Teensy SSB with EER

DF5SF

New member
Hello,
EER is an acronym for Envelope Elimination and Restoration. You amplify and limit the SSB output signal to a square wave with the frequency information.
Amplify that square wave with a nonlinear amplifier outside Teensy (class C or E) and restore the envelope information by modulating the power supply of that amplifier.
I am a newbie and my question is whether it is possible in Teensy to compute the envelope information (as PWM signal) and the frequency.
The frequency range is between 7 ... 15 MHz, better from 3 ...30 MHz.

Many thanks
Uli
.
 
Hi,
I know this QSX-SSB. I have it working here , but the 10bit AD-converter are not so good as 24bit AD.
I will try to "porting" this software into Teensy with better filter and waterfall.

Greetings
Uli
 
Hi Uli,

very nice project, it would be great to have the Teensy generate accurate signals for phase and amlitude in order to drive a Class E polar transmitter!

Generally, I think it is feasible with the Teensy (T3.6, better use the Teensy 4.1).

There are a few things to think about (my suggestions):

* as far as I understood the concept of polar transmitters, the most important thing is the very accurate timing of the amplitude and phase information as the input to the Class E PA. It seems that even commercially offered Class E polar transmitters have/had a lot of problems in this respect (you have probably already scanned the whole lot of very interesting links to papers on this subject in the µSDX groups.io forum ;-))

* "Porting" the µSDX receiver software to the Teensy would not be worth the effort, I think. There are already several very good approaches to an SDR Rx using the Teensy (search the forum for DerekR, Convolution SDR, bmillier, chipaudette, Bob Larkin, there are also some others). So, my opinion would be: take one of the existing approaches and expand the software with your Class E [EER] approach !

* I would advertise to you to use some potentially interesting code parts from my own development (Teensy Convolution SDR) which I would consider most interesting for your Ham transceiver: 1.) fully automatic real-time software I & Q phase and amplitude correction with about >60dB opposite sideband rejection, 2.) Fast Convolution filtering with on-the-fly calculation of the filter coefficients (enabling freely chooseable filter bandwidth and shape), and 3.) frequency domain noise reduction specifically designed for speech. Feel free to use the code parts under the open source license (github DD4WH)

* the first decision you would have to make is whether you take an Audio library approach using the audio library (maybe using DerekRs excellent lib as a start or chipaudette lib for floating point calculation) OR use a queue to get the audio samples and do everything connected with DSP inside the loop (this is the approach in the Convolution SDR)

* There was also some German guy who is also active in the µSDX forum who has already extended the Convolution SDR code with CW Tx. So maybe you search for that source code on the µSDX forum

* My time is rather limited, but I am really happy to help you with DSP code or with suggestions!

73 de Frank DD4WH
 
Its possible to sum two PM/FM signals to get SSB (or any QAM modulation)

The average phase of the two phase modulators is selected to match the wanted instantaneous
phase, the cosine of half the phase difference sets the instantaneous amplitude.

If there was a way to arithmetically sum the two signals efficiently at high power, it would solve
the problem of avoiding linear amplifiers :)

Unfortunately I think this inevitably means a non-constant load impedance to the power stages,
which is problematic at most RF frequencies.
 
Back
Top