[RTTY] Best RTTY program/hardware

[Kok Chen]

I have not done RTTY for as long as some people here, but here are my  
personal observations.

Long, long ago, I had used an MFJ 1278 for SWL'ing.  I shouldn't even  
mention having owned one... there goes my reputation.  Chalk it up as  
stupidity of youth.

I have used a few stock KAM Plus, and one which I modified so that the  
switched capacitor filters are closer to that needed for RTTY.  I  
still own a beige KAM Plus and a black KAM plus.

The modified one is a significant improvement, but still not that  
great.  The problem with many stock TNCs is that their filters are  
designed for Pactor and are way too wide for weak signal RTTY use.   
After modification, the filters in my modified KAM Plus were too  
narrow for Pactor; that is why I have always kept more than one KAM  
around.

A PK-900 (long ago given away) came in somewhere between the modified  
KAM Plus and an unmodified KAM Plus.

Garry NI6T had also modified the PK-232 to match RTTY (it is actually  
much harder to modify than the KAM Plus, since the PK-232 didn't use  
switched capacitor filters, which only involved resistor changes) and  
the modified PK-232s were not just used at 3D2CU but other DXpeditions  
had borrowed them.

If you choose the hardware route, get a HAL ST-8000.  There is nothing  
better in the hardware world, period, punkt.

If you can't get hold of a used HAL ST-8000 (not the ST-8000A, they  
are very different animals), I have found the Timewave 599zx to be  
quite solid.  Both the 599zx and the ST-8000 have AFSK regenerators so  
you can use them as front ends to other TNC/TU that you already own.   
They both also have analog AGC, and have decent dynamic range as far  
as hardware modems go.

I have also used the HAL DSP-4100 and the HAL DXP-38 side by side with  
the KAMs and they are both better than the KAM as long as you adjust  
the audio level properly, but neither can match the ST-8000.  The  
problem with the DSP-4100 and the DXP-38 is that they only have 14-bit  
ADCs without an analog AGC stage -- so their dynamic range is  
limited.  You have to carefully adjust the audio level so you don't  
clip the codec on strong signals, and when you do that, it won't work  
well on very weak signals.  You pretty much have to wire a pot ahead  
of them and ride the pot manually.

I have never used an ST-6000, but many people on this reflector have.   
Lets say I have never heard a bad word about them.  Plus, like the  
ST-8000, there is a nice crossed ellipse display built into the 6000  
(not as nice as the one on the ST-8000, but better than the tuning  
mechanisms that you find on the rest of the field).  For weak signal  
DX'ing it is imperative that you are able to tune a signal within 10  
or 20 Hz under poor SNR conditions.  It does not matter how good the  
demodulator is if you cannot *accurately* and *quickly* tune an RTTY  
signal in.  Except for clicking on a waterfall, the crossed ellipse is  
one of the few indicators that can do *both* well if you are tuning  
with a VFO knob.

However, any modern software modem that is worth its salt should beat  
even the ST-8000.

As finely crafted as they are, the ST-8000 is still limited by what  
you can do in hardware.   Its countermeasure to multipath causing mark/ 
space overlap is quite rudimentary, for example.  But at least the  
ST-8000 has some, albeit simple, circuitry to handle overlapped mark/ 
space.  The most modems that are built as TNCs don't even have a way  
to switch in a different demodulator under multipathed conditions.  If  
you want to dig out a fluttery signal across the pole, you need good  
multipath performance.

After saying all that, the best way to convince yourself is to do your  
own tests.  If you run Windows, get a copy of AE4JY's PathSim and if  
you run Mac OS X, get cocoaPath.  Both are free.  Run a clean RTTY  
signal through different propagation and SNR conditions through these  
HF Channel Simulators and make recordings of them to tape or sound  
files.  cocoaPath has built in RTTY, PSK and CW generators, so you  
don't even have to first create a clean AFSK signal yourself.

Then play the recorded audio back into each of the modem that you are  
testing.  All modems will get precisely the identical inputs.  The use  
of HF Channel Simulators is how the industry test demodulators (the  
prevailing standard is the ITU-R F.1487).  Then, make your own plots  
for the modems you are testing, like what Alex VE3NEA has done here

http://www.dxatlas.com/RttyCompare/

73
Chen, W7AY