Hello Frank, hello all,
> I do not understand the interest in special hardware type
> modems like the Baycom. Although I have used the Baycom
> (with the TI chip) and the Poor Man's Packet modem (using
> the Am7910), I have found it much more convenient, and less
> expensive (time and material) to use an ordinary sound
> card. Sound cards are cheap and don't require dedicating
> serial or parallel ports that are better used for other
> applications.
There are some good reasons to still use something like BayCom in our days.
The first reason is: There are still frequencies that one cannot operate more than
1200 bps on, due to bandwidth limitations, for example.
The second reason is that the SoundModem driver must be considered experimental, and
is sometimes not easy to set up. Even more, it seems that Thomas's support for that
driver has slightly vanished. He does not want to enhance tools like smdiag or
smmixer any more and even starts ranting if someone talks about their limitations.
The third reason is that exactly aligning the soundcard is sometimes very difficult
due to the lack of reliable diagnostic software (smdiag does not work on most newer
kernels, it is even missing in some of the newest ax25-* packages).
And, if you finally managed to get some soundcard to decode incoming packets, the
detection rate is often much worse than with a BayCom in hardware.
The fourth reason is that the (few) soundcard designs supposed to work with
SoundModem are not available any more. That is because ISA cards are generally
vanishing.
For PCI cards, there's no SoundModem support available.
Here, it sounds a little bit ironic that Thomas Sailer, developer of the SoundModem
driver, has written the Linux Kernel sound driver
(http://www.ife.ee.ethz.ch/~sailer/linux/pciaudio.html) for the Ensoniq AudioPCI
(http://www.ife.ee.ethz.ch/~sailer/linux/es1370.html) and the S3 SonicVibes PCI sound
card
(http://www.ife.ee.ethz.ch/~sailer/linux/sonicvibes.html)
but seems to be too busy to extend SoundModem.
The fifth reason is that if you want to avoid long PTT delays you'll have to use one
of the parallel or serial ports from the computer to drive PTT. Yes, there us also a
circuit that uses the MIDI port but a) it looks complicated, compared with the
circuits for the serial or parallel port or VOX, b) it depends on a correct MIDI port
setting (in other words, you'll have to make sure it is enabled and not working as a
joystick port accidentally).
The sixth reason is that one can easily have more than one modem of that type
connected to the computer. But how about two or even more soundcards? Besides the
problem of not having enough ISA slots on recent mainboards there is the problem of
confguring them correctly. PCI cards are, unfortunately, not supported yet.
> Now there is a new hardware design using a new chip that
> reportedly replaces the TI chip.
>
> Am I missing something here? Is there a significant
> advantage to using such an approach over using a soundcard?
I mentioned six reasons. Maybe, there are even more.
Cheers,
Gerd
--
Gerd Roethig
Universit�t Leipzig, Medizinische Klinik u. PK I
Johannisallee 32, 04103 Leipzig
Tel. (0341) 97 12622, Fax (0341) 97 12515
