05 Sep 2010 - 02:31 GMT


AMPSEQUENCER Technical bits



CW poses serious stresses to exciter and amplifiers components.
Generally when the key is closed, it  will take 6 to 9 milliseconds for the transceiver to develop full power output, mean while non vacuum mechanical relays achieve full closing typically after 15 to 25 milliseconds depending on relay type and the manner which is driven.  The closing time found on QSK vacuum relays varies around 6 to 10 milliseconds.
If you are driving a non QSK amplifier with CW speeds greater than 30 characters per minute, chances are that the first dot can be left out entirely while the first dash is shortened to the point of being mistaken by a dot.



The figure shows typical timings.
A 5 millisecond rise time envelope has been generally adopted to reduce occupied bandwidth. With this shaping factor, it takes about 3 milliseconds for the exciter to deliver 50% power to the amplifier relay contacts.

When fast vacuum relays are used such as in QSK amplifiers, the ON-delay (time to transition from RX to TX) could be on the order of to 6-10  milliseconds, you can see that even amplifiers with vacum relays may not be HOT-SWITCHING free when operated on CW mode.

In amplifiers with mechanical relays the ON-delay can vary between 15 to 25 milliseconds or more.
If you do some math the first dot or dash on a CW transmission at 30 characters/minute will be completely lost or severely shortened.  Did you ever wonder why that first character is so hard to copy by DX stations when using the amplifier?
The above situation is not just a nuisance, but will lead your amplifier components to premature failure.

Lets analyze what really happens:
during the first few milliseconds of key down (be it CW or RTTY),  as the mechanical relay contacts travel from the receive position to the operating position, the air-gap between the contacts on the input  of the amplifier's "Normally Open" side becomes increasingly smaller. At some point the contacts become close enough to act as a small capacitor allowing the signal to pass and be amplified.

So even when the contacts are not fully closed, RF will be present on the input and amplified. As the contacts keep approaching, breakdown of the air dielectric will occur causing a spark to bridge the distance.
In most amplifiers the contacts that switch RF are either ganged on the same relay or two relays are driven simultaneously. In either case while the input RF is capacitively coupled to the input side of the amplifying device (tube or transistor) the output contacts are in fact also moving, hence they present at best a 50 ohm load coupled in series via the contacts air-gap, this is hardly the ideal matching situation.

Two things are to be considered:
first, for an instant, there will be no load connected to the amplifier output while it is delivering nearly full power output. Needless to say about the stresses caused by an infinite SWR on the amplifier components.
second, each one of the spark occurrences will vaporize a minute portion of the relay contact forming a carbon deposit, which will decrease the conductance of the contacts by a few Milli-ohm each time, eventually leading to contact failure.
Similarly on RTTY many software programs simultaneously deliver the audio output and close the PTT, which causes amplifier relay hot switching. Your transmitter is not too happy either since instead of a 50 ohm load it will see an open line for about 5 to 20 ms (depending on the amplifier). Chances are that the amplifier finals will develop full power over an open line, causing stress and early component failure.
Hot switching is the leading cause for relay failures in amplifiers.


One way to prevent it is by activating the amplifier relay (s) before keying the radio.
To safeguard your costly radio and amplifier components we have developed the AMPSEQUENCER a microprocessor control board capable buffering your keying while sequencing the amplifier relay and only then passing your keying to the rig.
A nominal PTT delay of 30 ms introduced by the AMPSEQUENCER this is more that enough to allow even a couple of cascade relays to fully close before the RF signal is applied. For CW mode, the delay and buffering can be user programmed between 10 and 80ms.



At US$39.oo plus shipping, this may be the best investment you will make to protect your amplifier!!


NETCERTUS information

© (2002-2010) NETCERTUS
http://netcertus.com/cgi-bin/netc/net.cgi?AMP.tech