PA Performance Degradation Possibilities - RF Cafe Forums
Post subject: PA performance degradation possibilities
Posted: Fri Jul 07, 2006 10:38 am
Wed Jul 06, 2005 5:54 pm
I would really appreciate any comments in this regard..
What are the possibile causes for degradation in PA's performance
Post subject: PA performance degradationPosted: Fri Jul 07,
2006 6:10 pm
Joined: Wed Feb 22, 2006
For a PA's performance to degrade,
something has to change.
That something can either be the
active device(s), the bias circuitry, or the input/output matching
Normally, the bias circuitry doesn't change unless
it's underdesigned (for example, running 1/2 Watt dissipation into
a 1/4 W resistor).
The tuning of the input or output matching
circuitry can vary due to vibration, if adjustable devices are used
and vibration is present. This condition is relatively easy to fix.
However, the most likely change is in the active device, as
a PA section usually runs hot. Heat is the enemy of semiconductors.
Another issue with some active devices is "metal migration" - as
current density in the metallization on the active device increases,
the time to failure decreases.
are worse than others as far as degradation over time goes.
Posted: Mon Jul 10, 2006 7:49 am
Mon Jun 27, 2005 2:02 pm
As fred47 suggested, the worst enemy of semiconductors
is heat. Heat over time causes internal damage to the transistor's
junction and degrades the output power capabilities (P1dB/linearity
P.S: If you will operate a PA over time you
will see that the Ids increases and also the output power, this
is a drift of the Ids with temperature.
You will see in
many applications that there is a temperature compensation network
in the Gate made of thermistor. This forms a negative feedback that
reduces Vgs and hence decreasing Ids.
subject: Posted: Mon Jul 10, 2006 10:26 am
Joined: Wed Jul 06, 2005 5:54 pm
Thanks Fred and IR.