R: [AMPS] fine tuning amp pi-network
Maurizio Panicara
Maurizio Panicara" <i4jmy@iol.it
Mon, 28 Jun 1999 11:57:16 +0200
This is a multi-part message in MIME format.
------=_NextPart_000_005D_01BEC15D.5DB95960
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
>Anyway, in optimizing the L for the output pi-net what are the =
recommended
>methods? I see not much difference btwn 1.8 and 1 uH of L on the output
Within a certain loaded Q range thare are not so much differences but =
output=20
power is not the only parameter to check and deal width before going "on =
the air"=20
or saying that the amp came out well.
>Anyway, just wondered how some of you HB'ers have optimized your tank
>circuits for each desired band.
>
>So how do we optimize the tank guys? I do have a lab quality grid =
dipper as
>well as the MFJ 259 SWR Analyzer.
>
>Lane
>Ku7i
>San Diego
I don't like the cut and try procedure, better if the amp closely match =
the projected parameters.=20
All considered, it's faster to calculate and verify than making a lot of =
blind changes.
For a PI network, this is my bench method with amplifier OFF:
1) Keep connections short to avoid stray inductances, and coil far from =
box walls=20
to preserve Q.
2) If using ceramic tube (expecially when paralleled) stay with them =
far from box =20
walls to avoid unwanted extra plate capacitance. The residual =
(that sums to the
Tune capacitor one) could be too high and consequentely only high =
Q PI can =20
apply to 10 and 15m. The inherent problems in recovering a good =
efficiency=20
from a very Hi Q PI network can't be easily solved, so, better to =
avoid it.
3) Calculate Rin of your amplifier (known simplified formulas, with =
plate voltage, =20
current drain and amplification class are on many books) and =
realize a load=20
resistor (dummy) equal to Rin.
4) Estabilish a reasonable loaded Q value for the PI network (9 -12 can =
be OK in=20
most of the cases).
5) Calculate your PI (symple formulas, thousand of normalized tables, =
or even final=20
values are available on many books ).=20
In other words find out the coil and capacitors values associated =
with selected Q=20
and Rin (Rout supposed to be the standard 50 Ohm) of your amp.
6) Verify that found values can be effectively obtained by available =
components.
7) Realize a coil whose Q is considerably larger than loaded Q (this =
implies a lot=20
of care in coil positioning too) and having a size proportioned to =
the current flow.
8) Place the resistor equivalent to Rin, across PI input (tube in the =
circuit but without
plate choke connected).
9) Tune plate and load capacitors checking, from the output of PI (ant. =
socket, close
or by-pass the antenna relay somehow) that the return loss (match) =
is extremely=20
good and easy to obtain.
10)Connect the plate choke and verify that the return loss doesn't =
change or that
there is not to move capacitors to obtain it again. (if the last =
happen, the choke
has not enough reactance to leave PI unaffected, or self resonates =
too close to=20
the used frequency. Change the plate choke or its winding and =
retry !).=20
11)If using a spectrum analyzer with tracking generator, like I do, =
check also that
plate choke has no resonances (test to be performed band by band) =
along
all the amp specrtum. =20
=20
In amplifiers having good design and potentially good =
characteristics, not the
PI network but a wrong plate choke is the responsable of poor =
performances=20
(in one or more bands) or even the reason of main failures.
10)Store somewere Tune and Load numbers, and later, checking the =
amplifier
ON in a dummy load (at designed output power level), verify that =
they are=20
close to those found and recoreded at point 5.=20
This will confirm that everything is working correctly as planned.
This is not the only possible way to proceed, but I hope it helps Lane =
and someone else to spare time and to get success.
=20
73,
Mauri I4JMY
i4jmy@iol.it
=20
------=_NextPart_000_005D_01BEC15D.5DB95960
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN">
<HTML>
<HEAD>
<META content=3Dtext/html;charset=3Diso-8859-1 =
http-equiv=3DContent-Type><!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 =
HTML//EN"><!DOCTYPE HTML PUBLIC "-//W3C//DTD W3 HTML//EN"><!DOCTYPE HTML =
PUBLIC "-//W3C//DTD W3 HTML//EN">
<META content=3D'"MSHTML 4.72.3110.7"' name=3DGENERATOR>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV>>Anyway, in optimizing the L for the output pi-net what are the=20
recommended<BR>>methods? I see not much difference btwn 1.8 and 1 uH =
of L on=20
the output<BR></DIV>
<DIV><BR>Within a certain loaded Q range thare are not so much =
differences but=20
output<FONT color=3D#000000 size=3D2> </FONT></DIV>
<DIV><FONT color=3D#000000 size=3D2></FONT>power is not the only =
parameter to check=20
and deal width before going "on the air" </DIV>
<DIV>or saying that the amp came out well.<BR><BR>>Anyway, just =
wondered how=20
some of you HB'ers have optimized your tank<BR>>circuits for each =
desired=20
band.</DIV>
<DIV>><BR>>So how do we optimize the tank guys? I do have a lab =
quality=20
grid dipper as<BR>>well as the MFJ 259 SWR=20
Analyzer.<BR>><BR>>Lane<BR>>Ku7i<BR>>San Diego<BR></DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV><FONT color=3D#000000><FONT size=3D3>I don't like the cut and try =
procedure,=20
better if the amp closely match the projected=20
parameters. </FONT></FONT></DIV>
<DIV><FONT color=3D#000000><FONT size=3D3><U>All considered, it's faster =
to=20
calculate and verify than making a lot of blind =
changes.</FONT></FONT><FONT=20
size=3D3><BR></FONT></U><BR></DIV>
<DIV> </DIV>
<DIV>For a PI network, this is my bench method with amplifier=20
OFF:<BR><BR>1) Keep connections short to avoid stray =
inductances<FONT=20
color=3D#000000 size=3D2>,</FONT> and coil far from box walls </DIV>
<DIV> to preserve Q.<BR>2) If using =
ceramic=20
tube (expecially when paralleled) stay with them far from=20
box </DIV>
<DIV> walls to<FONT color=3D#000000 =
size=3D2>=20
</FONT>avoid unwanted extra plate capacitance<FONT color=3D#000000><FONT =
size=3D3>.</FONT></FONT><FONT size=3D3> The</FONT> residual (that sums =
to the</DIV>
<DIV> Tune capacitor one) could be too =
high<FONT=20
color=3D#000000 size=3D2> </FONT>and consequentely only high Q =
PI <FONT=20
color=3D#000000 face=3D"" size=3D3> </FONT>can </DIV>
<DIV> apply to 10 and 15m. The =
inherent=20
<FONT size=3D3>problems <FONT color=3D#000000>in recovering a =
</FONT></FONT><FONT=20
size=3D3><FONT color=3D#000000>good</FONT></FONT><FONT size=3D3><FONT =
color=3D#000000>=20
</FONT></FONT><FONT size=3D3><FONT=20
color=3D#000000>efficiency </FONT></FONT></DIV>
<DIV><FONT size=3D3><FONT color=3D#000000> =
from a very=20
Hi Q<FONT color=3D#000000> PI network </FONT>can't be easily solved, so, =
better to=20
avoid it.</FONT></FONT><BR>3) Calculate Rin of your amplifier =
(known=20
simplified formulas, with plate voltage, </DIV>
<DIV> current<FONT color=3D#000000 =
size=3D2>=20
</FONT>drain and amplification class are on many books) and realize a =
load=20
</DIV>
<DIV> resistor (dummy) equal to =
Rin.<BR>4) =20
Estabilish a reasonable loaded Q value for the PI network (9 -12 can be =
OK in=20
</DIV>
<DIV> most of the cases).<BR>5) =
Calculate=20
your PI (symple formulas, thousand of normalized tables, or even final =
</DIV>
<DIV> values are available on many books =
). </DIV>
<DIV> In other words find out the coil and =
capacitors values associated with<FONT color=3D#000000 size=3D2> =
</FONT>selected Q=20
</DIV>
<DIV> and Rin (Rout supposed to be the =
standard 50=20
Ohm) of your amp.</DIV>
<DIV>6) Verify that found values can be effectively obtained by =
available=20
components.<BR>7) Realize a coil whose Q is =
considerably<U><STRONG> larger=20
</STRONG></U>than loaded Q (this implies a lot<FONT color=3D#000000 =
size=3D2>=20
</FONT></DIV>
<DIV><FONT color=3D#000000 size=3D2></FONT><FONT =
size=3D2> =20
</FONT>of care in coil positioning too)<FONT size=3D2> <FONT =
size=3D3>and having a=20
size proportioned to the current flow.</FONT></FONT><FONT=20
size=3D3><BR></FONT>8) Place the resistor equivalent to Rin, =
across PI input=20
(tube in the circuit<FONT color=3D#000000 size=3D2> </FONT>but=20
without<BR> plate choke connected)<FONT=20
color=3D#000000 size=3D2>.</FONT><BR>9) Tune plate and load =
capacitors=20
checking, from the output of PI (ant.<FONT color=3D#000000 size=3D2> =
</FONT>socket,=20
close</DIV>
<DIV> or by-pass the antenna relay =
somehow) that=20
the return loss (match) is extremely </DIV>
<DIV> good and easy to obtain.</DIV>
<DIV>10)Connect the plate choke and verify that the return loss doesn't =
change=20
or<FONT color=3D#000000 size=3D2> =
</FONT>that<BR> =20
there is not to move capacitors to obtain it again. (if the last<FONT=20
color=3D#000000 size=3D2> </FONT>happen, the =
choke<BR> =20
has not enough reactance to leave PI unaffected, or self resonates too =
close to=20
</DIV>
<DIV> the used<FONT color=3D#000000 =
size=3D2>=20
</FONT>frequency. Change the plate choke or its winding and retry !). =
<BR>11)If=20
using a spectrum analyzer with tracking generator, like I do, check<FONT =
color=3D#000000 size=3D2> </FONT>also =
that<BR> plate=20
choke has no resonances (test to be performed band by band) along</DIV>
<DIV> all the amp =
specrtum. </DIV>
<DIV><STRONG> </STRONG></DIV>
<DIV><STRONG></STRONG><STRONG> </STRONG>In =
<FONT=20
size=3D3><FONT color=3D#000000>amplifiers having</FONT></FONT><FONT =
size=3D3><FONT=20
color=3D#000000> good design and potentially </FONT></FONT><FONT =
size=3D3>good=20
</FONT><FONT size=3D3>characteristics, not the</FONT></DIV>
<DIV><FONT size=3D3><FONT color=3D#000000> =
PI network=20
but a</FONT></FONT><FONT size=3D3><FONT color=3D#000000><FONT size=3D2>=20
</FONT></FONT></FONT>wrong plate choke is the responsable of poor =
performances=20
</DIV>
<DIV> (in one or more bands) or even the =
reason of=20
main failures.<BR></DIV>
<DIV> </DIV>
<DIV>10)Store somewere Tune and Load numbers, and later, checking the=20
amplifier<BR> ON in a <U><STRONG>dummy =
load=20
</STRONG></U>(at designed output power level), verify that they are<FONT =
color=3D#000000 size=3D2> </FONT></DIV>
<DIV><FONT color=3D#000000 size=3D2></FONT><FONT=20
size=3D2> </FONT>close to those found and =
recoreded=20
at point 5. </DIV>
<DIV> This will confirm that everything is =
<FONT=20
color=3D#000000 size=3D2><FONT size=3D3>working correctly as=20
planned.</FONT></FONT><FONT size=3D3><BR></FONT></DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV><FONT color=3D#000000><FONT size=3D3>This is not the only possible =
way to=20
proceed, but I hope it helps Lane and someone else to =
</FONT></FONT><FONT=20
color=3D#000000><FONT size=3D3>spare time and to get =
success.</FONT></FONT><FONT=20
size=3D3></FONT></DIV>
<DIV><FONT color=3D#000000 size=3D2></FONT> </DIV>
<DIV>73,</DIV>
<DIV>Mauri I4JMY</DIV>
<DIV> </DIV>
<DIV><FONT color=3D#000000 size=3D2><A=20
href=3D"mailto:i4jmy@iol.it">i4jmy@iol.it</A></FONT></DIV>
<DIV><FONT color=3D#000000 size=3D2></FONT> </DIV></BODY></HTML>
------=_NextPart_000_005D_01BEC15D.5DB95960--
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm