> From: Chuck Hast <[EMAIL PROTECTED]>
> To: [EMAIL PROTECTED]
> Subject: Re: 9600 packet query
> Date: Sunday, January 30, 2000 9:02 PM
>
> Jose
> Que gusto de nuevo...
>
> I have been doing a bit of 9k6 stuff both on the amateur said and on
> the ham side. So far experience has shown that depending on the radio
> (more RX than TX) deviation can be anywhere from 2.5 to 3khz, with
> 2.7 being about the best compromise.
>
> Here in Tampa we have some links between our FPAC switches which
> use Motorola Radius radios, and these radios do not like anything much
> beyond 2.5khz with 2.7 being about as high as I will go to talk to such
> a unit.
>
> We also have TEKK radios, these units can go the full 4-5 with no problem
> but since we also have other radios we keep it down to 3 khz except those
> links where we have a Radius on one end, in which case we throttle back
> to 2.7.
>
> As always it is the RX filters that cause the grief, (got pretty good at
swapping
> filters in the TEKK radios, setting them up for 19k2. The TX will go much
> higher, seems I tried something back years ago when I worked at PacComm
> where I pulled the filters out and ran 38k4 thru two radios but they were
as
> broad as barn doors.
>
> Really the higher the speed the easier it is to modulate the radio, it is
on the
> RX end where the problems appear, the higher speeds do not have such low
> freq components as do the lower speeds, i.e. 9k6 has a 23hz component,
4k8
> 19k2 is about 46 or there abouts, getting that low freq part through some
TX
> particularly those that are synthesized can be a test of patience,
>
> I have even taken wireless mic TX and made short haul high speed (well
> after 1k2 and 9k6 b/s 64 and 128 ARE high speed B-b) links out of these
> devices...
>
>
> Chuck Hast e-mail
[EMAIL PROTECTED]
>
> ---- FPAC ----
> www.fpac.qth.net
Hi Jos� Chuck and everybody,
It is interesting this thread.
I had the same case that Chuck with the motorola Radius
but I use another simpler radio that the Motorola Radius.
I modified two Radios Kenwood TR-7950 to work it
with 9k6 TNC, two KPC9612+.
I changed it two diodes of the discriminator for two new
diodes of germaniums 1N60A or 1N34A both are excellent
I also changed the ceramic filter (CFW-455F) for (CFW-455B)
for the output of the discriminator (RXA).
I used a no-polarized filter (NP) of .47uf 25v
in the (TXA) input, I changed an resistance of 220k for 68k in direct
modulation
I didn't use the (XCD) since the KPC9612+, I put it in CD SOFT non needed
the
SQ. Then I adjusted the IF applying to the receiver 12.5khz of deviation
with a tone of 1,000hz signwave using a (IFR-1200A) Monitor Server.
After of this calibration to the IF, I entered it sign at a normal level
of 4.5khz and I verified the sensibility and this was very good.
The final part was to adjust the TX and I proved different positiones nivel
I adjusted the deviation from the TX to 3.5khz, then the level of
modulation of
the TNC I adjusted to 2.8khz this was the trick so that it worked perfect.
then I making some connection of test I put the TXDELAY for 10ms FAST
Switch
It is works very fast, this is working until moment on the air.
The Radio that better it responded me was a old model of AZDEN the
model is 4000, I didn't have to make it many modifications.
And it works much better than the Kenwood TR-7950.
73's Jose (hi8gn)
[EMAIL PROTECTED]
http://hi8gn.dynip.com/
