Hopefully Hans can see this. We know QDX is designed as a transceiver driven by a computer to work. So it means you always have a computer available along with QDX. Now if you feel like CWing as well, you need another QRP transceiver (QCX*, KX2/3, IC-705 etc) which means one more device to carry and switching antenna back and forth while switching between FT8 and CW.

I'm wondering does it make sense to add the same key-shaping circuit in QCX to QDX so it can support transmitting CW as well? That would essentially turn QDX into a QRP version of FlexRadio (although CW/Digital only and operated locally). With a single USB cable carrying I/Q, tone and CAT, the computer becomes a panadapter and SDR. Since QDX has a high performance ADC, I think DSP software in computer would make it a high performance CW receiver. On TX side, we can either use logger software's built-in keyer or paddle/key with WinKeyer to control QDX via CAT/PTT.

I currently have a KX2 with I/Q output mod. To achieve the equivalent of the FT8 + CW + Panadapter setup, I would need a HiFi USB sound card (larger than QDX), another USB cable for CAT, and 2 more cables carrying the analog I/Q signal and tone signal. While with QDX, only one USB cable is needed. So I kinda feel this QDX+CW is a neat solution. Any comments?

Zhenxing Han/KN6PZO

BTW, from another hobby (flight simulator), I learned the idea of using a MIDI controller as a control panel. The best of such things is Behringer X-Touch Mini which has a ton of buttons and encoders. I just can't stop imagining using it as an SDR control panel. Actually FlexRadio's iPad version supports MIDI controller as well.

Zhenxing Han,
?I like your thinking. I have no idea how to do that or if it's possible, but it sounds like a great idea.?
--
73, Dan? NM3A

I suspect it would be fairly easy to remote into someone else's QDX via TeamViewer.? I and K7ZB have done it for WSJT-X.? Worst case the audio for CW could be handled with something like Discord.

73,
Dave?? AB7E

Hi,
has anyone pursued this further so far? IMHO, one should be able to use the basic QCX key envelope circuit but with a stronger transistor, because the push-pull PA draws a higher current than the PA in the QCX. 1A minimum, better be on the safe side with a 1.5-2A transistor.?

I am working on this, so if anyone has a suggestion for a suitable transistor type that would be great!


73 de Martin, DK3IT

PS: One could also use a bias voltage controlled via PWM or DAC to provide the key-shaping in software (same as in the uSDX); the downside is that this rules out high-level languages like CircuitPython/Micropython for the control software, as they can't match the speed.?

Addendum: The MPS751 can handle 2A as per the datasheet (), so the main question is how we can drive it to that current with the desired timing of 5-10 mS.

Looking at the schematic, the TX line switches both the drive to the final and the reverse bias to the diode switches in the LPF.? If separate lines from the uC were used for these two functions, then is should be possible to key the drive to the final for CW rather than turning the Si5351 clock outputs on/off.? With a bit of timing between the LPF switching and the keying, fast breakin should be possible.? Would a key shaping circuit (series transistor to drive power to finals) be needed if the drive is keyed (finals are class D, which like class C is biased beyond cutoff with no signal)?
Iambic keyer, morse to text, and sidetone could be added to the QDX firmware for CW, or be done on the PC side.

And another thing, could the same tricks that were used to transmit SSB on the QCX (uSDX) be used on the QDX?? Some hw mods would be needed of course.

I loosely follow this thread. One basic question: How are you going to HEAR the CW?

On Sat, May 7, 2022 at 09:56 AM, John N0TA wrote:

I loosely follow this thread. One basic question: How are you going to HEAR the CW?

?It requires a computer or iPhone/iPad to either implement an I/Q decoder or use the decoded audio output from QDX, then implement a digital narrow band CW filter.

开云体育

The Rx subsystem is a standard SSB receiver, so you just feed the “USB Audio” from the radio into the computer speakers. ?Implementation details are different for each OS, but it’s pretty straightforward.

The ATS series by Steve KD1JV starting with ATS3 uses a clever CW shaping circuit producing a trapezoid envelope with timing independent from the power supply voltage.



I actually adjusted the slope of the shaping circuit to produce quite a reasonable PSK31 shaping in 2008. Based on my simulations such envelope is comparable to a slightly overdriven SSB transmitter.

https://kufr.cz/~ok1iak/temp/psk31-trapezoidal-envelope.pdf

One would need to extend QDX with Q6, Q4, D4, R13, R14, C37, R24, R26 (on the ATS-3b schematic, see below).



Q6 would be inserted between 12V and L14 (on the QDX schematic). Actually this circuit may save some PA FETs because the PA will be off for sure during power up sequence before the controller boots up and takes control of its output pins, namely the TX line, because the Q4 (from the ATS-3b schematic) is grounded.

For the shaping to work, the Q4 (from the ATS-3b schematic) would have to be keyed on together with the TX to the FETs, however the Q4 would have to be keyed off a bit earlier than TX line to keep the RF being generated during the trapezoidal ramp down interval. Thus most likely such mod is not possible without Hans's firmware support. If one does not have access to firmware source code, one would have to resort to some form of ugly hackery as adding an external keyer / sequencer chip and convince QDX to produce CW continuously while keying the PA with the external keyer / sequencer chip, ideally having the external keyer / sequencer chip connected to PC with another USB cable. I am not really convinced I would want to do that. Thus please Hans, would you add a PA / CW shaping control pin to your firmware? Actually maybe once Hans implements support for the PA keying that is populated on revision 3 boards, maybe the PA keying pin will be timed in such a manner that it will be possible to use it to shape CW envelope.

On PC side, the PC application could either produce morse tone and make use of the QDX VOX function (I did not try QDX VOX yet, I am not sure about the timing) or one could command QDX over serial line to key up / down. Actually VOX may be better because audio output is synchronous while serial commands are asynchronous which may produce morse with faulty timing. In 2007 I wrote? an application PocketDigi for Pocket PC as a companion to ATS-3b to control DDS of the ATS-3b to transmit various digital modes. The concept was similar to QDX with the exception that a DDS was commanded with commands synchronously transmitted over sound card output and decoded by the ATS-3b internal comparator. My approach required a special terminal application (PocketDigi) which was a showstopper to many, while QDX decodes audio tone frequency thus it works with any terminal application.

Regarding SOTA and such, I wrote a prototype application for Android to stream IQ of a Peaberry SDR connected with an USB cable to another PC running HDSDR on Windows. I consider it to be a proof of concept for CW SDR using the Peaberry SDR or QDX on Android phone.

73, Vojtech OK1IAK

I found a way to listen to what the QDX hears. Now CW can be decoded by the wetworks between my ears. Keying is something that I have not figured out. I have a Morserino, and I was thinking of porting its audio (ICW) to the QDX. I have not figured that out.

Linus via groups.io <colour@...> wrote:

I found a way to listen to what the QDX hears. Now CW can be decoded by
the wetworks between my ears. Keying is something that I have not
figured out. I have a Morserino, and I was thinking of porting its audio
(ICW) to the QDX. I have not figured that out.

Keying is one thing (I guess it can be done with CAT commands), but QDX
lacks envelope shaping, so every key-on and key-off will generate clicks
(interference).

OK. Is that still the case if well formed audio is sent into a qdx, or will it also have a clicks??

Linus via groups.io <colour@...> wrote:

OK. Is that still the case if well formed audio is sent into a qdx, or
will it also have a clicks?

Yes.

QDX is a very clever hack -- it's not a SSB transmitter, where you feed
the well formed audio and the output signal closely resembles it, but it
analyzes the frequency of the input audio signal and produces a single
sine wave, with constant amplitude, based on this frequency.

For example:

- QDX is tuned to 7040 kHz (7040000 Hz)
- audio signal from the computer is 300 Hz
- QDX analyzes the signal and notes its frequency
- QDX adds 7040000 Hz + 300 Hz (working in upper side-band mode) and
determines that it has to produce a tone with frequency of 70400300 Hz
- QDX programs the built-in synthesizer to produce this exact frequency

This is repeated many times per second.

The amplitude of the input signal doesn't correspond to the amplitude of
the output signal.

Hi Hans! I hope all is well with QRPLabs!

I just had a nutty idea I thought I'd share with you. Please ignore this if you think it's silly or a waste of time or if you've already thought of it...

The B+ Voltage used for diode switching in the QDX should be proportional to the peak RF voltage at the output stage. Perhaps adding a divider and monitoring it (if you can find another analog input) can provide some kind of early warning about dangerous VSWR or other final amplifier problems?

Just one more thing you might want to add to list of things to "mull over"

stay well...

Mario AE0GL