions of the HL2 might be willing to take on design and layout of a complete new board. I think that with guidance from this group, we could come up with something that appeals to many people yet requires minimal gateware changes. Also, something that can be built from parts readily available to Art, but also acceptable for small batch manufacturing by Makerfabs.

73,

Steve

kf7o

[philip.j.s...@gmail.com]

I have both the original 125MHz version with the smaller FPGA and the 122.88 version with the larger FPGA. I'm currently using the 122.88 version as a FT8 skimmer to report to pskreporter (which I run). I tried doing the same thing with my HL-2 and running SparkSDR on a windows box. This latter allows me to skim different modes (wspr/ft4 and ft8) on more bands. However, it isn't as reliable as the single box solution of the RP. I find that the communications between the HL-2 and the windows box just suddenly stop -- toggling the power button in sparksdr brings it all back to life).

On the RP box, the processor runs linux albeit fairly slowly. I think it is a dual core processor. It has enough horsepower to do 12 (at least) FT8 decodes every 15 seconds. 

In terms of speed, 100MHz is not that slow -- I can recall when I bought my first computer that contained a pentium-133! Certainly the ability to have (for example) a real DHCP client, a simple web server would be huge -- it would probably allow the front-end to the radio to be implemented as a web app and just transfer the relevant i/q data over (say) websockets. Having a QRP TX capability would be useful for that use case. Further, having a USB (maybe OTG) interface would allow plugging in a wifi dongle and enable field operation with just a phone and the HL-2. The nice thing about having a real OS is that the drivers for the Wifi component are probably easy to come by. Would need to make sure that the USB interface is supported by the OS (the Zynq Linux comes with supported USB drivers for host/gadget/otg).

I only use a single ADC on the RP boards -- I only have a single HF antenna (a big longwire)

As far as the lack of an LNA is concerned, it doesn't seem to make much of a difference. For a while I ran both my HL-2 and RP connected via a splitter to the same antenna to compare decoding performance. I never really got any satisfactory answers as the decoding software was different and this (I think) turned out to dominate. This may be a testament to the robustness of FT8.

Does this help?

Philip

[Steve Haynal]

Hi Philip,

Yes, that is helpful information. Since this is now on the list, I've posted the rest of this thread below.

I also run with a long wire and currently "hermes-lite" has more spots than n1dq on pskreporter. ;) But I think that has more to do with the depth of decode done by SparkSDR versus the depth of the decode possible on the Zynq's ARM A9 cores. I ran some experiments last year with Pavel's FT8 decoder recompiled for a standard PC. I had to change settings in the code related to decode depth to get more spots. There are more details in this github issue.

Sorry to hear that your HL2 still drops ethernet connection. I remember you submitted some patches to try and fix this. Unfortunately I can't replicate the problem. My HL2 on my home network with DHCP and the one at a remote location with direct connection are both rock solid and up for weeks. I haven't heard of any recent reports similar to yours. Do you have suggestions for how I might replicate your problem? Do you want to send your HL2 to me for possible fix or exchange?

Regarding a soft processor, I am a big fan of the work by enjoy-digital with the vex risc-v processor. I would probably use that setup for cpu/ethernet/etc in a larger FPGA. This will run Linux, but you need about 32MB or more to do that reasonably. That requires the FPGA to have external DRAM which requires significant pins and PCB area. I might be able to squeeze in a hyperram which can be used for linux. Bare metal code is more likely. At my day job we use a soft RISC-V for DRAM controller training. It should be able to do DHCP, ICMP and other similar tasks like a MCU would.

73,

Steve

kf7o

Hi Philip,

Can we move this discussion to the list? I have followed the red pitaya for several years now and have high regard for Pavel's work to use it as an SDR. Some of the proposed projects do use boards with Zynq parts:

https://github.com/softerhardware/Hermes-Lite2/wiki/Future-Projects#fpga-boards

Since I haven't used the red pitaya, would you answer a few questions I have?

Are you using the SDRlabs 122.88MHz version with the larger 7020 part? I assume this is the case if you are skimming 12 bands.

Do you do any of the decoding on the processors in the Zynq part? I know Pavel has code to do this, but I've heard that those interested in the most decodes will decode on a separate machine where they can use larger FT8 and WSPR decode depths.

There are two ADCs. Are you using both, if so for what purposes? Diversity? Separate antennas for low bands and high bands?

From the schematics, I see there is no low noise amplifier on the ADC inputs. Even the openhpsdr radios with 16-bit 122.88 MHz sampling had up to 20dB gain on the inputs. Do you add external LNAs? Does the ADC have an internal LNA? Is it not much of an issue to operate without LNA? The schematics I am looking at are here:

https://redpitaya.readthedocs.io/en/latest/developerGuide/hardware/122-16/top.html

Since the SDRLab 122-16 is available at Mouser for $582 and is already good hardware with good SDR support, how do you think the HL2 could/should improve on this? I don't want to compete directly and think any project I do should differentiate from the red pitaya. We could lower the price by building a $100 daughter board for a $90 Bajie (7020) board. This would be similar to the SDRLab 122-16 but would be 76.8 MHz at 12-bits with LNA. We would have to use a two board setup for the lower price. We could offer more functionality and include a full QRP TX end for a full transceiver. 

A soft RISC-V CPU core is possible but would run even slower, in the 100MHz range. I doubt any meaningful decoding could be done. This could be used mainly to simplify and standardize interfacing and control. Do you have other applications in mind?

73,

Steve

kf7o

 

On Sun, Dec 5, 2021 at 6:39 PM Philip wrote:

Steve

I missed the survey cutoff (my bad), but my interest would be in using an FPGA with a CPU core (either soft or hard). I use the Red Pitaya SDRs and the benefits of having a real processor (that can run a real OS) are huge. They use Xilinx Zynq parts. In particular, I can run a 12 band receiver and an FT8/WSPR decoder on each band and feed the results into other systems (e.g. to pskreporter). 

Philip

[philip.j.s...@gmail.com]

One minor point -- my HL-2 does not need to be power-cycled to recover, I press and unpress the soft button in SparkSDR. I still haven't managed to capture what goes wrong as it happens infrequently and I don't have a good way to capture all the traffic between those two boxes. 

Philip

[Philip Gladstone]

On the topic of new HL2 -- it'd be nice to leave a location to add something like an ESP32 module. This could be used to provide wifi / bluetooth at very low cost. Provided that (at least) the primary UART and boot control pins are attached to the FPGA then the ESP32 firmware can be uploaded via the HL2. It might also make sense to connect up either an I2C connection or an SPI connection to convey (for example) the audio data (could also use an I2S port). This approach would not be suitable for transferring huge amounts of data (the various interfaces are limited to 10Mbit/s or 1Mbit/s), but this is perfectly adequate for an audio stream.

Philip.

--
You received this message because you are subscribed to a topic in the Google Groups "Hermes-Lite" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/hermes-lite/z1pJAi9N3-Q/unsubscribe.
To unsubscribe from this group and all its topics, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/69a2116e-3bf7-4f7a-b83e-8c52f2fe6033n%40googlegroups.com.

[Steve Haynal]

Hi Philip,

I agree easier interfacing to an ESP32 module would be nice. We've discussed it on the list in the past. I don't think I will be able to add a footprint for an ESP32, but I do plan to add FPC connectors as described on the wiki for as many extra IO as I can. This should allow for an inexpensive carrier card with ESP32, RPi zero 2, or other footprint to live in the slot above the HL2 board and connect to the FPGA.

73,

Steve

kf7o

[Art]

Hi Steve. 

I was a little late in answering, as there were simple matters unrelated to the radio.
Of course, I am ready to help in the development of a new version of the PCB for HL2.1. I am not sure that it will be fast, since I am engaged in other radio amateur projects. Plus, I didn't quite understand the new structure. What chips Makerfabs can offer for assembly. When there is at least a block diagram with the names of the chips, I could prepare the components and start drawing the circuit and routing the board.

понедельник, 13 декабря 2021 г. в 07:37:21 UTC+2, softerh...@gmail.com:

[Art]

If everything works out, I am also ready to undertake technical support, repair and maintenance of the new version of the transceiver in Eastern Europe and some CIS countries.

понедельник, 13 декабря 2021 г. в 18:48:49 UTC+2, Art:

[Steve Haynal]

Hi Art,

Okay. Sounds good. Thanks for your help. For the HL2.1, I am considering making small changes to the existing HL2 to port to an Efinix FPGA. I was thinking that you and others might want to work on a HL3. Please feel free to propose a block diagram and possible chips. I will run any chip selections by Makerfabs for small batch manufacturing feedback. I will also point out any design decisions which might require difficult gateware changes. I don't want to take on a major role in a HL3 but some are interested.

73,

Steve

kf7o

[Art]

Hello everyone.
I started researching PHY interfaces, and a couple of questions arose:
1. The B50612D chip does not have a 125 MHz clock output. We need this clock speed for the FPGA. What to do with this signal? This is a simple interest. I know it is not available in the build service. But I can buy these chips in China and test them in operation.

2. Do you need a pull-up resistor on the RX_clock signal? Steve (kf7o) pulled up that signal in the HL2. And in the Hermes transceiver and similar ones from the OpenHPSDR project, this signal will be pulled to the ground through 1K? Will work in both ways, but which one is best?

What are the opinions?

вторник, 14 декабря 2021 г. в 08:36:10 UTC+2, softerh...@gmail.com:

[Steve Haynal]

Hi John,

Thanks for the feedback. Yes, any HL2.1 will be compatible with the existing N2ADR filter board or at most a slightly modified one which will be available in the same way as the current one.

73,

Steve

kf7o

On Tue, Dec 14, 2021 at 12:07 AM John Bain <john...@gmail.com> wrote:

Hi Steve and group, I had an order in for the last build that didn't go ahead because of the crazy times we're living in with shortages of even the cheapest of parts. I got my refund almost immediately from makerfabs as soon as the production run was cancelled. I would have really preferred to be getting the hermes with the filter board and case to have been on its way to me instead of the refund. I've obviously missed my chance of getting one of the version 9 or build 9 boards (i'm not really sure of the naming convention used for the board but I think it was along the lines of V9 or Build 9)now because of the chaos in the world at this point in time. 

 I did the survey and it seemed to me like the best bet for me would be the version 2.1 hermes lite. I am really interested in the idea of the 2.1 with maybe even some slight upgrades/changes to the board to allow for the use of a different fpga Steve but I wondered if any of the changes would mean that the N2ADR filter board couldn't be used with the new 2.1 version or if I'd still be able to get an order in for the 2.1 with the filter board and enclosure all from makerfabs at the same time. I'd much prefer to just be able to place an order for the HL 2.1 + filter board + enclosure from makerfabs like I did the last time before the build was cancelled. I don't have any interest in vhf/uhf on the HL2.1 so an hf only version would work out fine for me, I know a lot of people are looking for the vhf/uhf option but me personally I don't need that capability. 

 I know its probably a long way off Steve but I wondered if anything of your design might be available from makerfabs in the new year possibly Steve or is it going to be a longer term thing because of the amount of work involved for you and other people involved in the design and building and testing any new design or changes to the design/board layout or whatever. I'm only an end user waiting for the HL2/2.1 to come back into production so that I'll hopefully get the chance to own one of these fantastic radios in the future. I have no technical knowledge of any kind to offer you Steve or the group in general. The only thing that I have to offer would be a massive amount of hope that everything will be great and go smoothly with any new design to the radio. 

 Steve I thank you for the amazing work that you do and the amount of your own free time that you give up to bring these amazing radios to the community, and the same thing for anyone else involved with the project I thank you all for your amazing work. 

 I believe that I will get one of these amazing radios in the future and hopefully it won't be in the distant future. Anyway sorry for cluttering up the list with my useless end user comments or whatever. 

 Many thanks again for the amazing work that you do Steve and everybody else involved with the project. 

 Very Best Wishes and Very Best Regards to Everyone on this list and the other few HL2 lists that appear from time to time. Merry Christmas and Happy New Year to everyone, yeah I know its a bit early for that but I might not be able to write/post nearer the time. Keep Healthy and stay safe to all. 

All the best, John Bain, Glasgow, Scotland.

You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/50ac41e2-daf6-4aee-ad43-06f4cf5475adn%40googlegroups.com.

--

John Bain

[Art]

Hi Steve.
I apologize for the technical issues in the topic, which may not really fit, but I don't see a clear separation. And I don't want to litter the Group with another not very significant issue. Perhaps a regular forum on a free SMS system would be more convenient for such moments. But judging by the activity in the Group, fewer people are interested in technical issues.
My opinion is to investigate technical issues as much as possible and gain more experience in using different nodes for SDR technologies. Most likely, technical innovations in the field of DDC technologies should be expected in the next decade.

среда, 22 декабря 2021 г. в 06:51:57 UTC+2, softerh...@gmail.com:

[Steve Haynal]

Hi Art,

Sorry for missing your earlier question. I think the group is what we make it to be, and I'd like it to be technical.

The various pullups configure the KSZ9031/21 when power is applied. The chip first looks at the value on these pins and sets a configuration. The B50612D does something similar where certain pins have dual purpose. See page 44 of the real datasheet:

https://github.com/wuxx/Colorlight-FPGA-Projects/blob/master/doc/B50612D-datasheet.pdf

Also in that datasheet and on page 44 it says that pin 42 can be a 125MHz clock to the MAC, or FPGA-MAC in our case. I think it is possible to not use this signal and instead use the RXC, but it is more compatible with the existing gateware if it is used.

73,

Steve

kf7o

[Art]

Hi Steve.

Thanks for the answer
I would like to test two chips in operation: RTL8211CL and B50612D. I don’t know which one to choose yet. I want to switch to a new PHY chip in my designs.

1. RTL8211CL has a low cost. It's available. Has good documentation. The RTL8211E version has a 140 mA lower consumption current (400 and 260 mA, respectively) on the 1.05 V bus than the RTL8211CL version. But I am afraid that there will be interference from the built-in DC-DC at 1.05 V. This is the only thing that upsets me.

2. BD50612D is now also available only in China. But there are fears that these are leftovers from the assembly of radio electronics. And it may be withdrawn from sale in the near future. And if you look at the offers on Aliexpress, you can see that sales are carried out by a couple of sellers with different store names.

3. Technical question:
- can I keep the pull-up resistors compatible for RTL8211CL / E and B50612D chips as done in the Hermes versions in the OpenHPSDR project? I am researching the Hermes / Angelia source code, and would like minimal changes to the gateway during the testing phase of new chips.
- what problems can arise in the configuration of both chips in the gateway? I know that for RTL8211CL / E it is only changing the values ​​in the internal registers through the MDIO. Does the B50612D chip require more configuration efforts?

среда, 22 декабря 2021 г. в 09:54:44 UTC+2, softerh...@gmail.com:

[Steve Haynal]

Hi Art,

The pull-ups will probably be different for a different PHY. You have to read the datasheet, understand and design. I would have to do the same to give you a definitive answer. The same goes for changes to the gateware to support each. You have to have a good knowledge of the gateware and the device and then decide. I do not know enough about the B50162D or RTL8211CL/E device to give good advice here. Another interesting option in a 40 pin package is:

https://www.lcsc.com/product-detail/Ethernet-ICs_Realtek-Semicon-RTL8211F-CG_C187932.html

You could always choose not to use the on-chip regulator and supply voltage externally if there is noise from the on-chip regulator.

73,

Steve

kf7o

[Art]

Hi Steve.
Yes, I know some aspects of the use of PHY. There are enough examples on the Internet network for configuring the RTL8211 chip in various modifications. There are also examples of B50612D configuration, but I tend to the first option. Because I think that the Broadcom chip can disappear from sale at any time.
I think that figuring out how to configure the PHY on Verilog is not a very big problem.
The datasheet says that the RTL8211F chip has a built-in DC-DC voltage converter. And RTL8211FD - has a built-in LDO. But the latter is not on sale.
I was also looking for an external LDO  1.05 V    in a small case. I found a good version of the APL5915. But the problem is that the SO-8 case cannot be called miniature. I wonder if it is possible to find miniature inexpensive LDOs at 1.05 V in China? I searched in many stores and did not find a suitable one.

I will think further. Good luck and thanks for the answers.

четверг, 23 декабря 2021 г. в 08:26:44 UTC+2, softerh...@gmail.com:

[Steve Haynal]

Hi Art,

Two places to find parts commonly available in China are:

https://www.lcsc.com/

https://jlcpcb.com/parts

73,

Steve

kf7o

[Art]

Hello group!
For someone in the Group, the New Year has already come. For some it will come soon. I am glad that people who are close in their thoughts have gathered in one place.
Happy New Year 2022. All the best. I hope the New Year 2022 will bring us interesting communication and good results in our hobby.

пятница, 24 декабря 2021 г. в 07:45:51 UTC+2, softerh...@gmail.com:

[Steve Haynal]

Happy New Year to your Art and the rest of the group!

73,

Steve

kf7o