[pskmail] Re: The hardware modem idea (sri, long rant)

  • From: Per Crusefalk <per@xxxxxxxxxxxx>
  • To: pskmail@xxxxxxxxxxxxx
  • Date: Tue, 01 May 2012 12:44:40 +0200

Hi Demetre,

Thanks for the input, this is of course an interesting subject.

If we were going to develop a synchronous protocol with very tight
timing requirements, like pactor, then the real time requirements would
be correct. But, we are not doing that, asynchronous ARQ modes don't
have any problem with reasonable speed switching time. 

Also, I think some of the modes being used now are very robust. A few
weeks ago I was away skiing and I brought my FT-817 and a fishing rod.
The antenna I made was too short and I forgot my radials at home (only
had one real radial). At times I could hear the servers anyway and I was
able to send my position and a few short emails using only 3 watts and
Thor modes. I can't imagine how bad my signal must have been but it
worked.

Btw my portable soundcard interface is a small home made thing that is
attached to the cables between the radio and the pc (no extra box
there).

Regarding the QRM and finding a clear frequency I am really glad that we
stay within 500 Hz all the time. Other systems can sometimes decide to
increase from 500 Hz to 1600 or 2400 Hz in the middle of traffic
exchange and that has caused some QRM. If we decide to use wider modes
later I hope we can agree to try to avoid that issue. I like to stay
narrow (filtered) when QRP portable so I will support strict 500 Hz
channels. The hidden transmitter problem is helped in part with the
european servers talking to each other through the big ear server. When
my server transmits on 10148 it notifies the others and they wait, even
if they can't hear the actual transmission.
 
So, no, I still don't think we need a hardware modem.

73 de Per, sm0rwo


tis 2012-05-01 klockan 10:28 +0300 skrev Demetre SV1UY:
> On Tue, May 1, 2012 at 5:53 AM, John Douyere <vk2eta@xxxxxxxxx> wrote:
> 
> [snip]
> 
> >
> > But still, I have a genuine question I would really like to have an answer
> > for, and believe me absolutely exempt from ulterior motives: for those of us
> > who have used/are using Pactor modems and Winlink why would we be interested
> > in Pskmail? There must be some reason?
> >
> > All the best,
> >
> > 73, John
> 
> As I said in the past, PSKmail has the right general idea about an HF
> Radio e-mail system. It offers far more features than WINLINK2000
> (most important is the ability to send radio e-mail directly, not via
> a MAIN server controlled by one body), plus APRS BEACONS in realtime,
> TWITTER, KEYBOARD CHAT, MAPS, WEBPAGES etc.
> 
> Rein's ingenuity is unlimited here but, and here is the big but, it
> lacks decent modem support, not a speedy one necessarily, but it lacks
> a ROBUST HF MODEM.
> 
> This does not have to be an SCS MODEM necessarily, although since the
> P4dragon appeared in the market last June, there are many PTC-II
> series modems available on the cheap now.
> 
> I hope you John, will be able to improve on this need because your
> "andPSKmail" is a great idea. Not sure if the FLDIGI is a good modem
> to follow BTW. FLDIGI is FB for what it was designed to do but not for
> ARQ. I think you need to make the ARQ mechanism be in the MODEM (or
> virtual modem) not in the PSKmail application itself. The modem should
> preferably be in a separate small box with it's own CPU that serves
> only the ARQ MODEM. You have to use a Soundcard interface even if you
> use the soundcard of your PC so a separate small modem box will not
> make the system bulkier!
> 
> All of today's Operating Systems suffer from one thing! LATENCY! So
> you will never be able to build a ROBUST ARQ MODEM if it has to run in
> the same Computer (PC or Android or other phone). You will always get
> FREEZES and/or DELAYS because the computer has to attend to more
> things than serving the MODEM.
> 
> The Computer no matter which O/S it uses, has to do interrupt calls
> for everything, i.e. keyboard, display, mouse, disk operations,
> networking, calculations, decode/encode the MODEM signals just to name
> a few!
> 
> No matter how fast a CPU is or will become in the future, you will
> always have this LATENCY problem, because of the many devices that are
> attached to it. That is why it is important to have a smaller CPU that
> does only one thing. DECODE/ENCODE the MODEM signals.
> 
> You might be able to make a MODE such as PSK31, RTTY, PACKET RADIO,
> SSTV and other asynchronous modes work FB in an environment like this
> (i.e. using the Soundcard of a WINDOWS or LINUX or Android Computer as
> a MODEM) but these modes are not ROBUST and also they will never be
> fast enough. The WINMOR guys have done something along these lines,
> but even WINMOR is far from being ROBUST. It fails horribly at the
> presence of the slightest QRM and our HF Bands are full of QRM!
> 
> You just can't expect to find a clear frequency to conduct your radio
> e-mail business  today. Even if all the Radio Amateurs listened before
> they transmitted, there is always the HIDDEN TRANSMITTER SYNDROME that
> will create QRM. This will only get worst in the near future when the
> Solar Activity will increase.
> 
> I hope something good will emerge from this conversation.
> 
> Take care.
> 



Other related posts: