As I understand it (probably wrong then!) to generate SSB it needs to be able to change both the frequency being generated and the amplitude of the RF at a rate greater than Nyquist limit (ie more than twice) for the required 3KHz(ish) bandwidth of speech.
As both of these are under software control in the QMX, then it's theoretically possible if the incoming voice can be sampled, analysed, the required calculations to generate the values of freq. and amplitude performed and the freq & amplitude controls applied fast enough (as well as doing all the other things that the software does to make the radio work).
That's about my limit of understanding!
?
I have just ordered a kit and it looks like an interesting candidate to
experiment with, software wise.
I thought the same, I saw the RSGB youtube video & was inspired to get one.
Having built the QMX+ I was really impressed with the performance and was thinking it'd be a shame to break it while playing with it, also Hans has indicated that the microcontroller on the QMX is locked such that it's not possible to use other than QRPLabs firmware.
Unless you replace the micro with an unlocked one, but then you can't subsequently program it with QRPLabs code as it's encrypted.
This is to allow QRPLabs to exist in the face of cloning from the far-east, something that I totally agree with, QRPLabs are very fair with pricing so there isn't really any issue.
Also the QMX is quite involved in all the control features that the firmware needs to drive the hardware without destroying it, so not an easy task to get something running that will 'just work'.
My solution was to keep the QMX+ for 'best' and keep the lid screwed on and buy a QCX+ as well.
I've just built the QCX+ this week and other than the single band & the CW only rx path, it's just as good in terms of performance with the added bonus of a re-programmable Atmega device, more through hole components so easier to tap-into & modify, lots of test-point/jumpers to do the same. My intention is have a play with this as a basic hardware platform.
I didn't fit the PA transistors to the QCX+ to ensure that I don't blow them or transmit where I shouldn't for now & I deliberately missed off the rx bandpass filter and after testing & aligning the analogue Rx section I've bypassed the low pass filter so I can play with the Rx on all bands.
I hope to start by keeping the current device operation & adding a separate micro with an audio codec to tap into the I & Q signals to have a play with some DSP before replacing the Atmega code with something that does what I want to try, either by reprogramming the existing device or removing it & using my DSP micro to also do the control signals.
I have subsequently discovered that the uSDX radio started life as a modification to the QCX so while they've moved on to their own hardware, their open source firmware has some similarities so will act as a good source of information.
?
I'm definitely a QRPLabs convert - there's still enough 'self learning' while also offering good performance at a very reasonable price.
My only wish is that I knew morse - and that's partly why I bought the QMX+, to help & motivate me into doing this.
?
Don't forget RTFM - the instructions are comprehensive but I still managed to miss a wire link that connects the BNC and the comments about needing a good soldering iron and taking longer to solder some parts due to the heatsink effect of the multi-layer PCB is definitely correct.
