If you dont send CW please pass over this text.
I use an FT857 for SOTA, not the best RX but an all round workhorse. I usually send CW using a double iambic paddle, in the past I used an external Bugambic keyer to generate the characters but to try and save a little weight I decided to try with the internal keyer circuit. Not mentioned in the literature, this is a Curis B keyer with dot and dash memories. My call contains 2 V’s and often I would find that I had been robbed of a dash. To overcome this I had to give the dash paddle a longer press, not easy when it is raining, freezing etc…
Today I decided to investigate. I set up a scope to look at the voltages across the dash paddle. I observed that the scope traces of my key closures were about 6ms long with steep sides, shorter than 5ms and the dash was not produced. The circuit diagram revealed J1011 was the key socket and that the supply to the dash paddle had a 1kOhm resitor (R1565) in series as a current limiter and a 1000pf RF supressor capacitor C1572 to chassis. I decided to try adding a bit of time constant to the dash.
After a bit of trial and error I found that adding 0.3uF (3 X 0.1uf ceramics) across the dash contacts produced a nice waveform and in conjunction with the 1kOhm resistor, an additional time constant of 0.3ms. On test, the problem of the missing dash had been solved, a test from 4wpm to 60wpm showed perfect dash preparation. The capacitor can be wired permanently across the dash (or dot is you are missing dots) key contacts. This may be a problem unique to my rig, however it may not. The FT817 uses similar circuitry so if you are being robbed of dashes (or dots) this minor external mod may suit you.