[hpsdr] New Hermes board

[Marc OLANIE]

Hi
Just the 2 cts of a non-technical guy (I'm writing stories for a living, far from soldering iron considerations). So please consider my thoughts as those of an "average low-end ham". I don't either understand most of the aspect of the Hermes architecture, thus I could write stupidities. Please, don't shoot the looney, hunting season is not yet open.

I totally agree with Steve remarks concerning the "open" aspect of the project. Kicad (as opposed to Altium) would be the best design/EDA option. 

I've read so many time "I'll manage to do my own front-end" (Hermes lite, red pitaya, Hermes boards etc). I definitely don't agree with this egotist approach.  Many reasons for this
-	Don't forget "unsk3lZ/un3l33t" people who cannot afford building their own. That's not an "open source/hardware" attitude imho.
-	Most front-end I saw so far were very retraining architectures (mostly lpf/BPF). Having a full coverage 120 MSPS rig just to be able to listen a few khz of hams bands is a complete waste of SDR power. Event "EB104.ru-like" filters don't seem to be a viable option
-	One "standard" has been established 9 years ago with the "alex" pinout. This variable bandpass filter approach is incredibly sexy and efficient and pervasive. It's only limitation is that you only can use one Alex board on only 1 filter's configuration. A small evolution of it allowing a real pair of independent filters for the two ADCs would be a nice add-in. Designed with Kicad indeed :-), and the patient support of "software guys" to enhance the original Alexiares SPI protocol This new version would be a nice retrofit extension for already existing new generation Hermes-like boards

Using the third Niquist zone for VHF would be a nice idea indeed. (I should test this hack on the Red Pitaya)

If I've understood the last developer's Teamspeak, It would be a great idea to consider the development of a new version of Cyclop able to support several software solution. Particularly MSAPy (open/python). Same remark about the basic VNA capabilities that should be kept without "unplugging" most or the hardware front-end of the Hermes V2.0. A specific switching board and optional coupler could be added to the hardware options to simplify plugging of the "analyzer front-end".

I have a slightly different point of view compared to Helmut DC6NY. 
-	Audio Codec is definitely not mandatory on a lan-distributed SDR network. It's an option for people wishing to recreate a "standalone" SDR, not for people who consider SDR as a new approach of hamradio like new gen. digital transmissions, new modulation schemes and waveforms, cognitive networks, signal analysis, wide band -if not spread spectrum- modes etc). 
-	300 USD limit is exactly what Hermes Lite and Red Pitaya are aimed at. I don't see the necessity of a new kid/clone on the block. Even with a 16 bits ADC. Hermes should be, from my point of view, THE high-end experimental ham oriented rig. The one you choose when you have understood how your HL or RP works. The one you definitely need to do "other things" (and sure not DXCC or Contests)
-	Undersampling, predistorsion yes. 
-	Modular building yes, thousand time yes for the reasons you explained
-	Diversity. perhaps. This kind of feature is just for some happy few having a dual antenna farm, probably less than 0.1% of ham community and less than .0001% of RF signal hackers. 
-	Real and independent use of both ADC for RX with adequate and independent filtering should be a better and smarter use of the second ADC in RX mode.
-	Totally agreed about "modular extension boards" like Niquist filtering, amps, baseband filtering, plug&play transverters and. amplifiers. I don't see the necessity to add a small formfactor "pennywhistle-like" amp on the signal/data-processing motherboard. Any kind of G6ALU driver or OPA267 or LDMos kW palet will do the job, depending on the power you need. Amp is a peripheral element of any rig, thus including Hermes, RP. (hermes lite V2.0 is an exception). Why trying to modularize the digital part of the rig and integrate the analog section? that's doesn't seems a logical method to me

- Being still compatible with the main development stream of Metis and mRX client software

- Oh, by the way, if the fpga could also be planned and used for more than predistorsion. My wildest dream : envelope elimination/restauration system (or Envelope tracking , or whatever evolution).
 
But I repeat, I'm not a tech guy. Could be wrong on certain options.
73'
F6ITU

PS : concerning the "board with already soldered high density components", I would remind that this kind of offer has a price. The question of having access to a professional pick'n place and reflow oven is not really important (ask your local hackerspace), but a matter of testing and quality control costs. Each board should undergo a series of tests depending of a procedure written by the pcb designer himself. This could double the price of the board. 

-----Message d'origine-----
De : Hpsdr [mailto:hpsdr-bounces at lists.openhpsdr.org] De la part de ad0es
Envoyé : samedi 3 juin 2017 21:21
À : hpsdr at lists.openhpsdr.org
Objet : Re: [hpsdr] New Hermes board

***** High Performance Software Defined Radio Discussion List *****