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

Datamode - FT8


I did two activations today in FT8 and also managed the first FT8 S2S, with OK1CYC (YO/WC-043 to OK/KR-001) on 7MHz. This mode is really fun in portable, at some point i was decoding at the same time on 20m stations from Indonesia and USA. I’ve used 1W or 2.5W trom a FT-817.

Razvan M0HZH / YO9IRF.


Really fun to make the FT8 S2S with Razvan today.
Thanks Razvan for waiting on the summit. It’s always trouble to timeplan with two little daughters hiking with me :slight_smile:

@G4TJC Simon, Helen, I will do my best to be on SOTA around 1800 GMT tomorrow to try another S2S.


My experience yesterday was mixed.

I was trying to get it going properly for about an hour before I gave up and went to CW. But then I realised my main mistake (rig on 20m, antenna set for 30m - oops!) and spent another hour trying again. I think I finally heard a reply to a CQ, only for him to fade away (a G). But I did finally achieve a few QSOs around Europe and was heard as far away as Brazil.

Trying to use Windows on an 8-inch tablet is horrible.


Heading out to YO/WC-001 right now, I should be on the air around 12:00 UTC, will be the 6th FT8 SOTA activation I’m doing this week. This is from YO/WC-052 earlier today:


Looking for ideas what to check…

I have my FT-817ND connected to my ACER Windows 10 tablet running WSJT-x v 1.8.rc2 for FT8.
I am using a Wolphi link interface for the Audio circuits and an RT-systems cable for the CAT commands.

Cat commands working fine - test CAT and test PTT work 100%.

Sending appears to work fine also (although PSK spotter is not reporting the transmission). I have a 40db level receive audio level coming in (I’ve tried up to 60db) the waterfall display shows signals but WSJT-x is not decoding the incoming data.

I have checked all settings against a working setup between my Windows 10 laptop (with the same version of WSJT-x and my ICOM IC-7300). Apart from the obvious comms port and rig type differences all other settings are the same. The configuration with the laptop and IC-7300 works 100% decoding incoming data and is spotted by PSK spotter on transmit.

I had thought that the windows tablet was not realising that an external “Mic” was plugged in previously but since reducing the resistance value acoss the 4 pin 3.5mm jack to 13K and checking everything, I’m happy the only input to the tablet now is that being output from the FT-817.

The fact that the waterfall looks “normal” but nothing is decoding is what I find hard to explain.

Any ideas? Anything that needs changing in the FT-817 perhaps? Perhaps audio corruption coming in somewhere?

73 Ed.


You did set the MODE to FT8, did you ?
(Just asking because I wanted to test it on another laptop too, and a fresh install of WSJT-x defaults to JT65 mode … kept me wondering for a while why it didn’t decode FT8 … hi)

And for those that want to build a very easy and quick interface, here some plugging of my blog :

Almost all components come from the junkbox, it doesn’t have to be complicated to work !

And on the 44/48 kHz matter : I checked my Linx 7" tablet running Windows 8, and it has the 48 kHz setting (and cannot be changed), so I’m lucky. Made 11 FT8 qso’s last week from ON/ON-027.



Hi Luc,
Yes set to FT-8 - all settings the same as the working configuration.

I see you are using a KX-3, my problem may only be an issue from the FT-817 using the data in and data out pins of the data socket on the back of the rig.

The Wolphi-link Interface works fine between the 817 and an Android device, so I would have thought it would be OK with windows as well. Maybe not. Unfortunately I cannot use the built in windows line monitor function as that send the monitored output to the same 3.5mm jack and so unless I build a break out cable I’m not going to get to that. At the moment my thoughts are around possible RF Ingress.



Hi Ed,
I had the same problem, Maybe you should synchronize the clock.



Clock is synchronised - I am running Nettime on the tablet.

HOWEVER, YAHOO !! I have found the reason it wasn’t working.

In case others are using an FT-817 - this is important.

I set the FT-8 software to select the digital mode in the settings which it does via the CAT commands. In the FT-817 there is support in the menus for different data-modes and the default set-up is RTTY. RTTY ALWAYS uses LSB irrespective of band and of course all modern data modes use USB irrespective of band. When I changed the FT-8 software to use USB mode rather than data, it started recieving and decoding!!
Of course it’s better to have the narrower filters of the Digital mode, so I went in and configured the “USER-U” mode in the FT-817 as the default digital mode on the FT-817.

I changed the WSJT-x FT-8 software back to use digital mode and PRESTO it kept decoding!! The drive level needed adjusting to remove any ALC and then checking PSK reporter and my 2.5 watts of FT-8 is being received by two lads in the North of Scotland - thanks guys!!

Big relief - the reason it wasn’t decoding was because the FT-817 was listening on the wrong sideband.

73 Ed.


And dont forget to change it back to LSB when you use voice on 40M. Dont ask why I know that, but thanks to VK2KEV for pointing this out to me on air last week when my CQ calls were unanswered.



Hi Compton,
He He, I could see that happening to me, however if I don’t switch the 817 from Data mode I wouldn’t get any modulation, so by default when changing back from FT-8 to Voice, I’d be changing mode from DATA to LSB (or USB for 20m etc) in any case.

73 Ed.


Actually you could do with the filtering being as wide as possible according to the manual.

Reading your original post I was about to post that you were on the wrong sideband - but then you found it!
Have fun


Hi Barry,
I also read in the meantime that a bandwith up to 5KHz filter is best, so as you say as wide as possible, which suggests I should use the USB rather than the DATA mode and let the software do the filtering. I have the Collins 2.4KHz filter for SSB in the FT-817 so I guess that will be my maximum bandwidth. main thing is that it is now working! I wonder why the option is in the WSJT-x program to select USB or Data/Pkt ?

Just need to try it from a summit once the S6-S9 (K=6) conditions pass!

73 Ed.


Hi Ed,
I don’t see the point of having a very wide pass band if only using FT8. The dial is set 2 kHz below the JT65 dial frequency so being able to listen much above 2 kHz seems pointless, regardless of what Guru Joe says IMO.

Not everyone is properly time synched so this is one reason they don’t decode. They may be on the margin of the tolerance and getting some decodes from you. Another increasingly common reason is two stations replying at once. Calling 50 Hz off is a revisit of the old “up 2” when working DX.

The latest version released at the start of Sept …r2 decodes faster and digs 4 dB deeper. Supposed to decode two signals in the same frequency but have only seen this work with about 20 Hz offset between them.

Lack of response can be due to noise at the other end or they are alligators. As it’s a DX mode primarily, pip squeak SOTA stations will find themselves competing with 1.5 kW big boys. S2S of course is not in the big amp category and chasers will call anyone they can copy, right down to -24 dB.

QSB is another factor. 20 dB swings in minutes is not unusual. If the band drops out during the synch times then no decode is possible.

All part of the fun.



A long time ago I saw someone questioning one of the very early JT modes on a forum and more importantly questioning whether the software author knew what he was doing. “Anyway who is this Joe Taylor that suddenly knows so much about weak signal modes?” went the question at the end of the post. And the reply was very simple “He has just another ham. Oh, and 1993 Nobel Prize winner.”

That is all.


He got the Nobel Prize for discovering binary pulsars, by listening to RF pulses coming from 21000 light years away, with receiver technology from the 1970s. From those pulses, he also calculated their mass, size, orbits, decay, with a 0.2% error between predicted and experimental data.

I’m sure he might know a thing or two about weak signals :).



Yes, Pulsar hunting has quite a lot in common. What makes it even trickier is that after travelling through light years of interstellar/intergalactic medium, which is very slightly dispersive (some frequencies traveller faster than others) the sharp pulses coming from the Pulsar have been smeared out into something you wouldn’t ever spot, except for the ones which are really big and close by. So that adds a big parameter space to search through, reconstructing clean pulses, before stacking pulses together to get something detectable. Years ago this meant adding a whole other piece of gear (a de-dispersal unit) ahead of the detection system. Nowadays I think it’s all in the number crunching.


I’ve met Joe; we both presented papers at the same conference many years ago. He’s a very personable bloke. I’ve used his software from day 2. So your suggestion I don’t know who he is or what he has achieved is misplaced.

Now I prefaced my post with “if only using FT8”. There’s the key.

Some people are (blindly) saying you must have a wide bandwidth purely because they read Joe’s suggestion. But they do not know why he is recommending that.

You need to do your own thinking. Why have a 4 kHz bandwidth for a mode limited to 2 kHz? If you are not interested in anything else than JT8, then a decent receiver set to a pass band a smidgen over 2 kHz gives you an unattenuated 2 kHz and that’s all you need.



Hi Ron,

Thank you very much for alerting me to the fact that Ver. …rc2 has been released. I was not aware of this, and had been soldiering on with …rc1 ! The update was not well publicised. The first I heard of it was your posting on here.

It is, indeed, much improved over Ver. …rc1. The sensitivity is now only 2 or 3 dB less than JT65, and it does cope very much better with QRM and multiple stations calling. There is now 2-pass decoding, and the decoding is much faster.

Unfortunately, an awful lot of people are still unaware of the new version, so often I can copy them with weak signals but they cannot copy me!

Walt (G3NYY)


Good Morning Walt,

I’m glad you like V…r2. I’m surmising that Franke and Taylor are wanting to do a few more tweaks before releasing a “proper” version. We are fortunate that Joe and his associates are prepared to give us free and for no charge megadollars worth of software. JT8 is the biggest thing to happen to AR since the first WSJT modes were released IMO.

I have mixed feelings about publicity of the new version. Already it is common for me to decode 20 stations at a time. The RX window scrolls up and I have to be quick to roll it down again to check who is calling before a new set of decodes pops up. We will soon need to double the frequency range used and then we will need a 4 kHz + bandwidth for RX and TX.

Probably multiple callers on the same frequency and local noise are significant factors in not getting a response. Add to that the high power being used by many - I look at my linear and so far have resisted flipping the switch.

The QSB is fascinating. 20 m is behaving like 6 m. Yesterday a ZS6 station came up out of the noise to a -6dB. We did not quite complete a QSO before he went down and out.

I’ve now had QSO’s on all bands from 160 to 10 m inclusive. Not many on 10 or 12 m but it’s wonderful to be able to use them in the current conditions.

I will be returning to a peak with JT8 as soon as I can sort some things out here but in the meantime usually look around on the bands from about 0600 UTC, starting with the highest I can see activity on.

I’ll be looking for familiar calls and activators on my screen.