- Howdy! My name is Charles, I live in San Benito, TX, in the southern tip of the state. Palm trees, South Padre Island, Mexico, spring break, and so on. It sMessage 1 of 3 , May 10, 2003View SourceHowdy!
My name is Charles, I live in San Benito, TX, in the southern tip of
the state. Palm trees, South Padre Island, Mexico, spring break, and
so on. It's kind of hot and sticky down here, right now.
I've recently set up the MixW Q15x25 package as recommended, and
tuned around for sigs. So far I haven't found anything, probably
looking at the wrong time or place.
I'm also interested in the earlier version, Newqpsk, that is included
in Thomas Sailer's "soundmodem" driver for Flex32. I've noticed that
Newqpsk is not as advanced as Q15x25, but the easy gateway between
Packet and Newqpsk that Flex32 makes possible makes it very
Is anyone here familiar with the latest release of "soundmodem" and
Flex32? I'm wondering about that version of Newqpsk's viability on
HF, since it is an earlier version.
Also: I've noticed that the MixW version of Q15x25 is limited to a
PACLEN of 256. - But the Flex32 Newqpsk has no provision in its setup
screens for PACLEN at all. I have a multicast application in mind
that requires a PACLEN of 1000... Does anyone here know how the
Flex32 version of Newqpsk handles PACLEN?
Sorry to barrage you guys with all of these questions, but I am
interested in this mode, and have to learn somehow.
One thing that I am wondering about is how Q15x25 behaves for file
transfers, etc... I'm thinking that if it drops characters or gets
garbage characters, then the only way to use it as a file server
would be the multicast approach, as it would give the additional
layer of error correction needed for reliability.
I still haven't tested-driven the multicast software with Flex32, so
my interest in that may be premature. In theory, it should work just
fine, except for the PACLEN issue.
If worse comes to worse, the author of the multicast software could
tweak the PACLEN down to 265 for this application, but I'd rather not
ask for that kind of favor unless it is absolutely necessary.
Besides that, reducing the PACLEN of the multicasts will reduce the
system's efficiency considerably by increasing the packet overhead.
One great thing about multicast with the 1k PACLEN is the reduced
overhead. It's almost all data, just like it should be for that
I've done some testing on VHF, and by increasing the TXDELAY quite a
bit, I even managed to connect to a buddy over the local voice
repeater. ( I had permission ) For 1200bd packet at least, the
Packet/Newqpsk gateway seems to work perfectly. Users can connect to
the packet node, then digi through my Flex32 setup and come out on
HF, in Newqpsk mode... Now to see how the HF end of it behaves!
Any help on this stuff would be greatly appreciated. I'd also like to
QSO, get a chance to play with the MixW version a little!
Charles Brabham N5PVL
- ... They are both exactly the same. They are actually built from exactly the same source code. So the features and performance are the same also. ... TheMessage 2 of 3 , May 11, 2003View SourceOn Sun, 11 May 2003, Charles Brabham wrote:
> I'm also interested in the earlier version, Newqpsk, that is includedThey are both exactly the same. They are actually built from exactly the
> in Thomas Sailer's "soundmodem" driver for Flex32. I've noticed that
> Newqpsk is not as advanced as Q15x25, but the easy gateway between
> Packet and Newqpsk that Flex32 makes possible makes it very
same source code. So the features and performance are the same also.
> Also: I've noticed that the MixW version of Q15x25 is limited to aThe newqpsk/q15x25 modem should not limit the PACLEN in any way. The
> PACLEN of 256. - But the Flex32 Newqpsk has no provision in its setup
> screens for PACLEN at all. I have a multicast application in mind
> that requires a PACLEN of 1000... Does anyone here know how the
> Flex32 version of Newqpsk handles PACLEN?
possible limitations come from the AX.25 protocol stack used. In MixW, the
internal protocol stack seems to limit it to 256. For FlexNet32 I don't
know if there is a limitation.
Anyway the FlexNet32 protocol stack gives the user very limited control on
the channel parameters (like paclen, frack, txdelay etc...). It has
complex algorithms to calculate those parameters automatically based on
the channel usage pattern. I'm not sure if those algorithms work well on
HF or with q15x25 altogether...
Unless there is a way to tweak these parameters your self (I'm not a
FlexNet32 guru so I cannot say for sure) this may be a problem for those
wishing to use FlexNet32 for q15x25 on HF.
> One thing that I am wondering about is how Q15x25 behaves for fileIt's an AX.25 mode. AX.25 (in connected mode) will guarantee 100% error
> transfers, etc... I'm thinking that if it drops characters or gets
> garbage characters,
free connection (or nothing if the channel is too bad).
Tomi Manninen Internet: oh2bns@...
OH2BNS AX.25: oh2bns@...
KP20ME04 Amprnet: oh2bns@...
- Thanks, Tomi! I experimented with FlexNet and 300 baud HF packet five years or so ago. As you would expect, FlexNet was too aggressive, not backing off onMessage 3 of 3 , May 11, 2003View SourceThanks, Tomi!
I experimented with FlexNet and 300 baud HF packet five years or so
ago. As you would expect, FlexNet was too aggressive, not backing off
on MAXFRAME and PACLEN enough for reliable HF packet.
The FlexNet system would be great for 300 baud packet, but I think it
would involve rewriting the FlexNet kernal. I'm not sure you could do
it with just a driver.
I uploaded a jpeg of a VHF multicast network here, in the PICTURES
section.... I'd like to try something like that, but on a larger, HF
It seems to me that multicast and Newqpsk were made for each other,
addressing each other's drawbacks quite well. I'll see what I can do
with it, and keep the group posted.
The multicast software I have expects a KISS TNC, which means I'll
have to use FlexNet's KISS driver in order for it to talk to FlexNet,
which means that most of FlexNet's control over the link will be
eliminated. It is my hope that this will allow me to digi the
multicasts through Flex32 to Newqpsk without too much interference
from the Flex32 kernal.
The multicast software will have to be modified to include a digi
field in the multicast packet header. If Flex32 has a 256 PACLEN
limit, the multicast software will have to be modified for the
smaller PACLEN and lose some efficiency. I'll avoid that if at all
That's the theory I'm working on, anyway. It may be that I am overly
optimistic, but I suppose that is better than the alternative!
There is an article about VHF multicast in the "Library" section at
the web address below, and a tutorial on the RadioMirror multicast
software in the "Tutorials" section.
There's a new tutorial on Flex32 with the soundmodem in there too.