[Elecraft] FTDX5000 Design Flaw

[John, KI6WX]

There have been a lot of comments on the reflector about the QST review of
the FTDX5000.  However, no one has noted a significant design flaw in the
transmitter that shows up in the review.  Refer to Figure 3 that shows the
composite transmitter noise.  It shows a noise level of -130 dBc/Hz from 10
kHz to 1 MHz (and probably beyond).

 

This transmitter when combined with a full power amp could wipe weak signals
in an entire amateur band for anyone living within a few miles of the
transmitter (see the math calculation below).  While a number of other
transceivers have this problem, I wouldn't expect a top of the line $6K
radio to have such lousy composite transmit noise.

 

This problem is created either in the radio's synthesizer or its transmit
amplification chain.  The K3 was specifically designed to minimize composite
transmit noise.  The K3 QST review showed a transmit noise level of -155
dBc/Hz at a 100 kHz offset.  This is 4 S-units less noise than the FTDX5000
at the same offset.

 

This is not a theoretical calculation.  I know of one case of composite
transmit noise where an amateur transmitter wiped out weak signal reception
across an entire ham band in a receiver located several miles away.

 

-John

 KI6WX

 

 

CALCULATIONS

 

Assume that we have a FTDX5000 transmitting CW on 20 meters followed by a
1.5 kW amp.  The transmit power is +62 dBm.  At a 100 kHz offset, the
transmit noise is -68 dBm/Hz.

 

Assume that the FTDX5000 transmit output is fed to an isotropic radiator (0
dB gain) on top of a hill and we have a receiver also with an isotropic
antenna in a valley with line of sight to the hill.  For directional
antennas, the sum of the antenna gains depends on where they are aimed and
could be greater or less than the 0 dB in this example.  For the moment,
we'll place the receiver 1 mile from the transmit antenna.

 

The path loss between the transmit and receive antennas is 60 dB, which
implies the receive power of the transmit noise will be -128 dBm/Hz.  The
normal atmospheric noise on 20 meters is about -144 dBm/Hz, which means that
the transmit noise will be 16 dB greater than the normal background noise.
This noise will be spread across the entire band whenever the FTDX5000 is
transmitting.  If it is transmitting CW, the receiver will hear noise
modulated in Morse code.  If it is transmitting SSB, the noise will vary
with the voice modulation peaks.  The receiver would have to be more than 6
miles away for the noise to drop to background levels.

 

Another way to look at this problem is how many S-units would the show up in
a 500 Hz receive bandwidth.  The total power in the noise is -101 dBm in the
500 Hz bandwidth.  S4 is -103 dBm, so the noise would be about a S4 signal
level.  Each time you halve the distance to the transmitter, the noise will
increase by 1 S-unit.  If you live 1000' from a FTDX5000, you could see a
noise level of S7.  You can reduce the noise by using a narrower filter, but
you would have to drop down to a 100 Hz filter to reduce it by 1 S-unit.

 

This calculation was done with the transmit antenna on top of a hill so we
could use free space radiation to calculate the path loss.  If both antennas
are on a flat surface of earth, the path loss will be somewhat greater, but
the exact magnitude requires using antenna radiation software such as NEC-4.