Other SOTA sites: SOTAwatch | SOTA Home | Database | Video | Photos | Shop | Mapping | FAQs | Facebook | Contact SOTA

Datamode - FT8



Well I’m happy with …r2 and not sure about the changes. I think Joe is wise to take time to let the mode bed in.

Build it myself? Crikey, that’s what I pay Joe and his collaborators to do for me. Life’s too short to be writing software. Also, it needs to be standardised as much as possible. Otherwise some will introduce 10 second tx periods and others will want 24 character messages. It was bad enough with JT56A,B,C.

Just set a new record: 29 dB difference between my report and the other guy’s report. Even giving a 9 dB QSB allowance that implies 7.5kW at the other end. Even if the 20 dB scale is in error by 3 dB that’s still surely special licence territory. Is 3.5 kW a normal amp output anywhere? Alligator licence class?



Actually, it probably means you have a lower noise floor than the other station. The reception reports are signal to noise ratios, not signal strengths.


Thanks for the information about reports, was starting to think that I should be going QRO, but it sounds like it is explained by a fairly low noise floor


You reckon? I live in a pretty average suburban area full of noise sources too numerous to detail. If the S meter is not reading well up scale then the coax is disconnected. A quiet rf locale it is not. There might be a few locations that are 30 dB worse, but you would not bother turning the rig on if you lived there.

FT8 is a DX mode and IMO most ops are running as much power as they can without having a melt down. World wide, 3 kW amps are not uncommon amongst the SSB gang and 1 kW amongst the CW gang. Most of these guys are now on FT8 and running flat chat.

That’s their choice and I’m OK with that if their signal is clean. My choice is to leave the amp off, wind the power down a bit and see who I can work anyway. The highest reading I’ve seen on a DX station is +8 dB so I can expect up to 32 dB difference in reports coming up sometime. But for now the record is 29 dB.



Yes, it looks like it was seeing too low a resistance from my audio attenuator. Now that I have increased it to about 10k it seems to be working. I suppose below a certain value it configures as if you have plugged in headphones, i.e. stereo plug shorts ring2 to sleeve, with input from the internal microphone.


Great news that you found it Simon. Yes it can be too high or too low. I ended up with 13K, so my guess would be anywhere between 10k and 20k would most likely work.

Now I’m just going to have to solve the problem of how to see and operate the tablets screen on a summit! I think I’ll be reducing the screen resolution and installing the biggest mouse pointers I can find!

73 Ed.


WSJT-x V 1.8.0.rc3 just released and available here:


(That’s the windows version)

Versions for other operating systems can be found here:

Change log here - https://sourceforge.net/projects/wsjt/files/wsjtx-1.8.0-rc3/wsjtx-1.8.0.log/download


I finally got things working better and had two successful activations from GW/NW using FT8.


Thanks for the link to the rc3 program. I have downloaded it and installed it.

It works fine, but I do wish they would include documentation when they change the software! The “Lock tx/rx” tickbox has been changed (again) to “Hold tx freq”, and the various shift, alt and ctrl key functions are far from easy to grasp.

Walt (G3NYY)


Hi Walt,
I searched around for updated documentation when I got the alert about rc3 - all I could find was the change log file that I linked to. I presume a fully updated doco file will come later - perhaps when it’s no longer just a release candidate.

I think the main change is the frquency calibration mode that has been added:

Improved frequency calibration

Measure check box added to FreqCal mode, check to record to fmt.all
with current calibration correction disabled, uncheck to see the
impact of the current calibration parameters.

The fmt.all file is now optionally renamed to fmt.bak when a
calibration solution is accepted. This allows users to preserve an
fmt.all file that they might have edited for best fit.

A calibration procedure might proceed thus:-

1) select FreqCal mode,

2) step through suggested calibration test frequencies deleting those
that have no usable signal,

3) enable “Menu->Tools->Execute frequency calibration cycle” and check
that suitable signals are present,

4) select a suitable FTol and T/R period,

5) check “Measure” and let the cycle complete a few times to gather

6) uncheck “Measure” to complete the data capture, optionally tidy the
fmt.all file with your favourite editor,

7) push “Menu->Tools->Solve for calibration parameters” and accept if
you like what you see,

8) sit back and admire your accurately frequency calibrated station.


Hi Ed
There is some documentation on FT8 when you select Help > Local user guide (not the online user guide)
See item 6.6 in the menu on the left.
73 - Luc


Yes, Luc. But it is out of date. It relates to version rc2, and some of the links in the text are dead.

Ed … many thanks for the update about frequency calibration. Very useful.



Hi all,

I was reading now different threads and I’m not sure about best practice for FT8 SOTA.

Better use freetext macro to give my ref?

Standard CQ call
Using /P as suffix being portable is deactivating the locator.
and then follow with locator in
other call +my call +4digitLOC
other call +my call report
and RRR or R73 ?

or as suggested recently:

Looking forward to your thoughts.

73 Joe


I loaded the new version yesterday, but hadn’t used it so had not noticed the ‘Hold Tx Freq’ change.

From the web site:

  • Click somewhere on the waterfall to set Rx frequency (green marker on waterfall scale).
  • Shift-click on the waterfall to set Tx frequency (red marker).
  • Ctrl-click on the waterfall to set both Rx and Tx frequencies.
  • Double-click on a signal in the waterfall to set Rx frequency and start a narrow-band decode there. Decoded text will appear in the right window only.
  • Ctrl-double-click on a signal to set both Rx and Tx frequencies and decode at the new frequency.

That will confuse me when I want to start using it !


Hi Stewart!

Yes, it has taken me a few days to get used to having to hold down the Ctrl key when I want to respond to someone’s CQ call, or to move both my TX and RX frequencies on the waterfall. Version rc3 seems to be designed to encourage people to work “split”, which I feel is generally unnecessary on FT8. (When I call CQ, more than 95% of people reply to me on my own frequency.)

Hope all is well with you. I am thinking of coming along to the Cheltenham club one of these days, so may see you there.

Walt (G3NYY)


As it is for other modes, the advantage to working split is that callers are not all piled on top of each other creating a situation where no-one’s signal is decoded. This is supposed to be a weak signal mode. Often weak signals just cannot get through the pile up, as I know running only a few watts to a mediocre antenna. The downside of working split is a massive increase in the bandwidth required. I have already seen wall to wall signals on the waterfall with 15 or more decodes per period, so is working split a recipe for mayhem? I think I will stick with rc2 for now…

73, Gerald G4OIG


One newbie FT8 question that I haven’t been able to find an answer for: if I have “Enable TX” and respond to a CQ (i.e., double click on cq’ing callsign), but the dx comes back to somebody else - will the software keep initiating a transmit call every 30 seconds or will it know that the dx came back to somebody else and inhibit further calling by me?

Thanks & 73,
Barry N1EU


Hi Gerald,

Yes, there are pros and cons in relation to working split.

  1. There is a difference between a “weak signal mode” and a “low power mode”. FT8 is a weak signal mode. I use the minimum amount of power to enable me to be received by the distant station. This can be anything between 5 watts and 50 watts, dependent upon conditions and the distant station’s receive capability.

  2. In all versions of WSJT-X including .rc3, if the “Call 1st” box is checked when you are calling CQ the software will always preferentially reply to a decoded caller on your own frequency, rather than to someone calling on another frequency. Only if there is nothing readable on your own frequency will the software look for callers on another frequency. Therefore, calling “split” can often be self-defeating.

  3. Version .rc3 has implemented two-pass decoding. This often enables two stations on exactly the same frequency to be decoded, so a QSO can result even in severe co-channel QRM conditions.

Walt (G3NYY)


Hi Barry.

The latest release (.rc3) does exactly as you suggest in your second option IF you are transmitting on the same frequency as the CQ caller. I.e. it inhibits further calls if the CQ-er starts working somebody else. However, if you have chosen to answer “split” - i.e. on a different frequency from the CQ-er - your further calls will not be inhibited.

There is some logic to this! It is unlikely that you will have chosen the same split frequency as the other caller.

Walt (G3NYY)


I always work split on FT8. Even though G is hardly DX I have often had more than one caller reply to my CQ on my TX frequency. Even though I can see a strong signal on the waterfall the software struggles to decode anything.