Full duplex

331 views
Skip to first unread message

Ronald Meier

unread,
Jan 5, 2023, 4:05:10 AM1/5/23
to Hermes-Lite
On the wiki page under software / Full duplex it says " The Hermes-Lite 2.0 always runs in full-duplex, and software should show the received spectrum during transmit.". 

What does it exactly mean? Receive and transmit at the same time on different frequencies? Or is my understanding not correct?

73 Ronald

"Christoph v. Wüllen"

unread,
Jan 5, 2023, 5:13:14 AM1/5/23
to Ronald Meier, herme...@googlegroups.com
This simply means that receive does not stop when transmitting, and the data is
contained in the data stream from the HL2 to the radio.

Whether the SDR software makes use of the data or not is another question.

The RX frequency(ies) stays the same as it was before the radio went TX.

In "normal operation" it is NOT recommended to make use of the RX samples during TX
in the SDR program
unless one is in "split" or better cross-band mode, simply because if RX and TX are
on the same frequency, you will see your TX signal VERY strong in the RX from the
crosstalk in the T/R relay, and this signal will be so strong that it makes the
AGC go to the limit, which means that immediately after TX/RX transition your are
deaf for a short time, until AGC has re-adapted to the new level.
> --
> You received this message because you are subscribed to the Google Groups "Hermes-Lite" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to hermes-lite...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/45fa859d-038d-4d02-a00a-2edf01842035n%40googlegroups.com.

Ronald Meier

unread,
Jan 5, 2023, 5:26:07 AM1/5/23
to Hermes-Lite
Christoph, danke. Clear, than the text on wiki is not 100% correct as normally with full duplex something else is understood.

MfG, 73 Ronald

Op donderdag 5 januari 2023 om 11:13:14 UTC+1 schreef "Christoph v. Wüllen":

ron.ni...@gmail.com

unread,
Jan 5, 2023, 11:12:46 AM1/5/23
to Hermes-Lite
Some of the SDR software does take advantage of the HL2 transceiver being full duplex, by analyzing the received signal and performing adaptive pre-distortion the clean up the simultaneously transmitted signal.  

Also, I've experimented with transmitting some extremely weak CW signals from the HL2 (noise shaped fraction-of-1-bit IQ data), not loud enough to drown out all the other RF signals in the same band.

73, Ron, n6ywu

Steve Haynal

unread,
Jan 5, 2023, 2:32:19 PM1/5/23
to Hermes-Lite
Hi Ronald,

We've stuck with full duplex as defined originally by the openhpsdr RTL. RX and TX are always active in openhpsdr radios and the HL2. In other words, RX data is always sent even during TX. Enabling the full duplex bit just means that two separate internal oscillators are used so that RX can be a different frequency from TX. With the full duplex bit disabled, RX and TX are locked to the same internal oscillator.

73,

Steve
kf7o






On Thursday, January 5, 2023 at 2:26:07 AM UTC-8 rpmei...@gmail.com wrote:

Ronald Meier

unread,
Jan 5, 2023, 3:02:42 PM1/5/23
to Steve Haynal, Hermes-Lite
Thks Steve for the further clarification. I think I understand now. If I translate it for myself, HL2 has only one oscillator and therefore not able to full duplex in the "traditional way (like for example IC7610". But as HL2 is a SDR it is able to with software to have RX and TX always active. Correct?

73 Ronald



Op do 5 jan. 2023 om 20:32 schreef Steve Haynal <softerh...@gmail.com>:
You received this message because you are subscribed to a topic in the Google Groups "Hermes-Lite" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/hermes-lite/lqanCSByqYU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to hermes-lite...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hermes-lite/5f75a346-ce26-4c14-8c05-5297ccf421f4n%40googlegroups.com.

Steve Haynal

unread,
Jan 5, 2023, 4:18:53 PM1/5/23
to Hermes-Lite
No, the HL2 has 5 NCOs (numerically controlled oscillators) as the default gateware can have 4 receivers plus 1 transmit. Duplex off just forces the RX and TX to share a single NCO. When duplex is off, TX and one receiver must be the same frequency. Duplex on assigns independent NCOs to RX and TX. When duplex is on, TX and all receivers may be set to independent frequencies. Whether duplex is on or off, all active receiver data is always sent to the host computer during TX.

73,

Steve
kf7o

"Christoph v. Wüllen"

unread,
Jan 6, 2023, 4:35:45 AM1/6/23
to ron.ni...@gmail.com, herme...@googlegroups.com


> Am 05.01.2023 um 17:12 schrieb ron.ni...@gmail.com <ron.ni...@gmail.com>:
>
> Some of the SDR software does take advantage of the HL2 transceiver being full duplex, by analyzing the received signal and performing adaptive pre-distortion the clean up the simultaneously transmitted signal.
>
> Also, I've experimented with transmitting some extremely weak CW signals from the HL2 (noise shaped fraction-of-1-bit IQ data), not loud enough to drown out all the other RF signals in the same band.
>

This is a software problem. There are two options to transmit CW with an SDR:

a) create the IQ data of a true CW signal in the SDR software and send it to the radio.
piHPSDR does this (I have programmed that), but most programs have option b)

b) encode key up/down info in the lowest bit of the IQ data and
*tell the radio we are in the so-called CWX mode*
Then, the firmware of the radio will "translate" the one-bit
info into a nicely shaped CW signal


So I guess what you describe is the consequence of sending a 1-bit CW signal
and not setting the flag which tells the radio to decode it.

DL1YCF.

"Christoph v. Wüllen"

unread,
Jan 13, 2023, 3:50:14 AM1/13/23
to Steve Haynal, Ronald Meier, herme...@googlegroups.com


> Am 05.01.2023 um 22:18 schrieb Steve Haynal <softerh...@gmail.com>:
>
> No, the HL2 has 5 NCOs (numerically controlled oscillators) as the default gateware can have 4 receivers plus 1 transmit. Duplex off just forces the RX and TX to share a single NCO. When duplex is off, TX and one receiver must be the same frequency. Duplex on assigns independent NCOs to RX and TX. When duplex is on, TX and all receivers may be set to independent frequencies. Whether duplex is on or off, all active receiver data is always sent to the host computer during TX.
>

So the "duplex bit" for the FPGA (ADDR=0x00 bit=2)
and the "duplex mode" of the SDR program are two very different things.

SDR PROGRAMMERS: FORCE FPGA DUPLEX BIT "ON" BY DEFAULT!

The reason is, that common (often-used)
operating modes such as Split, CTUN, XIT lead to different TX and RX frequencies so they require duplex "on".
For example, in CTUN,
the NCO RX frequency stays at the center of the display, while the TX frequency follows the offset
which changes when clicking somewhere in the display. The same applies to XIT which lets the TX frequency
move away from the RX one.

If this really saves something in the FPGA, the SDR software could set the duplex bit according to the needs,
so if if neither Split, CTUN, XIT is used, one can tie the TX and RX frequencies together.
I always regarded
this as an unnecessary source of possible problems so I force duplex "on" in pihpsdr. I modified the comment
in old_protocol.c as follows:

//
// ALWAYS set the duplex bit "on". This bit indicates to the
// FPGA that the TX frequency can be different from the RX
// frequency, which is the case with Split, XIT, CTUN
//
output_buffer[C4]=0x04;

Yours, Christoph DL1YCF.



James Ahlstrom

unread,
Jan 13, 2023, 4:57:20 PM1/13/23
to Hermes-Lite
 Yes, Christoph is correct. Quisk always sets the FPGA duplex bit to True. If the Rx and Tx frequencies need to be the same, the SDR software should just make them the same.

Jim
N2ADR 

Alan Hopper

unread,
Jan 14, 2023, 2:20:11 AM1/14/23
to Hermes-Lite
Spark also always uses duplex, with multiple receivers you might want to transmit on any of them so it would get complex if the tx frequency were tied to one of them.
73 Alan M0NNB

Ronald Meier

unread,
Jan 14, 2023, 8:11:07 AM1/14/23
to Hermes-Lite
Thks all for the further clarification. 

73 PH7R Ronald

Op zaterdag 14 januari 2023 om 08:20:11 UTC+1 schreef ahop...@googlemail.com:
Reply all
Reply to author
Forward
0 new messages