Re: Diode T/R Switching - Long

---------

From: Ed Pacyna (pacyna@auratek.com)
Date: Wed Mar 19 1997 - 17:10:27 EST


T/R switching with PIN diodes offers numerous benefits (e.g. cost
effective, speed, reliability, silence etc.) and need not be a determinant
to performance.
 
The I in PIN stands for the intrinsic un-doped silicon layer (has very few
charge carriers to support the flow of electrical current) between the PN
regions and produces a very slow diode response resulting from charge
storage. The parameter of interest is the charge lifetime (the period
required to remove charge from the junction through recombination and
migration). So far as RF energy is concerned, the PIN diode is a resistor
whose value depends upon the dc bias current flowing. There are two
different types PIN diodes (resulting from differences in the doping), One
is for switching (R high, no bias and R low with bias) and the other is
used as a current controlled resistor. PIN diodes are used in various
VHF/UHF applications such as tunable attenuators, limiters, phase shifters,
AM modulators and switches.

At lower frequencies the charge lifetime is short with regard to a single
cycle of RF. The device in this case displays PN junction diode action and
distortion occurs (non linear diode action). Most PIN diodes are for VHF
and higher and often perform poorly at the lower end of the RF spectrum.
Another consideration is that the RF current swing must not deplete the
charge in the PN junction. To avoid this possibility, the carrier lifetime
needs a be a least a minimum value at operating frequency which is
determined by:

T = RF signal current through the diode / (2 pie x frequency x DC forward
bias current).

Although the presence of a PIN structure in the 1N4007 makes it seem
attractive to some experimenters and manufacturers as a cheap alternative
to RF specified PIN diodes. However, the 1N4007 was not intended for RF
switching use and it has been demonstrated that the insertion loss when
unbiased or reverse biased and the IMD performance when unbiased is
inferior to RF specified PIN diodes (Ref: Exploring IMD in RF switching &
tuning diodes, QST 12/94).

Being penny wise and dollar foolish and/or poor application technique has
resulted in plenty of egg on face (including some commercial
manufacturers). The purpose of a T/R switch is to isolate the receiver from
any RF power (including mismatched conditions) during transmit and to
provide a low loss, low distortion path during receive. Very high degrees
of performance and isolation can be achieved through the selection of
proper devices, circuit configuration (e.g. using series shunt diode
combination, quarter wavelength sections, reverse bias) and operating
conditions.

73

Ed, W1AAZ
 
>It had been written that:

>A PIN diode is built with a special characteristic. It acts like a regular
>diode, but its response to a rapidly varying signal is very sluggish.
>Thus, at DC, it looks like any ordinary diode, but at RF AC, it doesn't
>rectify very well. This is useful for RF switches because we don't want
>the switch to rectify our RF signal but we do want to be able to control
>switch conduction with DC currents. The 1N4007 used in the 38S rectifier
>happens to be built like a PIN diode.

>So, if we push a few milliamps of DC current through
>the diode all the time, then add a tiny amount of current from the RF
>signals coming in off the antenna, the diode will pass those RF currents
>without much apparent resistance. Add to this the fact that the PIN diode
>is just too slow to react to RF, and we have what looks to the RF signal
>like a switch in the closed position.
>
>Now, remove the DC current from the diode. No current is flowing and the
>voltage across the diode is zero. Now apply just a little voltage -- a few
>millivolts. Almost no current flows because you haven't come anywhere near
>the sharp bend in the exponential curve. The diode is behaving like a very
>large valued resistor, since increasing the voltage across it causes only a
>very small increase in current through it. A very large valued resistor is
>almost the same as an open switch, preventing signals from passing through
>it. Therefore, with no biasing current on the diode, the tiny RF signals
>from the antenna cannot pass through it to the receiver.
>
>In the 38S, a few milliamps of DC current are passed through the switch
>diodes in the forward direction in receive mode. The amount of current is
>limited by the series resistors R1, R2, and R3. In transmit mode, no
>current is passed through the diodes. Therefore, in receive, the diodes
>act like a closed switch, passing signals from the antenna to the receiver.
> In transmit, the diodes act like an open switch, cutting off signals.
>
>"But wait!", the clever person asks, "What about that huge transmit output
>signal -- many volts of it -- that appears on the antenna side of the
>switch?" "Won't that cause the diode to conduct, and pass at least part of
>the signal through to the receiver?"
>
>That's why the PIN diode is used. In its normal off-state during transmit,
>the diode just sits there not conducting. Suddenly, the transmit RF
>voltage rises at the beginning of the first cycle. The lazy PIN diode is
>too slow to react, and by the time it notices, the RF sine wave has gone
>back to zero again. Essentially, the PIN diode is too slow to work like a
>diode at RF frequencies, so the transmit RF doesn't turn it on very much.
>It needs some DC current flowing through it to turn on.


Search QRP-L Archives

[ QRP-L Archive | ]
[ 1993 | 1994 | 1995 | 1996 | 1997 | 1998 | 1999 | 2000 ]

---------

This archive was generated by hypermail 2b29 on Fri Jun 02 2000 - 11:33:17 EDT

kd4ab@kd4ab.org