Hermes Lite 3

[randol...@gmail.com]

How many are up to contributing to a new effort to create a HL3 with faster FPGA (iproved latency)  and 10 watts? HL2 is a briliant design, mine has been rock solid for over 2 years but time moves forward. IMHO

Randy

W7CPA

[Dalton Williams]

I would rather support an effort to move Thetis to Linux to get away Microsoft increasingly fragrant monetizing its users data.  Microsoft has in the past few months taken away the ability to install with local accounts and now requires an internet connection to install.  Also it now defaults all user data to be stored on Microsoft servers.

I have installed and tried the other Linux software for the HL2 and none hold a candle to Thetis.  They are all lacking the capabilities and design of Thetis.

Dalton Williams

Amateur Radio W5EIM

--
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 visit https://groups.google.com/d/msgid/hermes-lite/ed31e484-3619-48c7-9217-31726b6253bdn%40googlegroups.com.

[Noel Woodfield]

Agree 100% Dalton.

And, why do Microsoft have to issue software patches on a daily basis?

I would rather they developed their software at their place, not in my computers.

ZL2VN

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/CAE6DSQd-Qvaq6Rr24iNvRPCbrFq5oPMGrcdY63nrWRdHh6MgGg%40mail.gmail.com.

[Greg Treble]

Try Deskhpsdr on Zlinuc it's very good

Regards,

Greg VK3KV 

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/CAOY1zUew3-7ivLrgRirRZq7OHrfEHjxbWP2ExVpKAbS6yR%3Djww%40mail.gmail.com.

[dsolbe...@gmail.com]

What you are asking for already exists, the ANAN 10E.  l had both a Hermes Lite 2 and a 2+ and sold them both after I got the 10E.  14 bit ADC, 15-20 watts, and 6 meters.

I loved my HL 2's but I needed more power to drive my 1KW amp.

If we could have an HL3 it should be equal to the ANAN 10E.

73,

Don K9AQ

[Gerry mc laughlin]

I'm for it, count me in👍 

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/fbecfd85-b28f-406d-b1bd-290c27dea60dn%40googlegroups.com.

[Ezio Insinna]

Hi friends,

It already exists, look here

https://linoobs.com.ua/brick2-sdr-hermes-16-bit-remote-hf-transceiver-160-6-m-15w-ltc2208/

I bought it and it works fine. See my preliminary results on facebook

Of course it costs twice the price of the HL2

I had two HL2 and sold them because of the low power output

73

Ezio - F5MWA / K3EK

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/CAHjGiDiahDpYipPOea7pprMLSOJWu%2BTeq_DOLjw4WDnsULuF9g%40mail.gmail.com.

[G4ZAL]

When you say 'contribute' in what form were you infering - hardware/software development or financially? 

As already said, there is Anan 10E and Brick2SDR amongst others that already exist.

If we are looking to the future, I think we should look to Polar Modulation and the hardware/software to support that.
PE1NNZ & DL2MAN showed it can be done and Hans Summers (QRP-Labs) has a quite refined version in his QMX series.
Flex Aurora is the only commercial setup that I am aware of that has it for Amateur radio (other than QRP-Labs/(tr)uSDX).

I would also like to see a 'lighter' way of remotely connecting to a radio over the internet (Thetis etc is bandwidth hungry).

I am not smart enough to help with software/hardware, but I do think the smart way forward is Polar Modulation.

Just my 2c...

Nigel, G4ZAL

[Greg Treble]

From my understanding the Brick is a better choice than the Anan10e as it comes with bandpass filters in which the Anan10e doesn't.

Myself and many amateurs cancelled our Anan10e pre orders as a result of this as I was really looking forward to a step up from the HL2 in which I owned 2 units.

I sold one of my My HL2 and now this remaining has a Tx issue so looking at repairing and future upgrade options in which the Brick is most appealing at this stage.

I have used Windows with Thetis as well as Linux with Deskhpsdr and enjoy using both and really appreciate the less resource hungry and complicated Deskhpsdr.

Oh well if I cannot resolve my HL2 transmit overload tx drop out issue it looks like I need to sell off some supplies to raise funds for the Brick ha ha cheers Greg VK3KV 

Regards,

Greg 

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/ca34ea46-ec7f-4910-b570-a92134bc8878n%40googlegroups.com.

[Phil]

I have just received my revised ANAN-10E, and it does not produce the advertised 15-20 watts. I have been in contact with both Tony and Doug regarding this issue, and I will be returning the unit for repair or a refund. Aside from the power issue, the unit performs well. I get to see how good Apache Labs support is. Fun.

Regarding an updated HL2 or HL3, I am also interested. I agree that Polar Modulation is the way to go.

Best regards,

Phil

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/CAHwnNzPtsP810Dj6CJE-jSosEymz9_i3un8Y2g-%2BbmgD%2BNPxug%40mail.gmail.com.

[randol...@gmail.com]

Hadn't seen the Brick, will explore. Thanks. How's the latency just doing stuff?

HL2 FPGP ia not impressive.

Much thanks.

[V31KC]

Sounds like a great Idea. I'm in!

[Brian Dotson]

Brian Dotson, KN4RDF here.

I like the idea of being able to reliably output 10 watts for better linear amp drive. I have checked out the Brick 2, but am concerned that I have found exactly one radio amateur who has a live video demonstration of it, and no real reviews. The Brick community doesn’t seem to be established yet, and it seems safer to build on the Hermes Lite 2 platform to me.

The Anan 10E is out of budget for me at $800+.

I would like to contribute from a hardware design perspective. I also could help with low - level firmware coding as long as it is in C, and not C++. I am a retired electronics product design engineer, but new to this community.

My CV:

www.linkedin.com/in/brianrdotson

May I ask the group what schematic capture / PCB design format the Hermes 2 hardware currently exists in? I haven’t checked the GIT repository yet.

Cheers,

Brian Dotson

In a gentle way, you can shake the world!  

-Mahatma Ghandi

 

--

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 visit https://groups.google.com/d/msgid/hermes-lite/609bf240-6964-4c81-9c01-ffff5798c953n%40googlegroups.com.

[Brian Dotson]

Never mind the hardware source format question. I checked the GitHub site. I have done exactly one design in KiCad (5 or 6 years ago), but I can come up on that. I was an Altium Designer guy for most of my career, but since I decided to stand down my consulting business and retire, THAT is certainly no longer an option as the cost doesn’t suit hobby pursuits!

Brian Dotson

In a gentle way, you can shake the world!  

-Mahatma Ghandi

 

[Mike Serino]

I cannot speak to cost but I can report that I am very happy with the Brick 2 I just installed. I do not need an intermediate amp for tuning nor powering my Mercury 3 amp to full power. The attenuator seems to be improved and mic and speaker access from the front panel are useful. I agree that the support community is very limited and AI was not much help. It is protocol 2 but not exactly a HL2 or Anan 10e. I think the manufacturer should spend more time on documentation and integration with Thetis.

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/CACti%3DzFvCb9Ko92TzoPU8-MebzTUHLjOx8mU0AkecpeYHi8doQ%40mail.gmail.com.

---

I am using the Free version of SPAMfighter.
SPAMfighter has removed 72268 of my spam emails to date.

Do you have a slow PC? Try a free scan!

[Clifford Heath]

Many different suggestions and desires have been raised in this thread. Not all of them are compatible! Not all of the respondents were clear about which requests they support. So here is my summary of the issues, with some commentary. I hope it helps people keep the different suggestions separate. Please try to identify which issue you are addressing or agreeing with :)

At the bottom, I add a few new suggestions.

1) Faster FPGA (Randy)

It’s not clear how Randy thinks this will improve latency. There’s very little buffering and no network retransmission in the HL2 - no data is being stored which could contribute to latency.

2) 10W PA (Randy, Brian)

It should be fairly straightforward to incorporate an amplifier with more output into the existing design.

3) Port Thetis to Linux (Dalton, Noel)

Dalton doesn’t elaborate on which features of Thetis he likes that are missing from deskhpsdr, linhpsdr, quisk, and other Linux-capable software.

4) Completely different radios at a much higher price (Alan 10E, Brick)

These have more powerful Tx, and use much faster (122MSPS) and wider (14 bit) ADCs, so need more processing power. They also cover 6m because of the higher bandwidth.

The Brick appears to be in violation of open source licenses. At least I can find no mention of source code. Only community support is available, and it’s not clear how well that can work with a closed-source design.

It’s not clear how much a 14-bit ADC helps. HF is limited by the atmospheric noise. The main benefit would seem to be less need to rely on bandpass filtering.

5) Polar Modulation (Nigel, Phil)

While Polar modulation can increase the efficiency of a PA, the efficiency loss of the 5-15W in this class of radio is insignificant. It only starts to matter when you get to using a larger external PA. What PAs support polar modulation? How many people have one, and would benefit from using it?

To add a drive capability to the HL2 for an amplifier that supports polar modulation is possible, but it requires two DAC output paths, so the AD9866 would not be used for output. It’s not just something you can add to the existing design, though the FPGA is quite capable to doing the required work.

Including a much more powerful PA which uses polar modulation is definitely possible, but is a bit out-of-scope for this class of radio.

6) Bandwidth reduction (Nigel)

“Thetis is bandwidth hungry” only if you configure it for multiple channels, or not reducing from 384ksps IQ data. Without moving all the demodulation and PureSignal into the FPGA and use the network only for audio and waterfall data, it’s not clear that there is a problem here worth fixing.

The HL2 has very limited buffering (insufficient for TCP), which means network reliability is paramount. That can’t be fixed without using a design that has much more RAM. A new protocol won’t fix it.

7) VHF/UHF support (Clifford)

Some folk use a transverter for this currently. If we need to change the RFSOC, we could use one that covers higher frequency bands. I am currently designing a new SDR that does exactly this. Stand by for more info :)

8) More CPU power (Clifford)

The FPGA in the HL2 is currently implementing all of the Ethernet stack and the Metis protocols in hardware. By adding a $10 ARM processor alongside the FPGA, all that work could be shifted out of the FPGA, leaving just bare IP packets to go through the existing Ethernet connection. That would release half the FPGA to do other work, and make it easier to improve the network stack.

9) New WEB front-end as a replacement for Thetis (Clifford)

A new SDR program requires a new GUI, and one way to develop such GUIs is using a web browser with a localhost server. It’s possible to achieve very high levels of usability in an interface that can be easily improved, customised or personalised, and use web sockets to stream real-time data. This approach could remove all the operating system platform dependencies from the actual SDR software, completely separating the radio software from the UI software. This is the path I plan to take.

Clifford Heath.

[Brian Dotson]

Clifford,

Thanks for a very clear summary of everything that has been proposed on this topic, and your assessments of their value! At this stage, I haven’t dived into the hardware design much, but I can offer the following few comments on your suggestions - see below.

Brian Dotson

On Sun, Feb 8, 2026 at 3:01 PM Clifford Heath <cliffor...@gmail.com> wrote:

Many different suggestions and desires have been raised in this thread. Not all of them are compatible! Not all of the respondents were clear about which requests they support. So here is my summary of the issues, with some commentary. I hope it helps people keep the different suggestions separate. Please try to identify which issue you are addressing or agreeing with :)

At the bottom, I add a few new suggestions.

1) Faster FPGA (Randy)

It’s not clear how Randy thinks this will improve latency. There’s very little buffering and no network retransmission in the HL2 - no data is being stored which could contribute to latency.

[Brian D] I guess that I am relieved to hear that opinion. I think it makes sense not to expand the FPGA processing and keep the readily available and moderately press Altera/Intel Cyclone IV device. If we crank up the resource demand, the price could go way higher than the current $50 cost.

2) 10W PA (Randy, Brian)

It should be fairly straightforward to incorporate an amplifier with more output into the existing design.

[Brian] We could get this done rather quickly, I think, if we held the next design revision to just an augmentation of the existing linear PA design, and delayed the polar modulation proposal in item (5). If the other improvements are agreed upon, we could roll them all into Hermes Lite 3 and just release a Hermes Lite 2+ version ASAP.

3) Port Thetis to Linux (Dalton, Noel)

Dalton doesn’t elaborate on which features of Thetis he likes that are missing from deskhpsdr, linhpsdr, quisk, and other Linux-capable software.

4) Completely different radios at a much higher price (Alan 10E, Brick)

These have more powerful Tx, and use much faster (122MSPS) and wider (14 bit) ADCs, so need more processing power. They also cover 6m because of the higher bandwidth.

The Brick appears to be in violation of open source licenses. At least I can find no mention of source code. Only community support is available, and it’s not clear how well that can work with a closed-source design.

[Brian] I can't find source code either. Good observation!

It’s not clear how much a 14-bit ADC helps. HF is limited by the atmospheric noise. The main benefit would seem to be less need to rely on bandpass filtering.

5) Polar Modulation (Nigel, Phil)

While Polar modulation can increase the efficiency of a PA, the efficiency loss of the 5-15W in this class of radio is insignificant. It only starts to matter when you get to using a larger external PA. What PAs support polar modulation? How many people have one, and would benefit from using it?

[Brian D} Agreed. Assuming the proposed 10 watt output  PA is only say 50% efficient, we are still talking about getting rid of 10 watts of heat, and the thermal management needn't be challenging. We probably would need to add a little fan, and that would impact enclosure size? design? tooling? We should have lower hardware design complexity and risk (read that as we can have it faster) than taking on a much more leading - edge polar modulated design.

That being said, it would be worth doing a short study on this. Clifford, to partially answer your “how many people have one” question, Hans Summers designed one into his QRP Labs QMX design (only 5 watts). I own a QMX Mini which I assembled from his kit. He was inspired to do so, interestingly enough, by examining the work of the designers of the uBITX on the low end of the cost range to the Flex Radio Aurora 500 watt transceiver on the high end. In the latter, efficiency certainly made a difference! All three, the two predecessors, and the QMX, used designs that heavily depend on software implementation. He covers the details of his implementation in both hardware and software from about 4 minutes 30 seconds on in this 43 minute presentation at the 2025 Radio Society of Great Britain convention:

https://www.youtube.com/watch?v=21uE341GtGU

I think that polar modulation, along with easing the thermal management burden, could result in a cleaner SSB signal, perhaps even negating the need for Pure Signal processing? That is a wild guess; I don't know. The questions that come immediately to mind are:

a. If we do this, would we want to add an ARM processor to handle the software modulation scheme, increasing the cost? You already proposed that possibility in item (8) below, Clifford.

b. Could the software modulation be done in the existing Altera Cyclone IV FPGA? It sounds like we already have the FPGA heavily loaded.

c. Do we have the DSP whiz kids in this community to duplicate what Hans describes in our own open - source implementation? I am guessing that we have all the necessary hardware. Hans makes it clear on the QRP Labs website that his code IS NOT open - source.

To add a drive capability to the HL2 for an amplifier that supports polar modulation is possible, but it requires two DAC output paths, so the AD9866 would not be used for output. It’s not just something you can add to the existing design, though the FPGA is quite capable to doing the required work.

Including a much more powerful PA which uses polar modulation is definitely possible, but is a bit out-of-scope for this class of radio.

6) Bandwidth reduction (Nigel)

“Thetis is bandwidth hungry” only if you configure it for multiple channels, or not reducing from 384ksps IQ data. Without moving all the demodulation and PureSignal into the FPGA and use the network only for audio and waterfall data, it’s not clear that there is a problem here worth fixing.

The HL2 has very limited buffering (insufficient for TCP), which means network reliability is paramount. That can’t be fixed without using a design that has much more RAM. A new protocol won’t fix it.

7) VHF/UHF support (Clifford)

Some folk use a transverter for this currently. If we need to change the RFSOC, we could use one that covers higher frequency bands. I am currently designing a new SDR that does exactly this. Stand by for more info :)

8) More CPU power (Clifford)

The FPGA in the HL2 is currently implementing all of the Ethernet stack and the Metis protocols in hardware. By adding a $10 ARM processor alongside the FPGA, all that work could be shifted out of the FPGA, leaving just bare IP packets to go through the existing Ethernet connection. That would release half the FPGA to do other work, and make it easier to improve the network stack.

9) New WEB front-end as a replacement for Thetis (Clifford)

A new SDR program requires a new GUI, and one way to develop such GUIs is using a web browser with a localhost server. It’s possible to achieve very high levels of usability in an interface that can be easily improved, customised or personalised, and use web sockets to stream real-time data. This approach could remove all the operating system platform dependencies from the actual SDR software, completely separating the radio software from the UI software. This is the path I plan to take.

[Brian D] It sounds like that is going to happen; that you are already planning to move on it yourself, Clifford. I would also like to point out that it requires software developer(s) with very different skill sets than the proposed hardware and firmware improvements we are discussing above, so that the effort in the nine improvements proposed here do not have to impact the SDR program in terms of development time. They can easily be parallel efforts in this community given its size.

Clifford Heath.

["Christoph v. Wüllen"]

I wonder why nobody has raised the need for going Protocol-2.

This will automatically eliminate the "bandwidth" problem with PURESIGNAL
when working at higher sample rates, because in P2 the PS sample rate
is fixed at 192k and not tied to the RX sample rate.

> --
> 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 visit https://groups.google.com/d/msgid/hermes-lite/06632335-A397-4436-8F48-27E7285986A6%40gmail.com.
>

[Ronald Meier]

Personally I do not see any need for HL3 unless same price level as the current HL2 and Protocol-2. Reason: we have Square SDR around $600 (with same nice add-on), Anan 10E around $900, Anan G1 around $2000 and Anan G2 around $3.700. So enough to choose from.

Rather would see an easy starter DB for Thetis to avoid steep learning curve and make it more plug- and play.

73 Ronald PH7R

Op maandag 9 februari 2026 om 10:35:01 UTC+1 schreef "Christoph v. Wüllen":

[Brian Dotson]

Hi Ronald,

Of the alternatives that you mentioned, only the Square SDR fits my budget, yet it costs nearly twice as much as the HL2 without offering a power improvement. I specifically need 10 watts of output to drive my RM Italy HLA150 linear amp to 100 watts. As I said, the PA update is a hardware effort, and could easily happen alongside an upgrade of Thetis that is a priority from your perspective unless as a newbie to this community I misunderstand the project process somehow.

Regards,

Brian Dotson KN4RDF

[Ronald Meier]

Brian, I see your point. Many HL2 users use, for example a 50W PA (cheap Micro PA50 from Ali), which drive the main PA which deliver 200W-1.500W. If you have $$$ there are PA's which can handle 5W to get >1.000W output.

Again my point is that from an availability point enough choice already. 

regards Ronald

Op maandag 9 februari 2026 om 17:51:30 UTC+1 schreef Brian Dotson:

[Robert Nickels]

As one who started with Hermes-Lite 1.0,  I reckon there may be some of
us who'd be interested in seeing a "back to the future" project that
returns to the original roots of a low-cost direct digital conversion
SDR transceiver platform aimed at learning and experimentation.

73, Bob W9RAN

[Ronald Meier]

Bob, what about low cost QMX+ (polar modulation) getting to work with Thetis / piHPSDR :-)?

73 Ronald

Op maandag 9 februari 2026 om 20:31:25 UTC+1 schreef Robert Nickels:

[Ken]

My HL2 runs with sparksdr , currently with 4 virtual receivers on 40, 20 15 & 10m sstv running 24/7.

I would like to see the HL3 expanded to include 6m for ft8 and 2m for sstv, enabling me to turn off two more transceivers that currently run 24/7 in the shack.

It would mean two more antenna inputs to the HL3- whether that is possible or practical I don't know- but then this is currently a wish-list, not a technical specification🙂

I'm also not concerned with transmit except for 6m ft8. For me, all the rest are strictly receive only.

--
Cheers, Ken
bat...@gmail.com
https://vk7krj.com
https://vk7krj.com/running.html
----------------------------------------
A baby can be defined as an ego with a noise at one end and a smell at the other.
Your job as parents is to teach them to control all three.
My job as a grandad is to tell you how you are doing it all wrong!

[Ron Lewkowicz]

I suspect at some point the chips required to build the current HL2 will become unobtainium so this would be the most compelling reason to begin a new project. 

Thetis is not needed for Linux.  Linux has absolutely excellent applications available already. If there is a "feature" of Thetis missing it's likely a simple matter to add it into an existing Linux app.  

[Gerry mc laughlin]

A lot of the higher power auto-tuners require more than 5w to function. A 10w HL3 would fit the bill nicely.

--
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 visit https://groups.google.com/d/msgid/hermes-lite/860ce371-e69d-414f-94bf-ed431ec671a2n%40googlegroups.com.

[John Williams]

It is important to note that the AD9866 is currently listed as "not for new designs." It is an old chip with the data sheet published in 2004. End of life is probably not far off, in my opinion.

I am liking the ANAN10. Can't seem to find many details on it under the covers or reviews of it. 

John W9JSW 

---

From: herme...@googlegroups.com <herme...@googlegroups.com> on behalf of Gerry mc laughlin <gerry...@gmail.com>
Sent: Monday, February 9, 2026 4:45 PM
To: Ron Lewkowicz <ron.le...@gmail.com>

Cc: Hermes-Lite <herme...@googlegroups.com>
Subject: Re: Hermes Lite 3

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/CAHjGiDhvDfHeLE_b7W2_zwVWMvOUreWR47zyoWx_p3TVpTRAGg%40mail.gmail.com.

[Antonio Del Manto]

anan 10 does not have band pass filters, buy a brick 2 16 bit

[Brian Dotson]

@John Williams , your concern about the AD9866 approaching EOL is legitimate asi it is at the heart of the RF path and substituting another IC solution would very likely mean starting from scratch in terms of hardware and firmware. However, even though it is listed as "NRND" this is not a problem THIS year. Here is what Google's Gemini AI says after I queried it to go through ADI's published data on the part:

https://www.google.com/search?q=when+is+a+possible+end+-+of+-+life+buy+date+for+the+AD9866BCPZ&sca_esv=e8bb71c9cc7ce2dd&rlz=1C1CHBF_enUS931US931&sxsrf=ANbL-n4_zcAtJH9MFxRBGtxaJG8JBxp-FQ%3A1770848085809&udm=50&fbs=ADc_l-aN0CWEZBOHjofHoaMMDiKpaEWjvZ2Py1XXV8d8KvlI3o6iwGk6Iv1tRbZIBNIVs-5-bUj3iBl-UxHsANYwOkWWQqZAJJdwuRaSoLHfELMHAQFneUwKM50jpvR3lgPoPKPocSAVLVIC1tmtj18NQh34sJoMKkeMLmVT0BQXR6brJvbElwncIYMxPc2e7aIP7xyAYnsJEw_6y0zcNxiTmikcViftsA&aep=1&ntc=1&sa=X&ved=2ahUKEwjgiOyYu9KSAxVKOjQIHW9ZCA4Q2J8OegQIDxAE&biw=1076&bih=520&dpr=1.75&mstk=AUtExfAbVakNmbqB0Gk1L4dreQVwm5AU9NtLKBSBGOVTUkTWlHYNpsZI1djJYomdDESj72VRe8_wxoSIN2D7MWxUTPSqWVF_zLhNauUN1uqVqSJIVDe2lBaKIdhjAReuw4fySk3yEdtgkrZ7BQ69ys9QcspZ3N0AcOovV4NmHZSYSw3PizDcqkJsTeIWX0NM9p0tk7vsmv7hAkrtjmnAR1mLXmPxARzc1044-7GNqx6Ngj0BqADHlMlmgZ_MxE0e7lYGn1pV-mezU0bxSAefmHJanyVfMJcLqf26z7Aj5Xau35IKOt4cti0XNeNq3-ZbvH69lJzhCnF-ctpIX7ZioXWaX0k1aKOC8QA2Nsvppf1G4UA-x__NKfNjpNeOKNaD4j0sZrsygHKa5msyuhQf56FIg45TZpUd7ZiCsg&csuir=1&mtid=ugGNacHtOuzb0PEPjp--gQw

Octopart shows good in-stock status at distributors worldwide:

https://octopart.com/search?q=AD9866BCPZ

We can't know what the estimated number of years to end - of - life without one of the pricey subscription - based component lifecycle manager tools (like SiliconExpert) would give us, I don't have access anymore post - retirement. Does anybody know someone who could run a query in SiliconExpert?

- Brian Dotson KN4RDF

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/BL0PR12MB2562EEC325E9DAFCF009AF6BA465A%40BL0PR12MB2562.namprd12.prod.outlook.com.

--

["Christoph v. Wüllen"]

I think there is *no* reason to worry about the AD9866 NOW, but whether it should be the basis for a new
project that is expected to live for about 10-12 years is another question. On the other hand,
I am not aware of a newer 1:1 replacement.

But I have so far not heard an opionion on the FPGA model. Should it be "FPGA only" or rather
a close-coupled FPGA/CPU pair such as in the RedPitaya? The FPGA/CPU pair has some distinct
advantages but it need be discussed in terms of cost and availability of developer tools.

Yours,

Christoph DL1YCF.

[Robert Benedict]

If there is interest in about 10 watts out it seems reasonable to consider a polar modulated 10 watt HL2 in the same size case as today. 
I think 10 - 20 watts is a sweet spot for small rigs. It's interesting that today we only see polar modulation on ~5 watt rigs like the uSDX and QMX and the high end 600 watt flex with nothing in between.
The Brick 2 SDR 16 bit is 799 Euros, so  for those of us in the US about $950 before shipping, Trump Tax and shippers extra fees for handling the tax payments.

[Brian]

It would be useful to bring out the two signals used by an EER PA so that any power level could be implemented. The two signals are: 1) a PWM representation of the envelope; 2)a hard-limited version of the composite RF output representing the phase. It would also be helpful to provide complementary versions of each of those signals with dead time to eliminate crowbar currents in the active devices.

Brian K1LI 

--
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/RyYCaofvh8U/unsubscribe.
To unsubscribe from this group and all its topics, send an email to hermes-lite...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/a2f7c073-009a-49b3-b3a1-d3298e6c5e1dn%40googlegroups.com.

[Clifford Heath]

On 13 Feb 2026, at 3:52 am, Brian <nekv...@gmail.com> wrote:

It would be useful to bring out the two signals used by an EER PA

1) EER signals do not currently exist in the FPGA. To be "brought out” they must be calculated first.

The complex-to-polar conversion is sqrt(q^2+I^2) for amplitude and atan2(q, i) for the phase.

2) The amplitude signal is analog, not PWM.

The envelope amplifier is just a normal audio amplifier, but with a DC bias to 50% of the supply rail. Many ordinary audio amplifiers could be used, but the efficiency gains come from using class D. The PWM signals required by class D don’t have to originate inside the FPGA.

3) The phase signal is a normal RF signal with constant amplitude. It’s not PWM.

The phase signal is used to drive the output stage fully on/off, so it can be square waves. However, purity of the output signal requires that it should not have significant phase steps. Generating these signals within the FPGA is done from a master clock, which (in the case of the Cyclone IV) is under 400MHz. This forces the transition times to be quantised to 2.5ns steps, which will produce significant spurs in the output. The spurs will vary significantly with output frequency changes, so they’re not easy to filter out, or for PureSignal to correct.

The alternative is to produce analog output from an NCO as HL2 does presently, but low-pass filter it in the analog domain, before using a zero-crossing detector to generate the gate drive signals. The LPF has the effect of interpolating the time steps that were intrinsic in digital signal generation.

Bottom line: EER is not a trivial addition to the HL2, at least not we are to retain the signal purity that we love about the HL2.

Clifford Heath.

[Clifford Heath]

On 12 Feb 2026, at 10:35 am, Brian Dotson <audio...@gmail.com> wrote:
> @John Williams , your concern about the AD9866 approaching EOL is legitimate...

> Here is what Google's Gemini AI says

> We can't know what the estimated number of years to end - of - life without one of the pricey subscription - based component lifecycle manager tools (like SiliconExpert) would give us, I don't have access anymore post - retirement. Does anybody know someone who could run a query in SiliconExpert?

I can’t imagine why you think that either Google Gemini or SiliconExpert would have any advance or superior knowledge about ADI’s decision process regarding whether to stop manufacturing the AD9866.

They don’t. They can only summarise, invent or lie about published statements from ADI. Please don’t share AI results as if they are magic.

Clifford Heath.

[Brian Dotson]

Clifford,

Point taken.

[Brian Dotson]

Clifford (et al):

Thanks for that analysis. With the idea of a minimum 10 watt class AB PA, I found discussions about early attempts using a pair of RD15HVF1 MOSFETs, and from Claudio's notes, it "could oscillate under some conditions".. Here's a link to that discussion page for those who haven't seen it:
https://groups.google.com/g/hermes-lite/c/93jKwcudhg8/m/DXE1OoBWBAAJ

I am also concerned about getting the heat off the stock enclosure side panels. I think additional finned heat sinks might be needed.

It's looking more and more like an outboard 10 - 20W PA to drive my RM Italy HLA150 makes sense, with PTT timing handled properly across the chain, as someone in this thread suggested.

Someone in the discussion on the above page mentioned users pursuing an option of building a QRP Labs 10W linear PA kit, or the Pennywhistle kit (with my own switched filter setup). I would have to contact TAPR about a kit, or ask for Gerbers, and try to put together the parts myself. which is really old now. Either the QRP Labs or TAPR option would require a really low drive level, and of course I would need a switched filter covering my bands of choice.. The Pennywhistle wants 1/4 to 1/2 watt; QRP Labs PA wants much less as it has 26 dB of gain. What would be the best way to do that? How much drive can one get out of the Low Power RF output on the HL2?

- Brian Dotson KN4RDF

[Mike Lewis]

External Connections ? softerhardware/Hermes-Lite2 Wiki 17dBm on RF1.  I use it with my split IF transvers.,  RF3 is RX.

 

From: herme...@googlegroups.com <herme...@googlegroups.com> On Behalf Of Brian Dotson
Sent: Thursday, February 12, 2026 5:44 PM
To: Clifford Heath <cliffor...@gmail.com>
Cc: Hermes-Lite <herme...@googlegroups.com>
Subject: Re: Hermes Lite 3

 

Clifford (et al):

--

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 visit https://groups.google.com/d/msgid/hermes-lite/CACti%3DzFL%2BAtLx%3DoFTWW1%3DQwUkTS6GOf7ussUFz%2BrTk20K_3xdA%40mail.gmail.com.

["Christoph v. Wüllen"]

This was the first contribution which does more than scratching the surface. I think
it must be clear what the costs are. I have no idea about polar modulation but
some things are clear from 101 math and need be spelled out:

- producing a polar signal in the digital domain is trivial, but attention
has to be made to discretization effects, that is, the resolution and
the sample rate for the DAC need to be thought of a-new

- Often the discussion concentrates on the "IM3 peak", but what are the
consequences of discretization on the TX noise? In the conventional
design we had some hard lessons in the past and there should be
simulations at least that tell you how the sample rate and the
bit depth of the DAC affects that.

- we know how to make a decent class-B PA from a dozen of parts or so,
is there any idea what is the minimum effort (in $$$) to make a decent
EER-PA?

- Almost certainly, the amplitude signal to the PA must be delayed, to
compensate for the time the phase signal needs to past the first
stages. This means the output of the PA need to be sampled back into
the FPGA and the required delay need to be computed.
This may involve heavy computation which has to be done in the FPGA
(currently, predistortion is done on the baseband signals in the PC).

So I think some experts need to go through this point for point, and then
come out with how much a PA will cost and how good the FPGA must be.
This then has to be converted into a minimum $$$ for a finished product.
And only then, the discussion can continue.

> --
> 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 visit https://groups.google.com/d/msgid/hermes-lite/FF221681-72D6-407C-A0AD-A467CB35264A%40gmail.com.

[luc laureys]

@ Mike Lewis

Pls explain how? i want to do the same but no low power on the RF1 here...even with the pa off ,and full duplex on in the thetis software....

Btw...i dont have a RF3 here on the Hermes lite2..

Greets Frm ON7KEC Luc

Op vrijdag 13 februari 2026 om 09:30:07 UTC+1 schreef "Christoph v. Wüllen":

[Mike Lewis]

Before I had an I/O board I soldered UT141 from the HL2 pin header for RF3 to bring out to an SMA.

RF1 (TX low power) was already on the back panel.

 

The IO board brings out RF3 switched between Alt RX  and PURE SMA jacks.  RF1 stays as was. 

 

You can use the N2ADR IO Board Control program, or features in the SDR app of your choice, to select between ANT and Alt RX/Pure (RF3) for receive,  Disable T/R, and Disable PA. 

 

If using Pure Signal, then the SDR app will switch RF3 between the sampler on Pure Signal SMA and the Alt RX (or ANT) SMA for receive.  This depends on the SDR app to handle this, not all do.

 

Each SDR app does this differently.  If you have transverters then fewer apps support that.

 

I modified my IO board firmware to automatically switch the ant inputs because most SDR apps do not handle the I/O board yet. Based on then Dial frequency it switches to ANT for HF, Alt RX for 6M and higher.

 

SparkSDR, SDR Console, and Quisk manually set use of Alt RX, one for all bands.  

 

SDR++ an most others do not handle the IO board. Most do offer a single offset value for a single transverter.

 

Thetis and PiHPSDR handle Alt RX per band, including transverters and their offsets.

 

• Mike K7MDL

 

From: herme...@googlegroups.com <herme...@googlegroups.com> On Behalf Of luc laureys
Sent: Friday, February 13, 2026 12:58 AM
To: Hermes-Lite <herme...@googlegroups.com>
Subject: Re: Hermes Lite 3

 

@ Mike Lewis

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/9c701f9c-e74f-4c39-8cbd-b511df3c1706n%40googlegroups.com.

[nekv...@gmail.com]

Claudio, IN3OTD, has already implemented 256-kHz PWM for the envelope component inside the HL2 FPGA. Look back into the archives for details. 

Speaking as co-developer with K1KP of the Polar Explorer, which was the prototype for the CMCD PA in the FlexRadio Aurora, I have learned the importance of dead time between the phases of the drive signals to the class D push-pull active devices to prevent crowbar currents, whether at AF or RF. Dead time is easily created with just a few logic gates.

Adding a couple of comparators and logic gates to the output of HL2 doesn't seem like a heavy lift for the benefit of enabling RF output power up to the legal limit at relatively low cost. It's very much in the experimental spirit of the Thor project originally envisioned by the founders of openHPSDR.

[Steve Haynal]

Hi All,

All the discussion about a Hermes-Lite 3 is interesting. As a reminder, the Hermes-Lite 1 and 2 efforts never prioritized building an SDR to a certain set of specs. Rather, these efforts were focused on what kind of low-cost SDR could be built with the AD9866, an inexpensive commodity part which might be the heart of a decent SDR. So, much of what the HL2 can or can't do is dictated by the AD9866. The power output is 5W because I am a QRP operator. I still only run my HL2 with 5W and have worked the world. If people want more power, it may be easier to disable the onboard PA and build another external preamp which uses the HL2 low power output to generate 10W or whatever.

Only a few hundred HL2s were ever expected to be built and sold. It amazes me that there are now thousands sold and still selling, plus derivative projects such as the radioberry and square sdr.

Best of luck to all those working on a Hermes-Lite 3!

73,

Steve

kf7o

[Ed Marciniak]

The radioberry + juice pack is interesting.

Perhaps it would make sense to combine preamp/PA, filters, the juice pack, an Ethernet interface + audio codec.

I wish the radioberry had a place to solder a clock input to bypass the 76.8 MHz clock with something derived from 10MHz.

The question is, if you did a new spin of the Hermes lite, whether it’s worth the trouble instead of using the  radioberry as a component.

---

From: herme...@googlegroups.com <herme...@googlegroups.com> on behalf of Steve Haynal <softerh...@gmail.com>
Sent: Sunday, February 15, 2026 7:28:13 PM
To: Hermes-Lite <herme...@googlegroups.com>

Subject: Re: Hermes Lite 3

--

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 visit https://groups.google.com/d/msgid/hermes-lite/0b964d9b-2232-4bbc-b21f-30649d8682aan%40googlegroups.com.

[Clifford Heath]

The Radioberry is much more affected by the digital noise emanating
from the Pi, in such close proximity.

> To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/BL3P221MB0321B876FA94E689BD81A79AFD6CA%40BL3P221MB0321.NAMP221.PROD.OUTLOOK.COM.

[nekv...@gmail.com]

Marc, F6ITU, collected lots of Hermes and ET/EER related items into a repository on his GitHub. The work done by Rudiger, DJ1MR, is especially worthwhile.

[James Ahlstrom]

Here are some thoughts on getting more power from the HL2 to drive an external amplifier. They are based on an inquiry from Brian, KN4RDF.

It would be convenient to redesign the finals in the HL2 to 10 watts, or add a small internal amp and use the existing filter board. Unfortunately there are problems. The current finals run at 8.5 volts because they are specified to tolerate an infinite SWR at that voltage. That feature would probably be lost. I don't think the filter board would tolerate twice the power especially on the bands that use ferrite inductors. To find out, you would need to apply 10 watts for a time and then test for heating or burnout. I run my HL2 without a heat shim, but others feel a heat shim is required, and some have even added a fan. I think it will be a problem to dissipate the added power. I have had no luck designing MOSFET amps that run at 12 volts. They require a large bias current to achieve low IMD, and that adds to the power dissipation problem.

A highly efficient polar amp would be nice. I have no experience with these and have only a dim idea that they require a separate amplitude and phase output from the FPGA. But the HL2 has been pressured to support the maximum possible number of receivers, and no more space is available. A re-write of the gateware would be required.

The obvious solution is an external amp. The request is 10 watts, but a design that can produce a range of output powers would be nice. In this case the filter board would not be used. The HL2 output is quite clean without it. The filter board could be removed, or its filters could be de-selected in software.

Going from 5 to 10 watts is only 3 dB, and we would need a power attenuator for most amps. The -7.5 dB power adjustment could be used to reduce the power to 0.9 watts. Then we need a gain of 10.5 dB. The PennyWhistle amp has a gain of 19 dB and a linear output of 16 watts. So a low power attenuator could be used at the 0.9 watt level. Note that the PennyWhistle is built on a finned heatsink, which should worry those who think the HL2 can dissipate this power. So a viable design available today is the HL2 at -7.5dB output, going to a -8.5 dB one watt attenuator, going to the PennyWhistle without low pass filters, going to the final amp with low pass filters selected by the IO board or some other means.

A more aggressive design would use the 17 dBm output from RF1, and would require 23 dB gain to reach 10 watts. I am thinking of a design that could produce much more power but be adjustable.

Jim

N2ADR

[Brian Dotson]

Jim, and other interested compadres:

Thanks so much for taking a careful look at this, and your suggestion to answer on the forum, Jim.  I am considering two experiment options in the short term: 

1. Hans Summers at QRP Labs has a very inexpensive kit that has been offered since 2018:

https://qrp-labs.com/linear

It has more than enough gain (about 26 dB) to be driven from the RF1 output, and Hans claims that it can get to a clean 15W - ish on all bands with 13.8V supply.

1A. I have not looked at the max component height on this kit yet, but I was thinking of adding a second enclosure to my upcoming post - Chinese New Year order from Makerfabs, and actually attempting to cut a large rectangular hole in the top to mount this assembly upside - down with the heat sink out of the HLA2 enclosure's top half. The cutout would be over the existing final amp section of the board and would need to clear the edges of the PCB here, about 72mm long and about 36mm wide. If I could get clearance to avoid interference with the HL2's main board, that puts the added heat outside of the box to avoid adding heat to the HLA2's interior and threatening the FPGA and AD9866 IC. The mods to the main board are very minimal - mostly picking points to fly some RG-174 coax for the RF1 signal to the outboard amp's input, and routing the output back to the output section of the board on the output side of B98 after removing it: I would power the add-on amp straight off of the external 11 - 16V power connection on the load side of F1 after I upsize it to handle the additional current demand  I believe I can disable the on - board PA in Thetis. Worst - case, I would just remove J31 and R123 to remove power from the PA.

[Insertion point for the add - on PAs output detail.]

1B. Jim, you conservatively sized the series pass inductor and shunt capacitor sizes on your filter board for 1812 and 1206 footprints, respectively. Since I have SMT rework capability on my hobby bench, I gather that we don't control Makerfabs' kitting of the passives down to manufacturers and MPN's, so I  would just order (from DigiKey or Mouser) inductors that would meet the requisite current (0.5 amps continuous rating and Isat would do it) and capacitors rated at 100 volts, and re-work them (12 pass inductors and 18 shunt capacitors).

Please feel free to shoot holes in this idea, friends. Understand that I am not suggesting group consideration of these mods. If I attempt this, I will take full responsibility for what this plan is: a somewhat aesthetically ugly and possibly ill - advised experiment. 😅. At least I am not considering any mods to the HLA2 PCB Assemblies that cannot be easily undone.

2. Per your suggestion, Jim, just sell my RM Italy HLA150 Plus amplifier, and buy one of the new Micro PA100s that have been favorably reviewed by Kurt, HS0XPX, and a few others on YouTube. For those that want a more refined and possibly more robust amp to deliver 100W and have the budget, the 60dbm.com Neptune and the Juma PA100 have been favorably reviewed on other threads on this forum. Very recently, John Williams recommended the Neptune amp to me.

Lastly, I FULLY support a future polar amp effort for HL3, and would love to contribute to the hardware side of it! Looking at the available real estate, I think this is our only real choice if the PA stays on Main Board, and we keep the proven, compact enclosure as is.

Cheers,

Brian Dotson KN4RDF

[ron.ni...@gmail.com]

If one just wants more ADC bits while staying within the same OpenHPSDR Protocol 1 ecosystem, note that the RedPitaya SDR 122-16 (and the TRX-Duo clone) seems to be in production and available, and has open source software support for Protocol 1 (I've tested that against my own SDR apps).  In addition with the RedPitaya, you get dual ADC/DAC port for diversity Rx and Tx steering.  A project to add an outboard amp and filter board would create much of what's needed for a hypothetical HL2++. So that's another option.  73, Ron, N6YWU

[Scott Strehle]

I can chime in with some info on a project i have been working on this winter.. I built a Red Pitaya 122-16 radio and have been using it for the last 3 or four months.  My goal was to build up a dual 16bit sdr tranciever that would hang with the Anan Radios and work hpsdr.  I have got the red pitaya fed through a switching board and a band pass filter and on the output a 12 watt amplifier.   I hav a 31db step att built into the system to work with pure signal.  I also added a pi5 to the radio so i could run things self contained with a flat screen ontop ot the radio and pihpsdr.  I normally run thetis on a pc though.  

Everything has worked very well and I love running this radio probably because i built it myself.  I come from a past running HL2 as my only radio in the shack  I mostly work ssb phone and everything works great.  I am not interested in watching the whole band or skimming or anything like that only watching signals close to where my conversation is.  

My true observations about my build and setup are that there is not much difference between this dual 16bit radio and a HL2 i have not truly been able to use diversity yet but when i get set up for it i believe it will help a great deal.  As far as going from 12 to 16bit i have not really seen much real world difference.  the band pass filter doues work very well and helps a great deal.  

On the ouput side of things i usually only use abot 2 to 2.5 watts output to run my ldmos amplifier.  at that level the 4 mrf300 amplifier runs super clean and puts out about 600 watts or so.  I can push it to 1000w but pure signal works overtime so i just keep things down low and it works very well.

I can compare this build to a HL2 but that is about it i do not have an anan to compare things to so that I can confirm it works as well as it should.  All it other parts of the radio other than the Red Pitaya ar 60dbm parts and they have worked well.  If I had to guess as to how much I have into things I would say probably 1200 to 1500 usd.  It wasnt cheap but it has been fun and I have learned a great deal.  There are some pics on my qrz page if anyone is interested.

Scott

KJ7RHP

[Aubrey Thompson]

I agree!  This should be a priority, especially in view of what we have seen from Win11.  A couple of questions come to mind:

1)  Could a Pi 5 support Thetis so that a Radioberry/Thetis package would be feasible?

2)  Can we invent a touchscreen "device" that is as easy to use as a knob to make the front panel form factor all screen?

Aubrey

KN4OTN

On Friday, February 6, 2026 at 7:18:50 PM UTC-5 Dalton Williams wrote:

I would rather support an effort to move Thetis to Linux to get away Microsoft increasingly fragrant monetizing its users data.  Microsoft has in the past few months taken away the ability to install with local accounts and now requires an internet connection to install.  Also it now defaults all user data to be stored on Microsoft servers.

I have installed and tried the other Linux software for the HL2 and none hold a candle to Thetis.  They are all lacking the capabilities and design of Thetis.

Dalton Williams

Amateur Radio W5EIM

On Fri, Feb 6, 2026 at 4:28 PM randol...@gmail.com <randol...@gmail.com> wrote:

How many are up to contributing to a new effort to create a HL3 with faster FPGA (iproved latency)  and 10 watts? HL2 is a briliant design, mine has been rock solid for over 2 years but time moves forward. IMHO

Randy

W7CPA

--
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 visit https://groups.google.com/d/msgid/hermes-lite/ed31e484-3619-48c7-9217-31726b6253bdn%40googlegroups.com.

[Scott Strehle]

Pihpsdr already works very well on a pi5 and the software works very well.  It was not meant to be thetis but it has most all of the important funtionality needed to bet the job done.  I am guessing as it evolves it will come into what people desire.  I slapped a 15 inch hdmi touch screen on top of the radio and it worked great.  I use a mouse because it is what I am used to but it also has touch screen functionality.

It sounds like what people want is not a Hermes lite 2 but a Hermes instead.  In order to get that you need to spend more money to build what you want.  That is why the HL2 was originally designed and why there are so many people who love it.  It is a gateway sdr drug to get into the wonderful world of SDR.  There are other projects out there...  The Odessey sdr group is an excelent example but it seems that it has lost favor recently recently.  I purchased a used Chineese clone version of one recently and have yet to get the thing to work properly.  I understand the the original project radios were great.

Scott

KJ7RHP

[Nikos Magafourakis]

Aubrey Thompson <aubrey...@gmail.com>	7:20 a.m. (10 hours ago)
προς Hermes-Lite

I agree, Thetis on Linux ...

Nikos 

SV9CVJ

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/1dc4f631-3424-47a1-ad4d-c23ddc3827abn%40googlegroups.com.

[ron.ni...@gmail.com]

Can anyone provide a list, or describe what features they use on Thetis that they find missing or less useful in piHPSDR?  Is the difference mostly functional DSP and radio functions, or the MSWindows-appropriate UI display and controls?  (I don't have any Windows systems, so I can't do the comparison myself.)  Thanks, Ron, N6YWU

[Ryan Jennings]

I use Thetis.  My favorite features are noise reduction and ability to listen to more than one freq/band at a time.  I often have one in my left and another in my right.  I just do casual listening and some POTA hunting, but I like to stick in a place with one ear and hunt around with the other.  

[Gareth “68CT080”]

I will have to have a good read on your build. I went with a trx duo as the red is a bit too much on my pockets but love the diversity features. Have got the 10 watt amp, 100 watt low pass filter board, an the switch board from 60dbm.com also. Have just been using it as a receiver most of the time using thetis and I like it very much. I have also had it running as a basic transceiver with the 10 watt amp driving a micro pa50. Still lots to learn and finding good sources of information for mere mortals is hard to find. I have no background in electronics at all but slowly learning with this project which is so much fun. Some day I will get it completed until then the Hermes is my daily toy.

Best of luck with your build, there is nothing more satisfying as building and using your own build.

By the way I'm not a ham operator,  I live on 11 meters talking radio all round the world and swl else where, its all fun.

Take care.

[Ron Lewkowicz]

One feature in Thetis that is not in piHPSDR/deskHPSDR is the ability to set manual notch filters.  That said I've seen a youtube demonstrating setting MNFs outside their passband to reduce interference from nearby signals and I'm not sure if that is actually how MNFs work.  Reading the WDSP manual it seems setting a MNF makes an addition to a database which then only turns the MNF on once it becomes inside the current  passband. Anyway this is a discussion for another time. Maybe we could add MNF to pi/deskHPSDR or maybe not worth the trouble.

piHPSDR/deskHPSDR does have auto notch filter and I have to say it works superb.  I was not NCS today but working relay on a net and was puzzled when the net suddenly went quiet.  The NCS stopped talking. People stopped calling in.  Didn't realize what was going on until I looked up at my screen and saw someone putting a 20dB over carrier onto the net to cause interference.  ANF took it out completely. 

   

If someone wants to port Thetis to Linux then godspeed.  It will be one hell of a job.  It was never designed to be portable.  Given how good pi/deskHPSDR work though it doesn't seem worth the effort.  I bought an HL2 because of how much I liked piHPSDR.  I would still use it even if Thetis was ported.

["Christoph v. Wüllen"]

Well, noise reduction in Thetis and piHPSDR should be the same,
and having one RX on the left ear and the other on the right ear
can also be done with piHSPDR. The difference between these
two programs is mostly the "size" of the GUI.

> Am 18.02.2026 um 20:31 schrieb Ryan Jennings <ry...@rjennings.net>:
>
> I use Thetis. My favorite features are noise reduction and ability to listen to more than one freq/band at a time. I often have one in my left and another in my right. I just do casual listening and some POTA hunting, but I like to stick in a place with one ear and hunt around with the other.
>
> On Wednesday, February 18, 2026 at 11:46:47 AM UTC-7 ron.ni...@gmail.com wrote:
> Can anyone provide a list, or describe what features they use on Thetis that they find missing or less useful in piHPSDR? Is the difference mostly functional DSP and radio functions, or the MSWindows-appropriate UI display and controls? (I don't have any Windows systems, so I can't do the comparison myself.) Thanks, Ron, N6YWU
>
> On Wednesday, February 18, 2026 at 7:59:38 AM UTC-8 Nikos Magafourakis wrote:
> Aubrey Thompson <aubrey...@gmail.com> 7:20 a.m. (10 hours ago)
>

> προς Hermes-LiteI agree, Thetis on Linux ...

> To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/7e563e82-c46e-4c11-bdd1-eb164d4f15b0n%40googlegroups.com.

[Joey G. (Bonehead)]

I have to agree.  PiHPSDR is perfect for my project and have no issues.  I wanted a portable unit so I mounted a Pi inside the HL2 case and a touch screen on the outside.  Makes a pretty good QRP rig for portable use but I have a small amplifier I carry around if I need around 40w.  Added an encoder and a few buttons on the top to change bands and mode quickly.  No modifications to the HL2 board itself.  Here's a few pic if anyone else would like to make a portable system. 

[Brian Dotson]

Joey,

Nicely integrated! You have inspired me! Maybe I CAN use my soon - to - be - acquired HL2 for SOTA expeditions, after all! I have seen no accounts of portable ops with the HL2 on the web. Thanks for sharing your use case!

Cheers,

Brian Dotson KN4RDF

To view this discussion visit https://groups.google.com/d/msgid/hermes-lite/66168633-cdc3-4958-8bd0-f73e00da834cn%40googlegroups.com.

[ron.ni...@gmail.com]

I modified an SDR app that I wrote for my iPhone app so that I can use my HL2 for POTA/SOTA portable ops (in Apple's App Store, or email me for a TestFlight invitation to get a no-cost copy).  No modifications to the HL2 required.  CW and SSB RxTx supported.  An older used iPhone plus an ethernet adapter weighs not much more than a Raspberry Pi + display.  And with a longer ethernet cable, one ends up with a smaller lighter handheld SDR radio console than bolting on a Pi display.

73, Ron, N6YWU

[Clifford Heath]

On 21 Feb 2026, at 8:00 am, Joey G. (Bonehead) <jcgre...@gmail.com> wrote:
> I have to agree. PiHPSDR is perfect for my project and have no issues. I wanted a portable unit so I mounted a Pi inside the HL2 case and a touch screen on the outside.

Nice job!

I’m curious however whether the PI or its power supply (in such close proximity to the HL2) generate noticeable interference or noise?
The Radioberry is known to be noisier than the HL2 for this reason.

I have a similar project (incomplete) which will also incorporate a 10” touch-screen and battery pack.

Clifford Heath.

[Joey G.]

Thanks!

I have not noticed any noise on the bands I use.  I figured the step down converter dropping the 12v to 5v would but to my surprise, nothing.  I tested it out in the country away from most electrical noise. 

73

Joe

K5BHD