Post subject: PIN diode bias question... Posted: Wed
Nov 23, 2005 6:29 am
Joined: Fri Aug 27,
2004 11:25 am
Location: Cape Town, R.S.A.
A fundamental question regarding PIN diodes and
their biasing – How does one calculate the minimum DC voltage required
to ensure that a PIN diode remains OFF in the presence of RF?
In a specific system implementation, where a PIN diode RF switch
is used to cold switch 500 W RF pulsed signal, a shunt diode configuration
is used. In this existing system a reverse bias of 50V is used to
switch the diodes OFF, but it was shown experimentally that the
reverse voltage could be dropped as low as 25V before the RF began
to bias the diodes ON.
First of all I understand that 500W
RF power corresponds to a RF voltage swing of 450Vp-p, if 50 ohm
is assumed. This on top of a -50V bias (for switching the diodes
OFF) means that the signal is swinging between -275V and +175V,
which one would think would bias the diodes ON during the positive
cycle. The minority carrier lifetime of the PIN diode however is
related to the fact that this does not happen and the diode remains
OFF for the complete duration.
The question then remains,
what is the relationship between minority carrier lifetime, RF power,
frequency and other PIN diode specifications and this minimum reverse
bias voltage to ensure that the diodes remain OFF.
Am I missing
something obvious? I’ve searched through various application notes
and other literature, which somehow always seem to ignore this requirement.
Post subject: Posted: Wed Nov 23, 2005 11:44 pm
Joined: Tue Sep 07, 2004 3:09 pm
This is an interesting question.
As I know when used as attenuators, pin diode long minority
carrier lifetime drives the hability of the diode to provide good
In a two tone intermodulation
test the enveloppe being non-constant, I see this also as the hability
to resist bias change by the enveloppe modulation.
if the equations for linearity of a pin diode attenuator in high
attenuation condition would be useful... like the one from MACOM
or Alpha. You would be using a criterion for extremely low distortion
meaning the diode absorbs no power.