¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. QRPLabs
  2. Messages
Single Messages Topics Expanded Polls

Re: #gdx QDX Firmware Request: Phase Modulation (psk31, etc.) #gdx

 

Hello Ed

The attached picture shows a PSK 31 signal. A 1 is represented by a 0-degree phase shift and a 0 is represented by a 180-degree phase shift. Each symbol generation has a raised cosine shape. The phase change occurs at the zero amplitude point. This fact ensures that the bandwidth occupation of the mode is only 31Hz.

QDX could create the 180-degree phase changes. However it cannot produce the raised?cosine envelope shape because QDX does not have any way to control the RF envelope amplitude. It's all or nothing. If you did phase shifts at intervals of 1/31 seconds the result would be a much wider signal than the desired 31Hz. This is why people are telling you, correctly, that this cannot be done on a QDX.?

On the other hand - there is another interesting possibility. This article is of interest:?


It notes that a string of 1's is a single unmodulated carrier. A string of 0's has the appearance viewed in the frequency domain of TWO carriers at?+/- 15.5Hz from where the unmodulated carrier was (half the symbol rate). In the article the author discusses generating the necessary signals using two oscillators which are set either to the same frequency, or at?+/- 15.5Hz (above and below it). he states that as long as the frequency separation and timing is perfect, there is no need for envelope shaping.?

Why this is relevant for the QDX is that the QDX in fact does contain two oscillators (in the Si5351A). And each of these two oscillators is connected to its own pair of BS170s which amplify with the transistors operating in Class-D (arguably). Ordinarily, the two sides are set at the same frequency and at 180-degree phase amplitude; the output transformer adds up the two parts (well arguably subtracts), putting them together to produce the desired output power.?

So then what if you set the two oscillators to 7039984.5 and 7040015.5 Hz. Wouldn't you get two equal amplitude distinct tones at?+/- 15.5Hz (31Hz separation)? If any other unwanted frequencies would be generated (other than RF harmonics) then what would be the mechanism for generating those??

I suspect there is something wrong with the idea... but can anyone explain what exactly?

73 Hans G0UPL



On Thu, Nov 17, 2022 at 4:31 PM Ed G8BQR via <ed_go_2004=[email protected]> wrote:
Many thanks for the comments Robin & Allison,
G3PLX assumes audio frequency generation of the phase shift.
By generating the FT8 type frequency shift directly at RF it seems that the QDX is able to remove harmonics by using the RF low pass filters, rather than an audio frequency low pass filter and the linear following stages of SSB generation.
This is a brilliant system, one reason I got a QDX, for the idea alone, not because I particularly wanted to FT8 on 40m? :-)
(still interesting to try it out though)
Currently the QDX does not have the software ability to create a direct RF phase shift modulation, so psk31 is not possible, as Allison says.
However the Si5351a can be programmed to create a phase shift directly at RF.
In this case, perhaps the harmonics might also be removed by the RF low-pass filters, in the same way as for FT8.
("amplitude management" was definitely not the right way of expressing myself !)
Yes, i came up with this request due to seeing the post about FLDIGI, not sure what Mitch is doing, but I joined that group anyway :-)
Regards,? Ed? G8BQR
Join [email protected] to automatically receive all group messages.