[Elecraft] K3 Variable amp keying delay >>not long enough!!??

[Ian White GM3SEK]

George wrote:
>The FAQ for the K3 promised a 0-50 ms amp keying delay.  We got only 
>0-20 ms in the latest release.    I don't think 20 ms is enough in my 
>case, based on my primitive research.  Several "experts" have said mine 
>might need as much as 30-35 ms.  My amp's relay is a 30 yr old Potter & 
>Brumfield open frame model with a 120vac  coil.
>

That's correct in my experience also. In supporting the Triode Board and 
Tetrode Boards, I have often measured 30ms switching delays in typical 
older amplifiers that use 2-pole and 3-pole open-frame relays.

Such amplifiers are still being used in very large numbers - and indeed 
they are still being manufactured. Since the K3 will be used with a wide 
range of unknown amplifiers, a 20ms delay is not long enough to 
guarantee there will be no contact arcing when the RF appears.


>To make things worse, since the K3 will not switch 120vac, I must use 
>another relay with a 12vdc coil in the line between the K3 and the amp, 
>which in turn pulls in the P&B relay.  This exacerbates the matter by 
>taking even more time, and I'm afraid to try anything for fear it will 
>damage the K3 and the amp.  Am I wrong??  Anyone care to comment??
>
That is correct. Attempting to switch 120V AC will damage the K3, and an 
intermediate relay will add even more delay.

But we have to draw a line somewhere. At some point we have to stop 
blaming the transceiver manufacturer, and accept responsibility for 
modifying the amplifier.

The modern interface standard is that the amplifier PTT input has an 
open-circuit (unkeyed) voltage of no more than +12-15V DC, and that the 
amp can be keyed by switching this input to ground with a current of a 
few milliamps. All modern transceivers meet that interface standard, and 
the K3 meets it very comfortably.

Amplifiers that require high-voltage, AC or negative switching are 30 
years behind the times. If the owner wishes to use the amplifier with a 
modern transceiver, I think it's clearly his responsibility to bring the 
switching interface up to date. Every design is different, but 
modifications are out there on the web, and none of them should take 
more than an evening's work.

If Elecraft can move a little too, by stretching the maximum keying 
delay to at least 30ms, then everyone could be happy.

With a little more effort, some of these slow relays can be speeded up 
electronically, without the trouble of changing to faster relays. The 
link below gives details of a circuit that can be used to speed up any 
RF relays operating from a +12-24V DC supply:
http://tinyurl.com/6n8wjg


-- 

73 from Ian GM3SEK
http://www.ifwtech.co.uk/g3sek/boards