Re: [digitalradio] Move away from PSK63 /31 in NBEMS?
Hi all, I would like to test one of these ARQ modes. Anyone around for a test on 40- or 80m? I am calling cq on 3585 usb in Thor right now. 73 de LA5VNA Steinar kh6ty wrote: >> 1. Is it correct that just because a mode is listed in the NBEMS mode >> list, does not mean that you could not use it with flarq as an ARQ mode? >> > > Yes, but some modes cannot be used as an ARQ mode due to either too much > latency or ascii or control code support. The only modes tested with flarq > are DominoEx, MFSK16,32,64, PSK31,63,125,250, Thor (all speeds) and > MT-63-2000. The latency of MT-63 is so great, only the fastest speed is > enough to enable ARQ to work. > > >> Therefore, any of the modes could theoretically be used, even if not >> recommended or even if they might not work very well but flarq will ARQ >> any of them? >> > > See above. Of the available modes for ARQ, the ones listed for NBEMS are the > ones recommended for messaging. However, on VHF when signals are near the > noise, it is sometimes easier to tune with PSK31 (because the idle carriers > stand out more clearly against the noise background on the waterfall than > the multi-tone modes) and then switch to MFSK16 or DominoEx11 without > changing the tuning. > > >> 2. Could you explain how the MFSK modes are made more robust and yet can >> still work with other programs? >> > > We worked with ZL1BPU, one of the authors of MFSK16, to implement several > enhancements, one of which is called "puncturing", which averages the signal > level and if a signal sample is significantly above the average, it is > assumed to be a static burst and the AFC is momentarily disabled. The fldigi > MFSK modes are compatible with MFSK16 or MFSK8 in other programs, but it is > necessary to use fldigi in order to get the advantage of the enchancements. > > >> On another note, thanks for confirming what I suspected about the drift >> issue. I recently was told that I my preference for purchasing only TCXO >> rigs was overkill, even for VHF, but when you look at the ppm issues >> with modes such as MFSK, even a few Hz is a problem. >> > > Generally, not on HF so much, as transceivers tend to have less drift on HF, > especially on 80m and 40m, but definitely on VHF, where excessive drift can > often be a serious problem for some digital modes. > > 73, Skip KH6TY > NBEMS Development Team > > > > No virus found in this incoming message. > Checked by AVG - http://www.avg.com > Version: 8.0.138 / Virus Database: 270.6.5/1620 - Release Date: 19.08.2008 > 06:04 > > >
Re: [digitalradio] Move away from PSK63 /31 in NBEMS?
> 1. Is it correct that just because a mode is listed in the NBEMS mode > list, does not mean that you could not use it with flarq as an ARQ mode? Yes, but some modes cannot be used as an ARQ mode due to either too much latency or ascii or control code support. The only modes tested with flarq are DominoEx, MFSK16,32,64, PSK31,63,125,250, Thor (all speeds) and MT-63-2000. The latency of MT-63 is so great, only the fastest speed is enough to enable ARQ to work. > > Therefore, any of the modes could theoretically be used, even if not > recommended or even if they might not work very well but flarq will ARQ > any of them? See above. Of the available modes for ARQ, the ones listed for NBEMS are the ones recommended for messaging. However, on VHF when signals are near the noise, it is sometimes easier to tune with PSK31 (because the idle carriers stand out more clearly against the noise background on the waterfall than the multi-tone modes) and then switch to MFSK16 or DominoEx11 without changing the tuning. > > 2. Could you explain how the MFSK modes are made more robust and yet can > still work with other programs? We worked with ZL1BPU, one of the authors of MFSK16, to implement several enhancements, one of which is called "puncturing", which averages the signal level and if a signal sample is significantly above the average, it is assumed to be a static burst and the AFC is momentarily disabled. The fldigi MFSK modes are compatible with MFSK16 or MFSK8 in other programs, but it is necessary to use fldigi in order to get the advantage of the enchancements. > > On another note, thanks for confirming what I suspected about the drift > issue. I recently was told that I my preference for purchasing only TCXO > rigs was overkill, even for VHF, but when you look at the ppm issues > with modes such as MFSK, even a few Hz is a problem. Generally, not on HF so much, as transceivers tend to have less drift on HF, especially on 80m and 40m, but definitely on VHF, where excessive drift can often be a serious problem for some digital modes. 73, Skip KH6TY NBEMS Development Team
Re: [digitalradio] Move away from PSK63 /31 in NBEMS?
Just to clear up a few things: 1. Is it correct that just because a mode is listed in the NBEMS mode list, does not mean that you could not use it with flarq as an ARQ mode? Therefore, any of the modes could theoretically be used, even if not recommended or even if they might not work very well but flarq will ARQ any of them? 2. Could you explain how the MFSK modes are made more robust and yet can still work with other programs? On another note, thanks for confirming what I suspected about the drift issue. I recently was told that I my preference for purchasing only TCXO rigs was overkill, even for VHF, but when you look at the ppm issues with modes such as MFSK, even a few Hz is a problem. 73, Rick, KV9U kh6ty wrote: > PSK31 and PSK63 are still there for use in QSO's but neither mode is as good > as MFSK16 or MFSK32, or our new Thor modes, for ARQ messaging when high > static conditions on 80m or 40m exist, which is most of the time in the > summer or before and after a hurricane, so MFSK16 or MFSK32, wider, but the > same speeds as PSK31 and PSK63, are recommended for NBEMS messaging over > PSK31 and PSK63. The MFSK16 and MFSK32 modes are not your grandmother's MFSK > modes, but have been made more robust under static conditions on HF when > PSK63 and PSK31 get pulled off frequency by the AFC if there is a static > crash between ARQ transmissions. It was a tough decision not to include > PSK63 in the recommended NBEMS mode list, but the job of NBEMS is to > accomplish the most reliable messaging for any given speed, and the modified > MFSK modes are simply more reliable, although they are wider. > > For VHF messaging, PSK31 and PSK63 lose out again over DominoEx11 and > DominoEx22 at the same speed, because many VHF multimode transceivers have > no TCXO and drift too much to maintain tuning on 2 meters during an ARQ > transfer, whereas DominoEx has a wide tolerance to mistuning or drift, such > that no AFC is needed and provides a lower minimum S/N than the PSK modes at > the same speed. > > PSK31 and PSK63 are easier to recongnize among the background noise on VHF, > so it is often best to establish tuning with PSK41 or PSK63, check to see if > there is no adjacent frequency activity, and then switch to DominoEx for the > ARQ transfer. > > The popular IC-746Pro, for example, drifts 100 Hz between receiving and > start of transmitting, unless a TCXO is added, causing the loss of the first > few characters or words on PSK63, so I have switched our 2 meter digital net > from PSK63 to DominoEx. > > 73, Skip KH6TY > NBEMS Development Team > > >
Re: [digitalradio] Move away from PSK63 /31 in NBEMS?
>I note the new release of FL-Digi does not list PSK31 or 63 under > NBEMS modes. I realize that Dave , Skip, and others never really > intended NBEMS to have much use on HF , hence PSK250/125, but was > wondering why the option of slower PSK modes has been removed. Andy, PSK31 and PSK63 are still there for use in QSO's but neither mode is as good as MFSK16 or MFSK32, or our new Thor modes, for ARQ messaging when high static conditions on 80m or 40m exist, which is most of the time in the summer or before and after a hurricane, so MFSK16 or MFSK32, wider, but the same speeds as PSK31 and PSK63, are recommended for NBEMS messaging over PSK31 and PSK63. The MFSK16 and MFSK32 modes are not your grandmother's MFSK modes, but have been made more robust under static conditions on HF when PSK63 and PSK31 get pulled off frequency by the AFC if there is a static crash between ARQ transmissions. It was a tough decision not to include PSK63 in the recommended NBEMS mode list, but the job of NBEMS is to accomplish the most reliable messaging for any given speed, and the modified MFSK modes are simply more reliable, although they are wider. For VHF messaging, PSK31 and PSK63 lose out again over DominoEx11 and DominoEx22 at the same speed, because many VHF multimode transceivers have no TCXO and drift too much to maintain tuning on 2 meters during an ARQ transfer, whereas DominoEx has a wide tolerance to mistuning or drift, such that no AFC is needed and provides a lower minimum S/N than the PSK modes at the same speed. PSK31 and PSK63 are easier to recongnize among the background noise on VHF, so it is often best to establish tuning with PSK41 or PSK63, check to see if there is no adjacent frequency activity, and then switch to DominoEx for the ARQ transfer. The popular IC-746Pro, for example, drifts 100 Hz between receiving and start of transmitting, unless a TCXO is added, causing the loss of the first few characters or words on PSK63, so I have switched our 2 meter digital net from PSK63 to DominoEx. 73, Skip KH6TY NBEMS Development Team
Re: [digitalradio] Move away from PSK63 /31 in NBEMS?
> On Mon, 18 Aug 2008 21:04:22 -0400, "Andrew O'Brien" <[EMAIL PROTECTED]> > said: > I note the new release of FL-Digi does not list PSK31 or 63 under > NBEMS modes. I realize that Dave , Skip, and others never really > intended NBEMS to have much use on HF , hence PSK250/125, but was > wondering why the option of slower PSK modes has been removed. There is nothing special about the "NBEMS modes" submenu. It lists modes that NBEMS users are likely to want to use, but nothing that is not also available elsewhere in the main "Op Mode" menu -- it's just a convenience. I suppose you could use any of the modes that can handle the ARQ protocol, including the slower PSK modes. 73, Stelios, M0GLD. --
