CW delay with Paddles

[K5DN]

What is the delay from the paddle input to the time the rig keys at the other end?  Is it one character plus latency?

Bob
K5DN

[sm2o]

Hi

Yes every part (dit,dah etc.) of the character is sent direct over internet. So minium delay is the Internet latency.
But that works over internal LAN, Over Internet you should set a little Key delay  ( keyer settings) also then the radio side will buffer x ms before starting transmitting, and then you avoid corupted characters caused by Internet Jitter.

73 de mike

[sm0mdg]

But that works over internal LAN, Over Internet you should set a little Key delay  ( keyer settings) also then the radio side will buffer x ms before starting transmitting, and then you avoid corupted characters caused by Internet Jitter.

How much delay and what jitter buffer and delay do you recommend for a internet connection with 35ms ping time?

I have the default jitter settings on the radio side and tried 300 ms Key delay, but still get messed up CW.

73 de Björn,
SM0MDG
SE0X

[sm2o]

Normally with a good internet connection and 35 ms ping i should guess 30 ms would be enough.
If it still messed up with "key delay" = 250 ms it means that some of the UDP packets are delayed more than 250 ms over the Internet. That not much to do about I'm afraid. I think it's only possible to set 250 ms, if you really have set 300 I guess you have mixed up "key delay" and "lf-delay".

73 de mike

[sm0mdg]

Hi Mike,

It is indeed possible to set "key delay" to 300, although the info text says it ranges up to 250, guess I was too tired to take in instructions yesterday

I have been playing around with different settings and nothing really helps. I loose dots or dashes, or on some occasions get very long ones. I use Spectrum Lab to monitor the station's sidetone at the other end so I can check visually after changing settings.

Could this be the effect of connecting though WLAN on the radio side? I am using a UBNT Nano Station to connect the RRC-Radio to the DLink wireless router. Could this have an impact?

Any settings I could try in the WLAN link that might help?

73 de Björn

[sm0mdg]

I did some further investigation this weekend by connecting the RRC-remote box to the local network. Ping time is locally around 4 ms and CW is much better. Although a lot less frequent I still have the same problem with lost dots or dashes and sometimes they get very long. I guess there's always a risk this will happen, and if it happens with the frequency of my local connection its fine.

What can I try to improve the code when keying over the internet? Is there anyone out there with good success who can share his setup and settings?

73 de Björn,
SM0MDG

[HB9AZT]

Hi Björn

Did you ever check rfi-issues at your radio-rrc? Had also some cw issues, ferrite ring in the key-line solved the problems with keying. For a simple test reduce the transmit power to a minimum and check the keying.

73 de Mark, HB9AZT

[sm0mdg]

Thanks for your feedback Mark!

Yes, I did rule out RF feedback as I was actually running the tests without RF at all, just monitoring side tone using Sprectrum Lab on the remote side.

I am leaning towards something in my routers, possibly because I use a WLAN hop on the remote side.

I am also noticing the the first dot or dash is cut a little short in the first character, don't know the reason for this but it might be related to other things

73 de Björn
SM0MDG
SE0X

[sm5acu]

During five weeks in September-Oktober 2010 I ran my RRC-1258 version 1, modified for CW, using software version 2.15 over a connection with 100 Mbit/s at the Control side and ADSL 14/0,8 Mbit/s in the Radio side. Ping times were around 35 ms. Radio ICOM IC-706MkIIG.

I used the following Keyer settings:
LF delay ms: 50
Key delay ms: 0

I experienced no problems with the keying, and did not receive any such reports either.

I had not made any serious tuning of these parameters.

/Uno, SM5ACU

PS. I tuned the values for the remaining parameters for a Mobile HSDPA in the Control end during 2009, using an older software release. Now in September-October I received very good reports on my SSB modulation compared to others' running RRC-1258. Especially, I had no drops in the speech. The values I used were:

Advanced settings:

Control:
RX jitter buffer size: 16
RX jitter delay: 8
Audio packet size (ms): 10

Radio:
RX jitter buffer size: 16
RX jitter delay: 12
Audio packet size (ms): 10


DS