This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_001_01C05E2B.352509F0
Content-Type: text/plain;
charset="iso-8859-1"
> However, transistors can fail very rapidly from excessive voltage.
> You have zero-time to prevent that failure, since it is a punch-
> through failure. Temperature really isn't the major concern. Excess
> voltage is, and exciters with keying transients (common today) can
> really hammer the transistors with collector or drain voltage spikes.
> That's another reason why the optimum load impedance is
> generally set to be higher than the working impedance.
>
[Dave D'Epagnier] Tom, I know of at least one prominent RF
transistor manufacturer that would take issue with your statements about
temperature above. My experience as well shows that high temperature is
almost always the cause of failure in a properly designed solid state amp.
This high temp is usually cause by load mismatch. If voltage breakdown is a
problem then improper devices are being used and it's a poor design to begin
with.
> > a transistor you will see that the die has a typical response of 10ms
> or
> > so. So if you can react quicker than this (say in about 1ms) then you
> can
> > adequately protect the transistor. They do not instantaneously vaporize
> > under adverse load conditions.
>
> They most certainly do. Thermal concerns from excessive
> dissipation are almost never the failure mode. Voltage breakdown
> is immediately catastrophic.
[Dave D'Epagnier]
Again, if voltage breakdown is the problem, then all bets are off.
The wrong device is being used and there isn't a whole lot that can be done
for protection. If voltage breakdown were that big of an issue in amateur
gear almost everyone with an exciter would have blown it up by now due to
load mismatches or accidentally transmitting with no load connected at all
(improper antenna selected etc).
> Therefore simultaneous detection of forward
> > and reflected power is unnecessary, and a single A/D converter is
> > typically used.
>
> You MUST use a directional coupler, and you should always detect
> reflected power as well as forward power. Forward power sets the
> drive limits, and reflected power sets the SWR limit.
[Dave D'Epagnier] Yes, you need a directional coupler, but
simultaneous (instantaneous) sampling of forward and reflected power is not
necessary.
> > time of the amplifier is this quick if that's the method used for
> shutdown
> > under high SWR conditions. Severe load mismatches can cause the
> > transistors to see higher voltages, but typically not as high as the
> > breakdown rating of the device.
>
> Not so. They can easily produce such voltages unless the voltage
> breakdown of the device is several times the operating voltage.
> While commercial equipment often has enough headroom, amateur
> gear doesn't.
[Dave D'Epagnier] In that context, I guess you could say I design
commercial grade equipment at home. Take a look at Advanced Power
Technologies RF transistors. Every one of their RF power transistors has a
breakdown rating of 3x the recommended operating line voltage.
> One of the cures for the common PA failures in Drake TR5
> transceivers was to use a higher voltage output device.
>
> Have a look at Motorola's data, and you will see transistors are
> typically recommended to be used with supply voltages just less
> than half of the working voltage. In normal operation, that device
> would be coming close to peak voltage rating on every cycle at
> maximum output. If the collector impedance increases, the voltage
> can easily exceed twice the supply rail.
[Dave D'Epagnier] Transistor output impedances are generally much
less than 50 ohms. The output matching network steps this impedance up to 50
ohms. As a result the [Dave D'Epagnier] voltage on the collector (or
drain) of the transistor is much less than voltage at the 50 ohm load during
normal operation. The transistor should not even be close to breakdown under
this condition.
> I've measured voltages of 100 volts on the drains of FET's operated
> at 12 volts, under some conditions of load mismatch. If the device
> have a breakdown voltage of less than that, it will instantly fail. Just
> one RF cycle over the peak voltage breakdown will do it.
>
>
>
> 73, Tom W8JI
> w8ji@contesting.com
------_=_NextPart_001_01C05E2B.352509F0
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Dus-ascii">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2650.12">
<TITLE>RE: [AMPS] SWR Protection</TITLE>
</HEAD>
<BODY>
<BR>
<BR>
<BR>
<UL>
<P><FONT SIZE=3D2 FACE=3D"Arial">However, transistors can fail very =
rapidly from excessive voltage. </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">You have zero-time to prevent that =
failure, since it is a punch-</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">through failure. Temperature really =
isn't the major concern. Excess </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">voltage is, and exciters with keying =
transients (common today) can </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">really hammer the transistors with =
collector or drain voltage spikes. </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">That's another reason why the optimum =
load impedance is </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">generally set to be higher than the =
working impedance. </FONT>
</P>
<P><B><I><FONT COLOR=3D"#0000FF" SIZE=3D2 FACE=3D"Arial">[Dave =
D'Epagnier]</FONT></I></B><I></I> <FONT COLOR=3D"#0000FF" SIZE=3D2 =
FACE=3D"Arial"> Tom, I know of at least one prominent RF transistor =
manufacturer that would take issue with your statements about =
temperature above. My experience as well shows that high temperature is =
almost always the cause of failure in a properly designed solid state =
amp. This high temp is usually cause by load mismatch. If voltage =
breakdown is a problem then improper devices are being used and it's a =
poor design to begin with.</FONT></P>
<P><FONT SIZE=3D2 FACE=3D"Arial"> > a transistor you will see =
that the die has a typical response of 10ms or</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">> so. So if you can react quicker =
than this (say in about 1ms) then you can</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">> adequately protect the =
transistor. They do not instantaneously vaporize</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">> under adverse load conditions. =
</FONT>
</P>
<P><FONT SIZE=3D2 FACE=3D"Arial">They most certainly do. Thermal =
concerns from excessive </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">dissipation are almost never the =
failure mode. Voltage breakdown </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">is immediately catastrophic.</FONT>
<BR><B><I><FONT COLOR=3D"#0000FF" SIZE=3D2 FACE=3D"Arial">[Dave =
D'Epagnier]</FONT></I></B><I></I> <FONT COLOR=3D"#0000FF" SIZE=3D2 =
FACE=3D"Arial"> </FONT>
<BR><FONT COLOR=3D"#0000FF" SIZE=3D2 FACE=3D"Arial">Again, if voltage =
breakdown is the problem, then all bets are off. The wrong device is =
being used and there isn't a whole lot that can be done for protection. =
If voltage breakdown were that big of an issue in amateur gear almost =
everyone with an exciter would have blown it up by now due to load =
mismatches or accidentally transmitting with no load connected at all =
(improper antenna selected etc).</FONT> </P>
<P><FONT SIZE=3D2 FACE=3D"Arial">Therefore simultaneous detection of =
forward</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">> and reflected power is =
unnecessary, and a single A/D converter is</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">> typically used. </FONT>
</P>
<P><FONT SIZE=3D2 FACE=3D"Arial">You MUST use a directional coupler, =
and you should always detect </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">reflected power as well as forward =
power. Forward power sets the </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">drive limits, and reflected power =
sets the SWR limit. </FONT>
<BR><B><I><FONT COLOR=3D"#0000FF" SIZE=3D2 FACE=3D"Arial">[Dave =
D'Epagnier]</FONT></I></B><I></I> <FONT COLOR=3D"#0000FF" SIZE=3D2 =
FACE=3D"Arial"> Yes, you need a directional coupler, but simultaneous =
(instantaneous) sampling of forward and reflected power is not =
necessary.</FONT></P>
<P><FONT SIZE=3D2 FACE=3D"Arial">> time of the amplifier is this =
quick if that's the method used for shutdown</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">> under high SWR conditions. =
Severe load mismatches can cause the</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">> transistors to see higher =
voltages, but typically not as high as the</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">> breakdown rating of the device. =
</FONT>
</P>
<P><FONT SIZE=3D2 FACE=3D"Arial">Not so. They can easily produce such =
voltages unless the voltage </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">breakdown of the device is several =
times the operating voltage. </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">While commercial equipment often has =
enough headroom, amateur </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">gear doesn't.</FONT>
<BR><B><I><FONT COLOR=3D"#0000FF" SIZE=3D2 FACE=3D"Arial">[Dave =
D'Epagnier]</FONT></I></B><I></I> <FONT COLOR=3D"#0000FF" SIZE=3D2 =
FACE=3D"Arial"> In that context, I guess you could say I design =
commercial grade equipment at home. Take a look at Advanced Power =
Technologies RF transistors. Every one of their RF power transistors =
has a breakdown rating of 3x the recommended operating line =
voltage.</FONT></P>
<P><FONT SIZE=3D2 FACE=3D"Arial">One of the cures for the common PA =
failures in Drake TR5 </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">transceivers was to use a higher =
voltage output device.</FONT>
</P>
<P><FONT SIZE=3D2 FACE=3D"Arial">Have a look at Motorola's data, and =
you will see transistors are </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">typically recommended to be used with =
supply voltages just less </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">than half of the working voltage. In =
normal operation, that device </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">would be coming close to peak voltage =
rating on every cycle at </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">maximum output. If the collector =
impedance increases, the voltage </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">can easily exceed twice the supply =
rail.</FONT>
<BR><B><I><FONT COLOR=3D"#0000FF" SIZE=3D2 FACE=3D"Arial">[Dave =
D'Epagnier]</FONT></I></B><I></I> <FONT COLOR=3D"#0000FF" SIZE=3D2 =
FACE=3D"Arial"> Transistor output impedances are generally much less =
than 50 ohms. The output matching network steps this impedance up to 50 =
ohms. As a result the </FONT><B></B><B><I> <FONT COLOR=3D"#0000FF" =
SIZE=3D2 FACE=3D"Arial">[Dave =
D'Epagnier]</FONT></I></B><I></I> <FONT COLOR=3D"#0000FF" SIZE=3D2 =
FACE=3D"Arial"> voltage on the collector (or drain) of the transistor =
is much less than voltage at the 50 ohm load during normal operation. =
The transistor should not even be close to breakdown under this =
condition.</FONT></P>
<P><FONT COLOR=3D"#0000FF" SIZE=3D2 FACE=3D"Arial"></FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">I've measured voltages of 100 volts =
on the drains of FET's operated </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">at 12 volts, under some conditions of =
load mismatch. If the device </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">have a breakdown voltage of less than =
that, it will instantly fail. Just </FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">one RF cycle over the peak voltage =
breakdown will do it.</FONT>
</P>
<BR>
<BR>
<P><FONT SIZE=3D2 FACE=3D"Arial">73, Tom W8JI</FONT>
<BR><FONT SIZE=3D2 FACE=3D"Arial">w8ji@contesting.com</FONT>
</P>
</UL>
</BODY>
</HTML>
------_=_NextPart_001_01C05E2B.352509F0--
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|