locked Re: Acceptable PA Power Supply Ripple?


Richard Lamont <richard@...>
 

On 06/03/2020 16:16, groups@... wrote:
On 06/03/2020 15:45, Steve Kavanagh via Groups.Io wrote:>> Has anyone any experience or opinion on what an acceptable power
amplifier supply ripple is for WSJT-X modes?
[snip]
73,
Steve VE3SMA
I'm surprised your licencing authority allows a 2% ripple.  I'm sure
ours would be very unhappy if it came to their attention.

Roger G4HZA
Our licensing authority (in the UK) requires that:

7(1) The Licensee shall ensure that:
(a) the emitted frequency of the apparatus comprised in the
Radio Equipment is as stable and as free from Unwanted
Emissions as the state of technical development
for amateur radio apparatus reasonably permits; and
(b) whatever class of emission is in use, the bandwidth
occupied by the emission is such that not more than 1% of
the mean power of the transmission falls outside the
nominal modulated carrier bandwidth.

In the case of an FSK transmission with a single pair of hum sidebands
spaced 100/120 Hz above and below the main signal, 7(1)(b) would be
broken when the sidebands exceeded -23 dBc.

The level of hum sidebands with 2% ripple will depend on (a) how you
define ripple and (b) the characteristics of the amplifier. A saturated
Class C amplifier would be more sensitive to ripple than a linear one.

If you define ripple as half the peak-to-peak ripple voltage expressed
as a percentage of the DC voltage and use the unlikely worst case of a
Class C amplifier, then you have something analagous to a high-level
modulated AM transmitter and 2% ripple will produce 2% modulation. At
100 % modulation, each AM sideband is -6dBc. So at 2% modulation, the
ripple sidebands will be 37 dB lower, i.e. -43 dBc.

In practice, most of us are using linear amplifiers in which, unlike the
Class C case, the gain is not proportional to supply voltage.

Therefore I reckon 2% ripple would comply with 7(1)(b) above with at
least 20 dB to spare!

Nevertheless, we should aim to keep our spurii as low as possible,
especially when dozens of stations are all using narrow slots in same
crowded SSB channel and one man's spurious is everyone else's raised
noise floor.

73,
Richard G4DYA

Join main@WSJTX.groups.io to automatically receive all group messages.