[Elecraft] [Elecraft_K3] TX CW discrepancy

[Igor Sokolov]

Here is the picture to illustrate the problem.
http://forum.qrz.ru/attachment.php?attachmentid=29596&d=1264954098

again, QSK is not used. This shows one of the long series of dots sent with 
external PTT.
73, Igor UA9CDC

> Igor Sokolov wrote:
>> It has started with complains of some of our contacts in the contest. 
>> They told us that the weighting of out TX CW is not standard.
>> Two channel oscilloscope was brought into shack and 3 K3 tcvrs were 
>> checked:
>> SN 1651 SW ver 3.06
>> SN 2929  SW ver 3.68
>> SN 3553  SW ver. 3.41
>> On all 3 units series of dots were sent using external keyer with 50% 
>> weight. While on the monitor output the weight remained to be 50% the RF 
>> output showed shortening of dots (and therefore lengthening of spaces) by 
>> 7 or 8 msec. This effect does not depend on WPM speed and therefore less 
>> pronounced at slower speeds.
>> Could somebody confirm this please. We do not want to trouble Elecraft 
>> without being 100% sure that the problem exists.
>
> What is CONFIG: TX DLY set to? I have noticed that setting it higher than 
> about 10 or 12 causes CW element shortening. This is the number of ms 
> delay from the time the KEY OUT jack goes low and RF is generated. A true 
> QSK amp should be able to work with the default setting of 8.
> -- 
> 73,
> Vic, K2VCO
> Fresno CA
> http://www.qsl.net/k2vco