cw vs. pulsed peak power

[Pat]

Could someone explain to me if there is a rule of thumb to determine
the peak pulsed power of an (linear class a/b) RF amplifier if the
peak cw power is known. For instance, if the amp is rated 300W (max
power) in cw mode, the max power in pulsed mode is definitely higher,
and surely depends on the pulse width (e.g. could be 1 kW for a 10us
pulse). Is there a rough rule of thumb to figure this out without
having to test it directly and risk blowing up the amp? The amp design
is based on the RF power mosfet MRF151G (the tech specs do not say
anything about pulsed mode ratings). thanks.

 

[Jan Panteltje]

Could someone explain to me if there is a rule of thumb to determine
the peak pulsed power of an (linear class a/b) RF amplifier if the
peak cw power is known. For instance, if the amp is rated 300W (max
power) in cw mode, the max power in pulsed mode is definitely higher,
and surely depends on the pulse width (e.g. could be 1 kW for a 10us
pulse). Is there a rough rule of thumb to figure this out without
having to test it directly and risk blowing up the amp? The amp design
is based on the RF power mosfet MRF151G (the tech specs do not say
anything about pulsed mode ratings). thanks.

This is not so easy to answer.
In all cases if you do not boost supply voltage then you are
still limited by the max current the FET can do.
If you stay in range of that you are limited by _thermal_ issues.
In on transmitter I have, the amp is 30W CW and about 150W pep (SSB)
with 2 junction transistors in push pull.
The issues are also the losses in your coils etc....
I am sure I could peak it a bit more.... but things would start frying.
Factor 5 should be possible, (SSB always assumed average of speech is 20%
of max power, so if I talk up to 150W SSB I use about 30W average.
It is _thermal_, watch the max temperature.

 

[Tom Bruhns]

Could someone explain to me if there is a rule of thumb to determine
the peak pulsed power of an (linear class a/b) RF amplifier if the
peak cw power is known. For instance, if the amp is rated 300W (max
power) in cw mode, the max power in pulsed mode is definitely higher,
and surely depends on the pulse width (e.g. could be 1 kW for a 10us
pulse). Is there a rough rule of thumb to figure this out without
having to test it directly and risk blowing up the amp? The amp design
is based on the RF power mosfet MRF151G (the tech specs do not say
anything about pulsed mode ratings). thanks.

No, there is no rule of thumb. It depends on the design of the
amplifier. Some are designed specifically for high peak pulsed power,
where the ratio of peak to average may be quite high, and some are
designed to run with a continuous unvarying output. I'd say it's
common for a linear amplifier designed for AM service to be able to
handle peaks at least twice the voltage (= 4 times the instantaneous
power) of the carrier, but that's just one example. Amplifiers
designed for a high peak to average power ratio are, I'd say, more
commonly not linear, however.

You can look at the peak collector/drain current and collector/drain
to emitter/source voltage ratings of the device; you can figure power
dissipations; you can figure the peak output voltage for a given
load...all those will go into finding the value you seek, but it also
comes down to what you expect for MTTF, too. Eventually a stressed
amplifier will fail--sooner rather than later as you increase the
stress. The stresses include voltage, power, thermal cycling, i^2*t
fusing currents, ... The power supply capabilities enter into it too,
of course: the supply may be capable of delivering a high current for
short periods, but a much lower average current--or it may not.

Cheers,
Tom