> As an alternative, how about switching the bias by means of a signal
derived
> from the audio, so the PA is switched before the RF gets there? Sounds a
bit
> like having relays and interlocks, doesn't it?
That can be arranged easily with a number of modern amateur transceivers and
an electronic bias switch like Rich's. Suitable rigs assert a TX signal
several milliseconds before allowing the RF envelope to rise. Since the TX
signal is asserted when the VOX is tripped, it is effectively an
audio-derived signal. If this signal is used to key the amp, the electronic
bias switch will apply operating bias voltage to the tubes before the RF
gets there. Rich's electronic bias switch uses an optocoupler to enable a
bias switching transistor. As I recall, the optocoupler is turned on when
the keying line is grounded, so the bias is switched long before the input
and output relays close.
For example, the Kenwood TS950SDX asserts a +12V signal a full 15
milliseconds before the RF envelope begins. The Yaesu FT-990 grounds a
back-panel pin 5 milliseconds before the RF starts. Instead of using the
rig's slow and noisy "linear" relay, it is often possible to use the TX
signal, although a simple solid state interface is required if the signal
has to be inverted (e.g., the 950) or the amp's T/R system uses a high
voltage/high current relay (like the 110VDC relay used by the stock SB-220.)
The ground-on-TX pin provided on modern Yaesu rigs can be connected directly
to a low voltage/low current amp T/R system (less than 50V@100mA for the 990
and less than 15V@100mA for the 1000mp.) Some care is needed not to short
the 950 TX signal line or exceed the voltage and current ratings on the
Yaesu lines. Damage to the rig may otherwise result (at a minimum, the
output transistor will have to be replaced.) I once modified a TS940 to get
at the +12V TX line, but as the line is not protected with a transistor
buffer, an accidental short can cause widespread damage.
BTW, the 950 service manual has timing diagrams that clearly show the delay
for both CW and SSB modes. In fact, the user manual provides instructions
for cutting a jumper to increase the delay time to 30ms for amps with very
slow relays. I had to use a dual trace scope to check the FT-990's TX signal
timing. The timing may apply only in CW mode -- I didn't scope the timing in
SSB mode, which would have required some sort of audio pulse with zero VOX
delay (or a storage scope.)
If the rig imposes enough delay, proper bias switching can be achieved even
if the amp uses a mechanical relay to switch the bias, provided that the
relay switching time is less than the delay. For example, the LK-550
switches the bias with a dip relay rated at better than 8ms switching time.
The 950 should be able to drive that amp properly, but the 990 might be on
the hairy edge. However, my guess is that the rise time of the typical SSB
signal is slow enough to prevent splatter on the first sound, and the VOX
hang time will take care of the rest of the transmission.
The timing considerations get more critical when attempting QSK with a
suitable mechanical relay (e.g., Jenning vacuum relay.) Here the bias and
relays switch on every element and the RF waveform rises very rapidly. The
rig and/or amp must provide enough delay to get the bias switched and the
relay(s) closed before RF is applied. Some amps (e.g., some by Ten-Tec, Amp
Supply and Alpha) allow interlocking through a key output signal. You
connect your keyer to the amp relay or KEY IN line, then connect the amp's
KEY OUT line to the rig's key input. This way, your keyer keys the amp
first, then when the relays are closed, the amp keys the rig. The amp may
use a relay sequencing technique and/or a time delay circuit to ensure that
the rig is not keyed prematurely. Alternatively, the Yaesu rigs have a TX
INH pin which must be grounded for RF to begin. In this case, the rig's
ground-on-TX pin is connected to the amp's KEY IN jack and the amp's KEY OUT
jack is connected to the rig's TX INH pin. This allows you to use the rig's
internal keyer. This works, although I've observed some slight chirping with
this system, either due to a flaw in the TX INH circuit or RF getting into
the rig via the TX INH line.
Unfortunately, the combined rig and amp delays shorten the length of each CW
element. This is not a problem for speeds up to about 20-25 WPM, but may be
unacceptable at higher speeds. The effect is a very choppy sounding signal.
One solution is to use a keyer that allows the weighting to be adjusted to
counter the effect. The transmitted signal sounds good, but the sidetone
signal may be annoying if you don't like weighting. If the amp uses a delay
circuit, it may be possible to adjust component values to shorten the
combined delay to the absolute minimum required by the relays. I suspect
that good sounding high-speed QSK CW can be achieved with a delay of
10-15ms, which should be enough to switch a vacuum relay.
73, Dick, WC1M
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|