[RTTY] who understands rtty code, strange thing happening, same characters always dropped

[Kok Chen]

On Jan 7, 2011, at 1/7    8:07 AM, Joe Subich, W4TV wrote:

> All of these characters either start or stop with a space (I has
> seen them listed both ways in code tables and can't track down a
> definitive "Bit order" - Chen?).

Serial protocol transmission all begin with the least significant bits  
of the code, right after the start bit.

Mark is the "rest" condition (i.e., stop bit sends Mark and Start bit  
sends Space).

A Baudot "1" bit is sent as a Mark and a Baudot "0" is sent as a Space.

(For a long time now, the ARRL Handbook has this Figure that shows how  
a Baudot "D" is sent.  I found it so useful over the years, that I  
retained the original figure [don't fix it if it ain't broke] when I  
wrote the "Unstructured Digital Modes" sectoin of Chapter 16 of the  
2010 and 2011 Handbooks.)

So what the characters Jim listed have in common is a Space just  
before the stop bit (a Mark).

Both Carriage Return and Line Feed has the same property (Baudot has  
no "Enter" character and uses a CR/LF pair).

However, the thing that puzzled me when I saw Jim's list is that the  
same thing should happen to the letters "S" and "N".

Whether it is start or stop bit related, I will still check to make  
sure that there is a proper pull up resistor (what Joe mentioned).   
But, I tend not to think it is pull-up related since that would cause  
other symptoms to appear (characters that have two 11 bits in a row  
before a 0 bit).

However, I would definitely check the Baudot baud rate that the  
software sets the digital interface to.

Jim mentioned the microKeyer, so I assume he is going through the  
microKeyer's serial protocol, and not "bit-banging" the FSK pin of the  
rig -- that would imply that the Baudot baud rate is determined by the  
microKeyer.

The MicroHAM keyers use 45 baud, not 45.45 baud.  But this should not  
be a huge problem (45 vs 45.45 baud should only reduce the SNR of the  
matched filter at the copy end by a fraction of a dB -- measurable,  
but not a big deal under real propagation conditions).

But a 50 baud setting, for example, will place the stop bit where the  
last bit should be at a 45.45 baud receiver.

So Jim, assuming that the S and N characters are outliers, I would  
check to make sure you are not setting the microKeyer to use 50 baud  
or something like that (leftover from working P5/4L4FN?!).

With the microKeyer, use 45 or 46.  Because of the way the microKeyer  
computes baud rates internally, 46 baud actually has a little bit less  
baud rate error than 45.

If that does not work, you can also try AFSK (whose baud rate is  
completely under the control of software).  If the microKeyer is  
connected also to the Packet connector of the Yaesu, you should be  
good to go.  Just make sure to adjust the Mic Gain of the rig so that  
no ALC shows.  With my FT-1000MP, I turn the RF Power knob to the  
maximum and control the power using the audio drive (Mic Gain knob),  
which I adjust for the power that I need, but keeping the ALC bars  
off.  I don't know if the 920 behaves the same way but give it a shot.

73
Chen, W7AY