[aprssig] "Best" packet decoder solution

[Ray Wells]

What you describe is typical for TNC's that use the Exar XR2211 FSK 
decoder IC, which is totally unforgiving of the high tone being as 
little as 1dB or so lower in level than the low tone. I don't know what 
decoder the KPC uses but the Exar 2206/2211 pair was commonly found in 
TNC2 compatibles such as the MFJ 1270B.

The most workable solution to minimise this problem is to use pre 
de-emphasised audio for the TNC, i.e. audio direct from the 
detector/discriminator. This overcomes the problem of a transmitter that 
is not using pre-emphasis or, a transmitter that is using pre-emphasis 
but it's been compressed back to "flat" due to excessive audio input to 
the limiter/compressor. Using pre de-emphasised audio for the TNC 
appears to suffer no ill-effects handling audio from a transmitter with 
correct pre-emphasis.

If you've ever had the opportunity to compare many off-air signals on a 
Service Monitor, you'll begin to understand why some packets won't 
decode reliably. Over deviation, under deviation, distorted tones and, 
incorrect tone levels (relative to each other) are not uncommon 
observations. Almost anyone can set their transmitter to better than 
ballpark with just a second receiver (scanner), a CRO (software CRO is 
fine) and some understanding of "what goes on".

Take a look at this excellent article for a better understanding  ....   
http://www.febo.com/packet/layer-one/transmit.html

Ray vk2tv

On 28/12/10 03:44, Lynn W. Deffenbaugh (Mr) wrote:
> Note that the one that is missing there is the KPC-3+ which, according 
> to what I've read, fails miserably if the signal is flat on the air 
> (not pre-emphasized) but de-emphasized on the receiving end before 
> being fed to the KPC-3.  The KPC-3 apparently doesn't like the low 
> tone being louder than the high tone resulting in a lack of decoding.
>
> Some have recommended that the audio interface for the KPC-3 be set up 
> so the transmit audio (KPC to radio) goes through the pre-emphasis 
> circuit, but the received audio (radio to KPC) be pulled directly from 
> the discriminator and before the de-emphasis (meaning, don't pull it 
> from the speaker output).  This would mean that "correct" on-the-air 
> packets (pre-emphasized, with highs louder than lows) would be decoded 
> ok as would non-emphasized (both tones equal) packets.
>
> I have a KPC-3+ here, but I haven't yet hooked it up to anything for 
> comparison to the TNCs listed in my previous e-mail.  I haven't made 
> time for this because apparently the KPC-3+ also has trouble when 
> operated long term in KISS mode which is my preferred TNC interface.
>
> Lynn (D) - KJ4ERJ - Author of APRSISCE for Windows Mobile and Win32
>
> Greg Clark wrote:
>> Thanks, Lynn, exactly the info I was looking for!!!
>>
>> - Greg
>
>
> _______________________________________________
> aprssig mailing list
> aprssig at tapr.org
> https://www.tapr.org/cgi-bin/mailman/listinfo/aprssig
>