Forum.PowerSDR.de

Just another general remark

Completely agreed with Rene opinion : original Hermes lpf is not the best you can find. But using an 9th order filter would have taken quite a lot of place on the board.
some real improvement could be done, using T68 -or T60 at least- toroids. The original filter uses T50 only. Using capacitors with a far better Q will help considerably too. In other words, keeping the very same structure than the one designed by Phil Harman but with better component will offer better performances without changing a lot of things.

Performances of the Mini Circuit RPL50+ are definitely not exceptional. I'll try to follow Rene's advice, bypassing this filter and adding another permanent hi-Q 55 MHz lpf.
Tnks for the idea

Marc

... just a short note to DC9EO

Please Edwin, keep on writting in German. I really can read your language (und ein wenig Bayerisch also) without difficulties (6 years of German at school) but I would feel ashame if you could see how much bad my German grammar is.
VY 73'
Marc f6itu

Hi
in fact, the "Alexiares" structure of the lpf section is in the TX path, and the 6m lpf filter is permanently "in circuit". In other words, when receiving, the signal goes thru a first "toroid based" antialiasing filter AND the MCL integrated lpf located at each RX input, just before the ADC driver.

I'm not particularly found of a "full smd" hpf section, as the Q and other limitations of each component is far lower than the one obtained with large toroids (ref: the PA3AKE writting concerning the IM of toroids in his "h-mode mixer frontend" ,
https://martein.home.xs4all.nl/pa3ake/h ... intro.html
chapter "back to the toroids and imd"

imho, the use of smd components in some commercial appliances has essentially been dictated by economical considerations (and the bpf architecture is another nonsense on a DDC/DUC architecture, once again, imho).

I may be wrong, but in this case, I'm not the only one
73'
Marc

ps : on my 14 bits Red pitaya board, I'm using the very same approach DC9OE described in his post

Hi Rene

>gibt es eine möglichkeit den SPI für Alex und den I2C fürs Audio gleichzeitig zu nutzen?

as far as I know, the I2S is sharing the same output with the Alex SPI bus... so I'm affraid it's not possible

>Betreibt von euch jemand den RP mit einem Alexsystem?
yes, I do

>Wenn ja, funktioniert das Problemlos?
so far, so good. The general behaviour is FAR better than the partial I2C order equivalences

>Ich teste es gerade und musste feststellen das es so eigentlich nicht richtig funktioniert/.../

For which reason did you need a stepdown? (if I understood well, which is not certain) the 12V part of Alex is totally separated from the spi input, which works at TTL level (and 3.3V signals comming from the Red Pitaya are far enough to trigger the pair of 74HCT04 ).

>Wenn ich nur die Bänder schalte wird irgendwann der LPF nicht mehr mitgeschaltet, reproduzierbar zu 90% von wechsel 40m <->80m, nach kurzen PTT betätigen gehts wieder.

That's a strange and anoying reaction for sure. I've remarked that the lenght of the SPI bus should be kept rather short to avoid troubles caused by the noise (and perhaps RF strays). But with rather short cables and ferrites, I swear, I never had random band switching.

> Pavel sendet ab einen bestimmten Punkt nur das 16bit RX Filter Signal,

I'm awfully sorry, but my German is really rusty and I don't catch your whole argument. If you only consider the port of the OpenHPSDR spi bus made by Pavel, it works with both LPF and HPF (independantly) and I can select the HPF I need when fiddling with the Filter setup window of PowerSDR.

But anyhow, I will test my configuration again and try to reproduce your problem. If there is a bug, I'm pretty sure a fix will be provided...

... And I will be interested in every testing configuration you could provide.

I feel terribly sorry not beeing able catching all the subtilities of your message and answering in German (even after 6 years of school)

73' René
De Marc f6itu

Answer to myself : the power variation of the "drive" cursor in PowerSDR gives a constant delta of 20 dB, and the "PA Gain" gives the possibility to "offset" these 20 dB, thus controling the max and min output power delivered by the Red Pitaya board.

If PA Gain is set to 38.8dB, the RP will deliver an RF signal from -10 to + 10 dBm (max power) on every bands (except on 6 meter, where the signal is lower, at 5.7 dBm)

That means that if you have an amplifier that needs, say, a source of -13 dBm to reach it's max output power, tuning the "PA Gain" at 61 dB will tell the Pitaya to never deliver more than -12 dBm (and giving -31 dBm when the "drive" cursor of PowerSDR is set to "1")
Your amp will never die or distord signal due to overdrive, you won't have to worry for your SSPA.

I asked another stupid question

Hi Red Pitaya Hams

I would like to know which power level do you use to reach the best tradeoff when transmitting with the Red Pitaya.

By default, if all TX params remain unchanged, the "full throttle" output power is around 0 dBm. The max output power reaches between 9 and 10 dBm when "power gain" settings is downed to 0.

Using a lower output signal would certainly extend the lifespan and lower the temperature of the RP, but the full dynamic of the signal will be also probably impacted.

So, between -10 and +10 dBm, which seems to be the best for your daily use ?

Many thanks to all
VY 73'
Marc f6itu

I've probably read more than 100 times Pavel's notes... how could I have missed this. Or, worst, refuse to see it.
"slow analog inputs can be used for the forward (Analog input 0) and reverse (Analog input 1) power measurement"

I really feel ashame not Reading The Fucking Manual

Really
Marc

Hi Ralf and Chatt

I'm afraid nobody can answer your question. In fact, everything depends on “what” you intend to measure, and “with which precision”.
In other words, basic applications are perfect to explain instrumentation and measure to students. But not more. Let’s just give an example: you cannot change the sampling rate when narrowing a spectrum window in the Spectrum analyzer application. You’ll have to develop and adapt (upgrade) the firmware and software part of your app before using it.
Another handicap will be the physical limits of the input and output ports. In terms of voltage, current, power… You’ll always have to add a frontend (voltage divider, buffer etc) between the Red Pitaya and the “Device Under Test”.
But if you don’t need accuracy and are looking for a variable geometry device, the RP will be perfect.
Imho
Marc

Thank you, Siegfried and Yevgueni

I was just remembering something like that, but couldn't retreive this info.

I'll try to add some buffering or limitation (Zener ? ) to avoid killing the gpio. Thanks again to both of you, you make my day :- DD

Marc f6itu

Hi

Amongst the many way PowerSDR mRX manages the forward and reflected power display, one consists using 2 analog input (plugged to a so called "commodity ADC on HPSDR hardware), those two lines comming from the FWD and REV power voltages delivered by a simple VSWR tandem bridge.
I'd like to know if such a possibility has been considered in Pavel's firmware (using for example analog inputs available on E2 connector)

Cheers
Marc f6itu

Hi Winfried
- did you configure PulseAudio to "connect" the linhpsdr input on "simultaneous output to audio internal analog stereo" ? as long as PulseAudio don't "see" linhpsdr, you won't be able to ear anything... you'll have to fiddle a little bit (output and input peripherals etc... try the different tab of the PulseAudio volume control window
- in LinHPSDR /configure button (main window)/ Tab "RX-0"/bottom of the popup windows, did your check the "local audio" checkbox ? did you choose the right "audio output" channel (in my config : simultaneous output to audio etc etc)

the "audio checkbox" is viciously hidden in a submenu... I had the same problem looking for it and trying to listen my rig

Hope it could help
73'
Marc f6itu

Hallo Thomas

I would say "no" if you are not using the VNA function, if you don't intend to use the FT8 digital mode, and if you regularly update your local system (with the "apk" utility)

from a more general point of view, I will upgrade anyway.... for security reasons. A new version of the kernel, a new version of the language, these are two good reasons to upgrade (I'm often using my Red Pitaya "SDR" on networks used by nice and efficient hackers, and even if nobody is a real blackhat, I would certainly feel better with the last fresh release).

This is a very personal point of view. From a pure "radio" perspective, keeping the "old" system is perfectly viable.

73'
Marc f6itu

Hi

Just a stupid question : what is the impedance on the sma fwd connector when something is plugged on it ?

the input Z of the AD8307 is 50 Ohms. The output network of Loftür's coupler is 50 Ohms also
https://3f7d437a-a-62cb3a1a-s-sites.goo ... edirects=0
the output of the fwd sma (ps out) is 50 Ohms again

so what is the impedance of 3 x50 Ohms loads connected together ? certainly not 50 Ohms. Unless you disconnect the AD8307 input circuit, you should add a 6 db splitter inside your coupler. (3x16 Ohm resistors in "Y" configuration, or a 3x50 Ohms splitter in "delta" configuration )

https://www.radio-electronics.com/info/ ... ivider.php

another solution would be to keep preciously your coupler aside, and add a capacitive or resistive pad.
https://www.webx.dk/oz2cpu/radios/miliwatt.htm
chapter "How To Measure High Power"

A capacitive pad only needs 2 high voltage capacitor

To give you a more precise idea, you can have a look on the last version of W6PQL's coupler with it's "pure signal" sampling output
https://www.w6pql.com/hf/hfddd-assembly3.htm
The 6 dB splitter is located in the forward branch.

You always can connect PS on your sma FWD output without precaution. But your power reading will be wrong and the insertion losses of your coupler will go up. The tandem match configuration is easy to build, but if you don't care, it's bandwidth will be narrowed and it's VSWR will increase (which is not really logic for a device used to measure VSWR)

73'
Marc f6itu

Bonjour Alexander

Did you check the email address written at the end of the article (dc2pd at darc.de) ?

73'
Marc

Hi Uli
The simplest way to analyze a I2C bus is to plug an entry level logic analyzer on the output of the Red Pitaya
https://www.amazon.fr/AZDelivery-Logic- ... alyzer+fx2

If you are in a hurry, your also can use an Arduino (nano, uno, whatever) and load an I2C sketch
https://playground.arduino.cc/Main/I2cScanner

if you intend to protect your I2C bus, you can add an ADUM isolator
http://www.analog.com/media/en/technica ... 0_1251.pdf
Several model can be used, choose the one you prefer

The only precaution you must have when using these components is to use a really close decoupling cap of the vdd rail. Adum are prone to oscillate, but a simple 100 nF will calm down the chip

hope this will help
good luck
Marc f6itu

Oh, certainly Sigy
I’ll try to be as precise as possible and will always answer (or try to answer) to any question
In a few words : We are a group of French hams who decided to build a series of “HPSDR compatible” rigs. Some decided to build an angelia pcb (anan 100D) from component level, others have chosen the Red Pitaya, and some have decided to build the KF7O’s Hermes Lite v2.0 (and I must admit, many of us are building the three configs)
This is the main reason why I tried to retrofit the Alexiares Frontend : “one filter to rule them all” , with some constraint :
- It has to be and stay “open hardware” (CERN licence for original designs)
- It must be 100% protocol compliant with the original HPSDR code
- It could be build by newcomers (all boards are using SMD component, but nothing under 0805)
- All design could be downloaded on github (kicad schematics and pcb)… Please fork as much as you want
- Dimensions must fit in the “cheapest” form factor sold by Chinese pcb makers (10x10 mostly)
- We could supply pcb for ham who don’t want to run a 10 pcb batch for each board (not yet in place, but it will)
- This must be a non-profit initiative for DIY addicts
The Red Pitaya “integration” is in fact the proof of concept and real life benchmark of these boards
Just some picture to illustrate (the box is a former 4 channel Gould oscilloscope)
This is the upper layer of the rig. The piece of paper with some orange traces is a “dummy” to mark the size of a full size Angelia board. On top, the red pitaya, on the bottom, the SPI control board and the two filters, hpf and lpf. This picture has been taken before a neat wiring
The next picture shows the “power” stage. In the lower right angle of the box, the blue pcb Is in fact an I2C board. A kind of DC2PD interface board, but with a dual buffer (ULN2003). This board could be used to control the two filters, but with less functionality compared with the SPI bus. I use it to control the “external” Penny connector
The small green and sqare pcb located close to the blue I2C interface is the Angelia OCXO (10 MHz). I’ll probably later add a specific ocxo for the Red Pitaya.
On the upper side of the rack, you can see 2 different switching power supplies. The big one is a 13,8V dedicated to the relay switching and the 20 W power amp (not on these pictures).The second -and smaller PSU gives a 8.5 volts
- for the 10 MHz ocxo (which has it’s own series regulators and filtering).
- For the Red Pitaya or Angelia 5V rail
- For the Angelia 3.3 rail
In fact, this PSU is able to deliver approximately 40 to 50 W. The 8.5 V is sent to a bunch of 3 series regulators (LM1084 for the 5, 3.3 and “whatever” output voltage) followed by a heavy-duty filtering board to avoid any switching noise coming from the PSU.
The last to pictures shows the CPU/display/USB audio board and partly the antenna switching system (not yet shielded when I took the pictures). The TX switching relay board can handle approximately 200W PEP.
As I already said, this UP board is just a proof of concept, for demo purposes and “quick and dirty” portable conditions. Most of the time, the computer part is shut downed and the rig uses like a “normal” networked radio station.
More enhancement are “under development” : an SSPA control unit (MCU, sensors, display), a new version of the Pure Signal coupler, a new design of the voltage regulation board… and I hope Manfred XQ6FOD will succeed building the EER (or ET) part of his amplifier. I’d definitely love to drive a “linear D class amp” :- )
If someone consider using my Github repository (https://github.com/F6ITU) , please download the most recent “branch”. The “master” trunk is considered as an ever evolving project and could not be considered as reliable.
I couldn’t find the time to clean the gerber files yet… it will be done during this summer… I swear
Once again, I wish to thank Pavel for its efforts and the implementation of Phil VK6PH “alex” protocol. Very few ham are considering using the SPI bus… they really don’t know what they miss
Vy 73’
De Marc f6itu

Hallo Red Pitaya Freunde

This Red Pitaya transceivers is using the SPI bus (Thanks to Pavel for this incredibly nice feature) and a personal retrofit of the Alexiares frontend, offering all the functions controlled by PowerSDR or LinHPSDR (filter section of the Setup menu)

This filter is a combination of low pass and high pass, able to cover the whole 0/60 MHz spectrum without hole (and not only small portions restricted to the ham band as most bpf do)
This frontend also includes an attenuator, a low noise amp for the 6m portion, 3 antenna selector, several external filters, transverter and Pure Signal switched input
External control (penny menu) is also available to command and control an external sspa and it’s lpf (via the I2C bus this time)

I'll probably made a write-up of the whole system and integration next July
Most of the frontend architecture is described at
https://wiki.electrolab.fr/Projets:Lab: ... es_Angelia#

Hope you'll enjoy it
73'
Marc f6itu

Hi Manfred
Never heard of "alpha & omega" before. And these FET are definitely better than FQP13xx. Thanks for the tip and the reference.
The 416 is hard to find in Europe, but TME is selling the AOT412 which is very close in term of characteristics
Seems nice for low cost RF amps
Thnks agn.
Marc

Hi Manfred
That's not frightening at all... it is strongly encouraging and tempting

just a stupid question : did you try to change the pair of IRF FET with a couple of FQP13N10 ? these are very cheap switching FET, same category of the IRF series, but have a lower input capacitance

anyhow, this looks really promizing . Thanks for sharing

Marc f6itu

Pavel marks a point (and I feel stupid indeed)
you can't control a master/slave I2C dialog without connecting both

Marc