[CW] "Amp Sequencer"

[N2EY at aol.com]

In a message dated 5/19/05 4:39:55 AM Eastern Daylight Time, n1ea at arrl.net 
writes:


> The web page (see below) says that it is "near QSK".

Which means it isn't really QSK. 
> 
> What happens is that it inserts 10 ms delay after you close the key, 
> switching the amplifier circuit relay (which will close within 10 ms) then 
> energizing the transmitter will will go into transmit - it will stay on 
> through the length of the pulse - then the sequencer will allow the amp 
> relay to close.
> 
> Switching pulses are sent first to the amplifier relay, delayed by 10 ms to 
> and then sent to the transceiver (but with the same length), ... ... ...
> 
> It is obvious to me now that the switching pulse to the amplifier has to be 
> lengthened so that it is around 15 ms (delay time of the pulse sent to the 
> transceiver, plus whatever time it takes for the r.f. to reach zero...)
> 
> So if we have a 20 ms pulse (a dot at 50 wpm), I think it figures as such:
> 
> 20 ms pulse sent.
> 
> Time = 0
> Relay voltage sent to amplifier relay.
> 10 second buffering delay by sequencer
> Amplifier relay is fully closed @ T+10 ms
> 20 ms pulse starts to be sent by the transceiver @ T+10 ms
> 5 ms rise time  finishes @ T+15 ms
> 20 ms pulse finished by transceiver @ T+35 ms
> 5 ms decay time @ T+40 ms
> 
> So we have a minimum of 40 ms that the linear amplifier relay will be open - 
> 
> which is equal to a 50 wpm dot and the associated interelement space (both 
> 20 ms long).
> 

Bingo


> So at 50 wpm, you will never hear between dots, but you will hear between 
> letters as that would be 60 ms - but the window will be only the same size 
> as a 50 wpm dot!
> 

Yep. 

> I've seen vacuum relays that switch in 6 ms close, 6 ms open - so that's 12 
> 
> ms.
> 
> I have no idea how long it takes for a key to be closed and a transceiver to 
> 
> start sending r.f.
> 

Depends on the rig. Some are very quick, others take more than a few ms.


> I wanted to put vacuum relays in my NCL-2000 but Carl, KM1H is saying that 
> because of the screen circuit, it isn't advisable, also because of hot 
> switching.
> 

I don't see where the screen circuit makes any difference.

The main factor is that the antenna must be connected solidly (switched) 
before any RF gets to the amplifier. 

---

Here's what I'd do:

First off, most "modern" rigs and amps aren't meant for real QSK. They're SSB
sets with a key jack. Been that way since the KWM-2. So the first order of 
business is to get or build a rig that's good for CW. 

The second part is to get away from the concept that the antenna must be 
switched at all. Why not leave it connected to the rig output at all times, and 
have a highspeed relay or electronic circuit to disconnect the rx input?

More than 45 years ago, W2LYH showed how to do this. His key operated a 
highspeed relay that, when open, connected the rx to the antenna, and when closed, 
keyed the rig. It was in QST about 1960. Of course his final was run at or 
beyond cutoff.

73 de Jim, N2EY