For JT modes you know when it’s good - the over decodes!
It’s very rare to get a bad decode and then it’s obviously gobbledegook.
So nothing beyond the normal exchange is required, but cramming the ref in at the end wouldn’t hurt.
I favour JT9 for its reduced band occupancy, especially for the lower bands. Better for pile-ups! I have yet to try it from a summit, as my ebay spares-or-repair netbooks are too under-powered for JT. Waiting for work to get me something decent!
VOX use and just an audio cable is what I only use with both my Yaesu and KX3. No need to CAT/DATA at all.
My CQ SOTA call macro includes the reference, as does my basic exchange macro, on my DroidPSK app. Gives efficient operating (potential of one QSO per minute) when activating.
The JT modes have some heavy limitations on what can be sent at any time up to a maximum of 13 characters I think. There are rules about callsign format and addition data and this contraint of the allowable characters is used by the software to decode the data when it is further in the noise.
These modes are much less conversational than PSK etc. but are much more easily decoded when many dB below the noise.
psk is easy for that. jt will be a bit harder.
I have to spend a bit of time chasing psk me thinks
edit: easer to type than do 
I have just been playing with JT65 on my Samsung NP-N310 netbook which has an Intel Atom 1.6GHz processor. I found that the standard JT65HF program worked okay on receive when there were few stations to decode and transmit worked fine. However it failed to complete the decode at busy times. I downloaded the HB9HQX version and hey presto the problem is solved as it receives for a shorter period. Nice interface as well.
73, Gerald G4OIG
Ok, this backs my results too. I tested it last year. So ill chech the latest version again!
Unfortunately the local environment at the home QTH is too noisy to copy much by way of DX on JT65, though I did copy a KP4 yesterday evening. If I could reduce the bandwidth I might fare better. It might be worthwhile trying my 500Hz filter in the SSB narrow slot. Of course this is unlikely to be an issue out on a summit where it is usually a much quieter environment.
Yes, the world of CPU specs. Oh for the days when you could look at the model and speed of a CPU and know how much “ooomph” it had. It’s got to the point that the only way you know if it’s fast or not is when you run tests, the names are misleading now.
What’s good to know is that there are solutions for lower performance computers. I ran WSJT-x on a modern Windows tablet that uses a modern Atom 1.33GHz processor and the decode of the samples was almost instanaenous. Running the same WSJTx release on an 8 year old eeePC701 and decoding the samples was very slow. But the new Atom benchmarks 5.5 times faster. But the tablets have significantly naffer audio hardware. The eeePC701 has a quite a high spec audio chip, your Samsung will be the same.
Just waiting for a slow boat from China to deliver some USB audio cards to try…
my benchmark yesterday was 10watt (battery + KX3) and a 35ft wire in central London got me ~3500km somewhere in Russia, maybe a bit more. On 20m.
I did test this too. seems ok with the new versions of JT software. Battery will work for 3-4+ hours of work for sure. (my estimation! I did 30mins only)
Hi Andrew, I was thinking the summit ref may need to go in the free text field (followed by the 73 field), and probably lengthen the QSO to 8 transmissions - unless we put in the summit ref and 73 in the last over (keeping it at 6 transmissions).
The 13 character maximum transmission makes it harder, but we just have to work around it - personally I would be happy with summit ref and 73 as the final transmission of the activator - repeated if the chaser doesn’t give a 73 for it. Guess we can only try it and see how it goes, ammending as necessary.
The suggestion of moving below the traditional slot for JT modes to simplify it for SOTA chasers and reduce congestion is also a good idea - if not enough contacts are made, you can alwasy go back into the usual area to get enough to qualify.
Matt
VK1MA
agree! a preconfigured free text will be perfect.
GW SE001 73
As activation can be like a pileup, we might find a short cut to the exchanges! 
so lets see if shortcuts work:
I call CQ M6VAR IO91 (1)
a couple of guys respond
CALL LOCATION (2)
I replay CALL M6VAR -13 (3)
M6VAR CALL R-22 (4)
GW SE001 73 (5)
then more guys calling CALL2 LOCATION (6)
etc etc.
So can be 5, can also eliminate (6) if activator calls another one from the original decode.
Santa paid me an early visit and left a USB audio dongle that supports 48kHz sample rates. I’ll have a play over the Christmas break with this and a cheapo tablet and if something looks viable I’ll try it from a summit. This depends on it being viable with the tablet and the WX being suitable. The current WX not being suitable for much other than hiding out of the wind, rain, lightning & snow.
I agree Matt. With Christmas greetings being sent at the moment, I have decoded many transmissions with no callsign in them, just greetings and 73. The frequency indicated by the software provides the clue as to who is sending the transmission, indeed WSJT-X has a window specifically for information on the working frequency. A full summit ref + space + 73 works (such as AA0/XX-001 73).
73, Gerald G4OIG
I assume that you already know this, 44k does the business for JT, more its just unneeded CPU usage.
The new “JTDX ver 17.3.2” software for JT modes specifically says you should use 48 kHz sampling rate. This is because it uses up to 4 consecutive passes to decode signals in the 12 second interval after each reception period.
73,
Walt (G3NYY)
My Acorn Atom had a 1 MHz processor speed (6502 chip), and 2 kB of RAM.
73,
Walt (G3NYY)
Thank you! I didnt know that!!
(Im not sure if the Atom CPU will handle more than a very basic single pass)
I can’t vouch for other software but I have WSJT-x running here with V1.7 on Win7 on a laptop and V1.6 on a Win8 tablet. The laptop has a full feature sound card (multiple inputs/outs multiple sample rates). The Win 8 tablet uses the USB dongle that arrived yesterday. Using the USB input set to “1 channel 48000 samples” WSJT-x is working using the laptop’s speakers and mic and a headset plugged into the USB dongle. If change the rate on the USB dongle to “1 channel 44100 samples” the it does not decode. Maybe other software will interpolate to 48kHz sampling.
The £2.01 USB dongle certainly works but it’s not HiFi! There’s a nasty sprog at 1kHz exactly but apart from that it’s fine.
Next is to make an audio lead for the 817 and check on the VOX keying.