SSB audio transmit delay in Flex receiver

  • 1
  • Question
  • Updated 2 years ago
  • Answered
I own a 6500 SDR with latest software and standard FHM-1 mike. We have a local group of HAMS that enjoy SSB discussion on a 6 meter net in the evening. The others in the group, using analog radios, criticize me for releasing the PTT too quickly and dropping the last letter of my call sign (N1THL), when I sign. I never was called on this when using my previous radio which is an ICOM 751A. Recently I read this article about a net using Vox that had an increase in "doubling" with HAMS that were using SDRs , created by the receiver delays in the SDRS. See http://www.n1kdo.com/sdr-delay-measured/
This article shows a measured receive delay of 164 ms for the 6500 and extending out to 300 to 400 ms for other SDR radios. Analog radios were near zero. This is more of a curiosity question.. Does Flex have any measured or calculated data for the SSB delay on the transmit side...from mike to transmit out? Flex makes a great 6000 series radio.. I've owned the 6300 previously.
Photo of spopiela

spopiela

  • 71 Posts
  • 10 Reply Likes

Posted 2 years ago

  • 1
Photo of pa0bie

pa0bie

  • 152 Posts
  • 20 Reply Likes
SDR radios will always have this kind of delays.  Try to set the TX-delay in radio-setup to put on 200 mSec. Then You have a delay on the start of the transmission and not any more on the end. Perhaps that will work better for You.
Photo of spopiela

spopiela

  • 71 Posts
  • 10 Reply Likes
I've added 200 ms and will give that a try. Thanks for your help!
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 8670 Posts
  • 3251 Reply Likes
I do not recommend adding a fixed delay.  I recommend doing what WA4SCA recommended which is to just wait a moment after speaking before releasing the PTT switch.  On SSB mode, the delay is in the neighborhood of ~50-100 ms (1/10 of a second).  Once you have your "muscle memory" trained, it becomes second nature.
Photo of Alan WA4SCA

Alan WA4SCA

  • 22 Posts
  • 3 Reply Likes
It takes time to do all the calculations in an SDR radio, and some delay is inevitable.  Depending on the software, you can often adjust parameters such as sampling speeds and number of buffers to reduce, but not eliminate it.  Of course there are always performance trades.

Another place this matters is in digital modes such as packet.  In order to transmit packets, I had to use a parameter called TXtail, which holds the transmitter for a selected time after the PTT is released. The last few bits were being cut off with TXtail set to zero, which is normal for analog rigs. It can take 200-350 ms, depending on many things. On voice, I have just reprogrammed my brain to add the needed delay.
Photo of Ken - NM9P

Ken - NM9P, Elmer

  • 3823 Posts
  • 1164 Reply Likes
One advantage of the very slight latency is that when using VOX I never seem to experience the cutting-off of initial syllables. And with proper adjustment of the VOX delay, I don't get cut off at the end.


But when using PTT or the footswitch, I did need to adjust my timing just a little because I was sometimes transmitting "November Mike Nine Pa....." in a pileup. But remember that the delay is less than a fifth of a second. Less than a single letter. It only takes retraining yourself to hang on for a slight moment to end the cut-offs. In a couple weeks you won't think about it any more.


Ken- NM9P
(Edited)
Photo of Barry N1EU

Barry N1EU

  • 495 Posts
  • 120 Reply Likes
Has anybody actually measured the ssb tx latency and come up with a number?  50msec?  100msec?  200msec???

Barry N1EU
Photo of Tim - W4TME

Tim - W4TME, Customer Experience Manager

  • 8670 Posts
  • 3251 Reply Likes
Thanks for the measurement Doug.
Photo of Ken - NM9P

Ken - NM9P, Elmer

  • 3823 Posts
  • 1164 Reply Likes
Thanks. Would be interesting to know if it changes with DX+, and a narrower filter of about 2.4 kHz. My guess is that it wouldn't be much different.
(Edited)
Photo of Doug Hall

Doug Hall

  • 178 Posts
  • 49 Reply Likes
Ken,
I see no appreciable difference in the TX delay regardless of PROC setting or TX filter bandwidth. TX EQ doesn't seem to affect it either. It's always a bit under 130 mS.
73,
Doug K4DSP
Photo of Ken - NM9P

Ken - NM9P, Elmer

  • 3823 Posts
  • 1164 Reply Likes
Thanks, Doug.  
I didn't think it would, but it is nice to know.  I do know that on CW, the latency goes up as filters become increasingly narrow, but mostly when under 500 Hz.  
Some day I must get one of those nifty 'scopes!

Ken - NM9P
Photo of Doug Hall

Doug Hall

  • 178 Posts
  • 49 Reply Likes
The Tektronix 'scope has been a real workhorse. My wife gave it to me almost 9 years ago as an anniversary present!

I've heard that the Rigol 'scopes are a nice (and less expensive) alternative.
73,
Doug K4DSP
Photo of Bob - W7KWS -

Bob - W7KWS -

  • 255 Posts
  • 35 Reply Likes
Tim,

How about including a PTT release delay in the software equal the the audio delay?
Photo of Paul Bradbeer

Paul Bradbeer

  • 154 Posts
  • 60 Reply Likes
Bob, that is an excellent idea.  To my mind, for SSB Ops this is a high priority...I love my 6500 but I shouldn't have to 'train' myself to delay release of the PTT to compensate for the radio's latency!