Title: Testers wanted: Low Latency TX/RX Switching (CW operators Delight!)
Post by: DB4PLE on 14. November 2018, 07:28:51
Hi,
we just integrated an experimental change, which makes switching from RX to TX and vice versa much more responsive. It should help CW operators working with higher WPM to get the first dit or dah out with little delay and without shortened length. Both issues are existing with the "old" approach of TX RX switching. For users of voice modes the difference will most likely not be noticable, for these uses the switching was already fast enough.
The effect will be most noticable on TRX with an STM32F4 processor (mcHF, Eagle) and here again on machines using SPI for display controller.
Technically the switching takes as long as it was before, but we now check 1500 times a second if we should switch (AND if it is safe to switch, which is not always the case) and if so, switching takes the usual very few milliseonds it used to take.
Nothing comes for free, and the drawback here is, that we (the developers) now have to be more careful that we don't attempt to switch while our normal code does something in the "same area" since switching now interrupts the normal program flow. Previously switching happened as part of the main control, where everything was done one after the other, so there was no risk of interfering. We tried to conservative, so it should work, but of course, we may be wrong.
Please test this change. Especially the signal waveform of the first morse signal is of interest and of course, if there are any noticable issues when switching from TX to RX or vice versa.
73
Danilo
|
Title: Re:Testers wanted: Low Latency TX/RX Switching (CW operators Delight!)
Post by: SP3OSJ on 14. November 2018, 11:46:22
Hi
Eagle has TX/RX relay.
TRX Eagle does not have a PIN diode.
73
Artur |
Title: Re:Testers wanted: Low Latency TX/RX Switching (CW operators Delight!)
Post by: DF8OE on 14. November 2018, 11:49:16
Hi Artur,
that does not matter. Delay was located in "waitstates" of UHSDR when paddle was pressed in an uncomfortable monent (for the software)... Hopefully now much better because of this delay is minimized now.
vy 73
Andreas |
Title: Re:Testers wanted: Low Latency TX/RX Switching (CW operators Delight!)
Post by: DB4PLE on 14. November 2018, 12:46:34
Hi Artur,
To be precise, the actual switching process has not been changed (i.e. it runs at the same speed as before). It is only the time between pressing the PTT button / paddle and the UHSDR TRX start of running the switching procedure which has been significantly reduced on average. As a consquence if you are not actively changing settings or are for instance saving the configuration to EEPROM while you press paddles, TX will activated almost instantenously in CW. Which also means, if the switching worked before without problems, it should do the same now.
To confirm this: I have a relay in my test machine's RF board and it seemed to be fast enough even for the very fast switching on the H7. It is the "normal" variant of the small OMRON latching relays used in the mcHF LPF section.
73
Danilo
|
Title: Re:Testers wanted: Low Latency TX/RX Switching (CW operators Delight!)
Post by: DF8OE on 14. November 2018, 19:50:27
I have tested using straight key. Much better than before. I do not have the skill to do high speed CW QSOs but I am able to key "cq dx de df8oe" and "vvvv" quite fast ::) I cannot observe any reasonable cutting of the first dot and signal sounds good.
vy 73
Andreas |
Diskussions- und Newsboard des DARC-Ortsverbandes I40 | Powered by YaBB SE
© 2001-2003, YaBB SE Dev Team. All Rights Reserved.
|