[pskmail] Re: Comparing digimodes and ssb voice...what is good/bad result?

  • From: John Douyere <vk2eta@xxxxxxxxx>
  • To: pskmail@xxxxxxxxxxxxx
  • Date: Thu, 16 May 2013 10:22:02 +1000

Hello Tarmo,

No problems. Yes what you describe is worth testing (squelch open and let
Fldigi extract the FSK modulation).

Some (cheap but also not so cheap) FM radios can have a fair bit of
non-linearity in the audio passband.

Therefore it may be worth while trying two or three audio frequencies. For
example 1000, 1500 and 2000Hz. You may find different results.

All the best.

73, John

On Wed, May 15, 2013 at 6:43 PM, Tarmo Huttunen <tarmopaha@xxxxxxxxx> wrote:

> Thanks John!
> You gave me a lot of helpfull information. And also one totally new point
> of wiew to consider.
> If I understand correctly you say that fsk modes like thor can be quite
> effective also in FM-modulation? So I suppose proper use then is to leave
> base receiver squelch open and flidigi decoding fm-noise? Very very
> interesting! I must try that.
> To have local communication normally in FM phone but if needed switched to
> digimodes is very nice idea. Like easy local net but possibility for
> extented range when needed.
> Well, considering more, there must be totally different backround noise on
> 10m than vhf or uhf. But surely worth to try.
> Btw, I have tried thor-modes with cheap pmr-radios, results were not good
> even squelch left open, but these were only some sort of "hey lets try what
> happens" -style sunday afternoon games.
> Back to my original questions... I now have confirmation that thor22
> should work better on my system. I try next to catch some triggered fast
> waterfall pictures with spectrumlab to see and show what is going on when
> transmiitting.
> Maany thanks!
> Tarmo
> On May 15, 2013 2:48 AM, "John Douyere" <vk2eta@xxxxxxxxx> wrote:
>> Hello Tarmo,
>> Sorry I noticed I didn't reply to your last question directly.
>> I would expect THOR22 under FM to be better than SSB voice.
>> The only things that can affect THOR decoding are settings on the
>> receiving Fldigi (notably selecting "Slow CPU" and not selecting the
>> filters under the THOR mode settings) and a rather large  difference in
>> audio card clock frequencies (I suspect something like 5000ppm or more).
>> Have you done the card calibration using the WWV mode in Fldigi. It is
>> quite simple.
>> Also when using direct audio cables or speaker to microphone (no radio),
>> do you get a difference in audio frequency when you do a tune. That would
>> indicate a mis-aligment of audio clock frequencies between TX and RX
>> systems. I easily get 10Hz difference with an 1500Hz audio tune for
>> uncalibrated cards. That's 6700PPM of difference.
>> Hope this helps,
>> 73, John
>> On Wed, May 15, 2013 at 9:33 AM, John Douyere <vk2eta@xxxxxxxxx> wrote:
>>> Hello Tarmo,
>>> I did similar test than you a few years back. I tested on VHF some data
>>> modes transmission both in SSB and in FM from the same locations. I
>>> progressively increased the distance until I had no reply under any mode /
>>> modulation from the Pskmail server. Using remote desktop control I was
>>> checking the decoding on the Pskmail server side as well, mainly for
>>> frequency alignment (explanation below). Here is what I noticed:
>>> 1. The FSK modes like Thor and MFSK will work almost the same under SSB
>>> than FM modulation, right down into the noise, to the point that it is very
>>> difficult to say if there is modulation or not being transmitted
>>> (especially with THOR8).
>>> So I deduce from my tests that digital modes were going through when not
>>> voice contact could be made at all, both under SSB and FM modulation.
>>> On the other hand the PSK modes were much less sensitive under FM than
>>> under SSB.
>>> 2. At VHF and I suspect even more so at UHF frequencies, the VFO
>>> accuracy is poor for digital modes. I was getting 100Hz+ of VFO
>>> mis-alignment with an FT-817 and an FT-857 in my tests. Therefore the BIG
>>> advantage of FM modulation is that there is negligible audio frequency
>>> shift even with a significant VFO shift, and therefore sensitive modes like
>>> MFSK16 or THOR8 can be used quite effectively.
>>> 3. There was an article in QST some years ago from Skip Teller about
>>> digital modes under FM modulation and he had the same conclusion. In fact
>>> his article triggered my tests.
>>> So, assuming we get the same or close sensitivity with FSK modes under
>>> SSB and FM, then you should get a sensitivity advantage using digital modes
>>> over voice regardless of the underlying modulation (SSB/FM) since an SSB
>>> voice communication requires a minimum of about +6dB of s/n to be
>>> understandable while digital modes like THOR22 can be decoded at about -8
>>> to -10 dB of s/n (using a white noise test).
>>> From memory THOR8 was readable down to -13dBs in my bench tests so it
>>> should be usable further than voice regardless of the modulation.
>>> So if distance/sensitivity is the target I would definitely stay with FM
>>> modulation and use the sensitive FSK modes. Based on what I saw you
>>> "should" get a better result than under SSB voice.
>>> Let us know what you find. Hope that helps. All the best.
>>> Regards,
>>> John
>>> On Wed, May 15, 2013 at 4:54 AM, Tarmo Huttunen <tarmopaha@xxxxxxxxx>wrote:
>>>> Hello all!
>>>> I need a little advice from more experienced pskmailers.
>>>> Last night I send some aprs beacons with albrecht2990
>>>> handheld/AndPskmail to my own base with thor22.
>>>> Distance was about 20km. Beacons were received mostly ok, about 80%
>>>> right. In that distance/path voice communication is easy, it is in about
>>>> the border of practical fm work, still usable (I have tried that place
>>>> before)  and phone ssb is easy and clear.
>>>> So queston is: what range should I expect from digimodes like thor22?.
>>>> Should thor22 have "longer" or "shorter" maximum distance on line-of-sight
>>>> communicatons as phone ssb?
>>>> I know there can not be accurate answer for this question. What I am
>>>> looking for is for example experiences like " I was able to work ssb when
>>>> psk500 didnt get throw...but thor22 did..." or something like that.
>>>> If we consider pskmail digimodes starting from most robust (thor8?) to
>>>> less robust (psk500?) so where we should consider limit for phone ssb
>>>> usually to be? Is it somewhere on thor-modes or somewhere on psk-modes or
>>>> how?
>>>> Without arq, is this thor22-80% right bad or good result on path where
>>>> phone ssb is mostly easy?
>>>> Tarmo

Other related posts: