[pskmail] Re: [pskmail] Re: Notes and questions related to last weekend's tests
- From: Rein Couperus <rein@xxxxxxxxxxxx>
- To: pskmail@xxxxxxxxxxxxx
- Date: Tue, 29 Jan 2008 09:53:49 +0100
Roger, the <SOH> only work well with PSK, as they spread the information over
the whole bandwidth.
We have been using a preamble in the past, but it showed that with the higher
speeds the syncing was not the problem.
The sync problem only exists with DominoEx. I can reintroduce a preamble with
the DominoEx modes if that is wanted..
73,
Rein Pa0R
> -----Ursprüngliche Nachricht-----
> Von: pskmail@xxxxxxxxxxxxx
> Gesendet: 29.01.08 01:44:09
> An: pskmail@xxxxxxxxxxxxx
> Betreff: [pskmail] Re: Notes and questions related to last weekend's tests
>
> Hi.
> In the latest versions of flarq Dave has added 10 X <soh> in a row and
> this seems to help a lot with syncing.
> I have been helping with crossed audio testing on a pair of pc's
> DominoEx was not one of the modes I tested with as it was mainly the psk
> versions and MFSK16 we were interested in.
> I will run some tests using DominoEX and report how it does with the 10
> <soh> leading string setup on flarq / fldigi.
> 73 Roger G6CKR
>
> On Tue, 2008-01-29 at 11:22 +1100, John Douyere wrote:
> > Hello, I did some more experiments with Pskmail last weekend while
> > camping and here are my notes and questions:
> >
> > 1. PSK63 vs 125 vs 250: I found that again the faster speeds more than
> > compensate for the loss of sentivity in getting the packets across
> > intact. Maybe in low noise, low signal would PSK63 be better, but I
> > haven't come across this condition yet in practice.
> >
> > 2. DominoEx: I had another go at this mode with little success. I am
> > insisting with this mode because I still think that it should be good
> > for QRN (static crashes) conditions. Looking more closely at the
> > issue, this mode takes far longer than the PSK modes to synchonize.
> > Typically my client's fldigi would only get the synchronization right
> > at the end of a short packet (in DominoEx11 mode). Solution may be to
> > improve the sync speed in fldigi or alternatively add a sync string at
> > the beggining like the old "rrrrrr" just for this mode. Rein, any
> > suggestions? I am happy to change and test the code on the Pskmail
> > side.
> >
> > 3. Some timing questions: when the client sends a packet and the
> > server does not receive it properly, the timeout seems very long
> > (roughly 10 to 15 seconds). Since it takes no more than 1 or 2 seconds
> > for the server to start transmitting a frame would it be ok if the
> > client waited say 4 seconds for an <SOH> character and assumed no
> > reply? I have tried to change the client's timeout but it does not
> > seem to alter this issue. Rein, how is this timeout calculated and
> > where can I find the logic in the code? Thanks.
> >
> > 4. I have updated my Mandriva distribution to the latest versions of
> > Pskmail client and server. All functions work but when I do a Getpos
> > from the client I get a "position unavailable" as a reply. I checked
> > on findu and all the positions are updated correctly both in unproto
> > and connected mode. What am I doing wrong? Where should I look for
> > more detailed feedback on the server?
> >
> > Otherwise had good success with this system. The CW filter I added
> > last week in my server does help with the city QRM. I managed to
> > linkup to the server on 40M with only the FT-817 bare at 5 watts with
> > a dipole in the trees at 4 meters above ground. Right up to PSK250
> > speed. QSB would affect 1 in 3 or 4 packets but it would get the job
> > done at the end. Great stuff.
> >
> > Regards,
> >
> > John (VK2ETA)
> >
> >
> >
> >
>
>
>
--
http://pa0r.blogspirit.com
Other related posts:
- » [pskmail] Re: [pskmail] Re: Notes and questions related to last weekend's tests
