Uwe wrote in message ...
in article , N2EY at
PAMNO wrote on 4/5/04 18:59:

In article , Uwe
writes:

I will just have to fiddle a bit more
with the pi network (since at the B+ voltages suggested here my plate
current would be way high) and I will have to live with the chirp.

Are you getting a "dip" in plate current? If not, the coil is probably too
large or too small. Unless you get a real dip, the output network isn;t right.

I have used a very similar transmitter with 350 volts on the plate, and the
dip
is clean and pronounced.


Jim, the original docs I got for this tx call, at 40 m, for a 15 turn coil
on the coil form provided with the kit, which I hear was 1.25" diameter.
If I use the formula for air coils this turns out to be roughly a 22
microhenry coil.

22 microhenries? I get more like 8 microhenries using the formula

L = (a * a * n * n)/([9 * a] + [10 * b])

where
a = radius of coil in inches
b = length of winding in inches
n = number of turns

The coil which works best with my tx is 8 turns on a 1,125" ceramic core.

But do you get a dip?

Be aware that the AC-1 went through some changes in its lifetime. Some
models used a filter choke, others did not. Some used a 730 uuf
loading capacitor, others just a single-section 365 uuf one. Coils
changed too.

To get guess work out of it I just bought and built a L/C meter and
measured my coil to have 2.7 microhenry. So I am way off, but it works, sort
of.

The air caps are 36 to 420pf at the plate and 15 to 728pf at the antenna, so
that seem right.

All this happens with B+200V and 35 mA plate current.


LC = 25,330/(f * f)

so for 7 MHz, the LC constant is 516. Your 2.7 uH coil should resonate
with 191 uuf.

Older ARRL handbooks give typical values for pi network for 50 Ohm antenna
loads and my values are in range for the caps but my coil is too small.

The ouput voltage on my antenna measured with a scope is up to 75 volts peak
to peak, with a 50 Ohm load that would mean I get more out of the tx than I
put into it and I am not of the sort who says this might happen.

So my conclusion is, and tell me if this sounds right, that I have an
antenna which is far from 50 ohm resistive at 40m and that that makes
everything weird.

That's defintitely part of the problem. What antenna are you using?
Have you tried a resistor or lamp load?

The dips in plate current are nearly imperceptible and they are not aided by
my 250mA full scale meter. They may be 2 or 3 mA.
I tune with the help of my scope.

The meter tells more. You can use a pilot light (#47, 150 mA) instead
of a meter.

Sudden thought: Where is the meter connected? Are you reading plate
current, or plate-and-screen current combined?

Here's something else to try:

Often trouble of this sort is due to the RF choke used. What RFCs are
you suing, particularly in the plate circuit? Although the LC meter
may say they are a certain L, in real life they may have all sorts of
unwanted resonances.

To test this idea out, do the following:

- Remove the plate RFC
- Connect the antenna end of the plate coil to the B+ where the RFC
used to be connected. This point should already be bypassed to ground
through a disk capacitor of about .01 uF
- Disconnect the "loading" capacitor
- Remove the plate coupling capacitor.

What you will then have is the 200 volts being fed to the plate
through the coil, with one end of the coil going to the plate supply
and the other end connected directly to the plate of the 6V6. The
plate tuning capacitor is connected between the plate of the 6V6 and
ground.

End result is no plate RFC and a parallel resonant circuit. There's no
connection for an antenna yet, but that's not important right now.

Test out the rig and look for the plate current dip. It should be very
obvious because there is no load connected.

This is just a temporary setup to see if the RFC is OK.

73 es GL de Jim, N2EY

"N2EY" wrote in message
om...

This is just a temporary setup to see if the RFC is OK.

My recollection is hazy but I seem to recall that when the loading cap was
open too far for the load the pi net was seeing, the dip got very shallow.
I'll bet his antenna is outside the range it can match. Time for Uwe to
gather up some of that coil-winding stuff and make a tuner.

73,

"PM"

in article , N2EY at
wrote on 4/6/04 12:29:

Uwe wrote in message
...
in article , N2EY at
PAMNO wrote on 4/5/04 18:59:

In article , Uwe

writes:

I will just have to fiddle a bit more
with the pi network (since at the B+ voltages suggested here my plate
current would be way high) and I will have to live with the chirp.

Are you getting a "dip" in plate current? If not, the coil is probably too
large or too small. Unless you get a real dip, the output network isn;t
right.

I have used a very similar transmitter with 350 volts on the plate, and the
dip
is clean and pronounced.


Jim, the original docs I got for this tx call, at 40 m, for a 15 turn coil
on the coil form provided with the kit, which I hear was 1.25" diameter.
If I use the formula for air coils this turns out to be roughly a 22
microhenry coil.

22 microhenries? I get more like 8 microhenries using the formula

L = (a * a * n * n)/([9 * a] + [10 * b])

where
a = radius of coil in inches
b = length of winding in inches
n = number of turns



Well, the way I use the formula is 1.25*1.25*15*15/((9*1.25)+(10*0.45)) =
351.5/15.75 = 22.3

O.45 is the length of the 15 windings.

Do I not use the formula properly??



My meter is built into my bench power supply (thats why it reads up to
250mA), so I am measuring plate and screen current.

I put in a second meter which would only measure the plate current but its
reading is practicly identical to the first one, as if there was no grid
current.

My antenna is a dipole of about 75ft. length each side, connected with a 50
ohm coax, no balun or such things.


I will need a few days to try out some of the things you and also Paul
suggested and it might really be a good idea to get an SWR meter and a
tuner.

All in due time and I will surely get back to you.

Thanks for the help


Uwe


The coil which works best with my tx is 8 turns on a 1,125" ceramic core.

But do you get a dip?

Be aware that the AC-1 went through some changes in its lifetime. Some
models used a filter choke, others did not. Some used a 730 uuf
loading capacitor, others just a single-section 365 uuf one. Coils
changed too.

To get guess work out of it I just bought and built a L/C meter and
measured my coil to have 2.7 microhenry. So I am way off, but it works, sort
of.

The air caps are 36 to 420pf at the plate and 15 to 728pf at the antenna, so
that seem right.

All this happens with B+200V and 35 mA plate current.


LC = 25,330/(f * f)

so for 7 MHz, the LC constant is 516. Your 2.7 uH coil should resonate
with 191 uuf.

Older ARRL handbooks give typical values for pi network for 50 Ohm antenna
loads and my values are in range for the caps but my coil is too small.

The ouput voltage on my antenna measured with a scope is up to 75 volts peak
to peak, with a 50 Ohm load that would mean I get more out of the tx than I
put into it and I am not of the sort who says this might happen.

So my conclusion is, and tell me if this sounds right, that I have an
antenna which is far from 50 ohm resistive at 40m and that that makes
everything weird.

That's defintitely part of the problem. What antenna are you using?
Have you tried a resistor or lamp load?

The dips in plate current are nearly imperceptible and they are not aided by
my 250mA full scale meter. They may be 2 or 3 mA.
I tune with the help of my scope.

The meter tells more. You can use a pilot light (#47, 150 mA) instead
of a meter.

Sudden thought: Where is the meter connected? Are you reading plate
current, or plate-and-screen current combined?

Here's something else to try:

Often trouble of this sort is due to the RF choke used. What RFCs are
you suing, particularly in the plate circuit? Although the LC meter
may say they are a certain L, in real life they may have all sorts of
unwanted resonances.

To test this idea out, do the following:

- Remove the plate RFC
- Connect the antenna end of the plate coil to the B+ where the RFC
used to be connected. This point should already be bypassed to ground
through a disk capacitor of about .01 uF
- Disconnect the "loading" capacitor
- Remove the plate coupling capacitor.

What you will then have is the 200 volts being fed to the plate
through the coil, with one end of the coil going to the plate supply
and the other end connected directly to the plate of the 6V6. The
plate tuning capacitor is connected between the plate of the 6V6 and
ground.

End result is no plate RFC and a parallel resonant circuit. There's no
connection for an antenna yet, but that's not important right now.

Test out the rig and look for the plate current dip. It should be very
obvious because there is no load connected.

This is just a temporary setup to see if the RFC is OK.

73 es GL de Jim, N2EY

In article , Uwe
writes:

in article , N2EY at
wrote on 4/6/04 12:29:

Uwe wrote in message
...
in article , N2EY at
PAMNO wrote on 4/5/04 18:59:

In article , Uwe

writes:

I will just have to fiddle a bit more
with the pi network (since at the B+ voltages suggested here my plate
current would be way high) and I will have to live with the chirp.

Are you getting a "dip" in plate current? If not, the coil is probably
too
large or too small. Unless you get a real dip, the output network isn;t
right.

I have used a very similar transmitter with 350 volts on the plate, and
the
dip
is clean and pronounced.


Jim, the original docs I got for this tx call, at 40 m, for a 15 turn coil
on the coil form provided with the kit, which I hear was 1.25" diameter.
If I use the formula for air coils this turns out to be roughly a 22
microhenry coil.

22 microhenries? I get more like 8 microhenries using the formula

L = (a * a * n * n)/([9 * a] + [10 * b])

where
a = radius of coil in inches
b = length of winding in inches
n = number of turns



Well, the way I use the formula is 1.25*1.25*15*15/((9*1.25)+(10*0.45)) =
351.5/15.75 = 22.3

O.45 is the length of the 15 windings.

Do I not use the formula properly??

You used the coil *diameter* where you should have used the coil *radius*. A
coil with diameter of 1.25 inch has a radius of 0.625 inch.

Compute

0.625*0.625*15*15/((9*0.625)+(10*0.45)) =

and see what you get.

My meter is built into my bench power supply (thats why it reads up to
250mA), so I am measuring plate and screen current.

I put in a second meter which would only measure the plate current but its
reading is practicly identical to the first one, as if there was no grid
current.

That's odd.

My antenna is a dipole of about 75ft. length each side, connected with a 50
ohm coax, no balun or such things.

150 feet total length? That's not resonant on 40 meters, and your SWR with 50
ohm coax is probably quite high.

A half-wave 40 meter dipole is about 66-67 feet long (33 feet each side), and
will have a fairly low SWR on 40 meters when fed with 50 ohm coax. The next
length that will give a fairly low 40 meter SWR is about 205 feet overall (102
feet each side). Such a dipole is one-and-a-half waves long.

These are "ballpark" figures, not exact ones.

How high is your dipole?

I agree with Paul Morphy that a simple dummy load is best for testing. His
suggestion of paralleled noninductive resistors is excellent.

I will need a few days to try out some of the things you and also Paul
suggested and it might really be a good idea to get an SWR meter and a
tuner.

That will work, but first get the rig working correctly into a dummy load.

All in due time and I will surely get back to you.

If it takes me a while to respond, it's because I'm away from the computer.

Thanks for the help

You're welcome.

73 de Jim, N2EY

In article , Uwe
writes:

in article , N2EY at
wrote on 4/6/04 12:29:

Uwe wrote in message
...
in article , N2EY at
PAMNO wrote on 4/5/04 18:59:

In article , Uwe

writes:

I will just have to fiddle a bit more
with the pi network (since at the B+ voltages suggested here my plate
current would be way high) and I will have to live with the chirp.

Are you getting a "dip" in plate current? If not, the coil is probably
too
large or too small. Unless you get a real dip, the output network isn;t
right.

I have used a very similar transmitter with 350 volts on the plate, and
the
dip
is clean and pronounced.


Jim, the original docs I got for this tx call, at 40 m, for a 15 turn coil
on the coil form provided with the kit, which I hear was 1.25" diameter.
If I use the formula for air coils this turns out to be roughly a 22
microhenry coil.

22 microhenries? I get more like 8 microhenries using the formula

L = (a * a * n * n)/([9 * a] + [10 * b])

where
a = radius of coil in inches
b = length of winding in inches
n = number of turns



Well, the way I use the formula is 1.25*1.25*15*15/((9*1.25)+(10*0.45)) =
351.5/15.75 = 22.3

O.45 is the length of the 15 windings.

Do I not use the formula properly??

You used the coil *diameter* where you should have used the coil *radius*. A
coil with diameter of 1.25 inch has a radius of 0.625 inch.

Compute

0.625*0.625*15*15/((9*0.625)+(10*0.45)) =

and see what you get.

My meter is built into my bench power supply (thats why it reads up to
250mA), so I am measuring plate and screen current.

I put in a second meter which would only measure the plate current but its
reading is practicly identical to the first one, as if there was no grid
current.

That's odd.

My antenna is a dipole of about 75ft. length each side, connected with a 50
ohm coax, no balun or such things.

150 feet total length? That's not resonant on 40 meters, and your SWR with 50
ohm coax is probably quite high.

A half-wave 40 meter dipole is about 66-67 feet long (33 feet each side), and
will have a fairly low SWR on 40 meters when fed with 50 ohm coax. The next
length that will give a fairly low 40 meter SWR is about 205 feet overall (102
feet each side). Such a dipole is one-and-a-half waves long.

These are "ballpark" figures, not exact ones.

How high is your dipole?

I agree with Paul Morphy that a simple dummy load is best for testing. His
suggestion of paralleled noninductive resistors is excellent.

I will need a few days to try out some of the things you and also Paul
suggested and it might really be a good idea to get an SWR meter and a
tuner.

That will work, but first get the rig working correctly into a dummy load.

All in due time and I will surely get back to you.

If it takes me a while to respond, it's because I'm away from the computer.

Thanks for the help

You're welcome.

73 de Jim, N2EY

"N2EY" wrote in message
om...

This is just a temporary setup to see if the RFC is OK.

My recollection is hazy but I seem to recall that when the loading cap was
open too far for the load the pi net was seeing, the dip got very shallow.
I'll bet his antenna is outside the range it can match. Time for Uwe to
gather up some of that coil-winding stuff and make a tuner.

73,

"PM"

in article , N2EY at
wrote on 4/6/04 12:29:

Uwe wrote in message
...
in article , N2EY at
PAMNO wrote on 4/5/04 18:59:

In article , Uwe

writes:

I will just have to fiddle a bit more
with the pi network (since at the B+ voltages suggested here my plate
current would be way high) and I will have to live with the chirp.

Are you getting a "dip" in plate current? If not, the coil is probably too
large or too small. Unless you get a real dip, the output network isn;t
right.

I have used a very similar transmitter with 350 volts on the plate, and the
dip
is clean and pronounced.


Jim, the original docs I got for this tx call, at 40 m, for a 15 turn coil
on the coil form provided with the kit, which I hear was 1.25" diameter.
If I use the formula for air coils this turns out to be roughly a 22
microhenry coil.

22 microhenries? I get more like 8 microhenries using the formula

L = (a * a * n * n)/([9 * a] + [10 * b])

where
a = radius of coil in inches
b = length of winding in inches
n = number of turns



Well, the way I use the formula is 1.25*1.25*15*15/((9*1.25)+(10*0.45)) =
351.5/15.75 = 22.3

O.45 is the length of the 15 windings.

Do I not use the formula properly??



My meter is built into my bench power supply (thats why it reads up to
250mA), so I am measuring plate and screen current.

I put in a second meter which would only measure the plate current but its
reading is practicly identical to the first one, as if there was no grid
current.

My antenna is a dipole of about 75ft. length each side, connected with a 50
ohm coax, no balun or such things.


I will need a few days to try out some of the things you and also Paul
suggested and it might really be a good idea to get an SWR meter and a
tuner.

All in due time and I will surely get back to you.

Thanks for the help


Uwe


The coil which works best with my tx is 8 turns on a 1,125" ceramic core.

But do you get a dip?

Be aware that the AC-1 went through some changes in its lifetime. Some
models used a filter choke, others did not. Some used a 730 uuf
loading capacitor, others just a single-section 365 uuf one. Coils
changed too.

To get guess work out of it I just bought and built a L/C meter and
measured my coil to have 2.7 microhenry. So I am way off, but it works, sort
of.

The air caps are 36 to 420pf at the plate and 15 to 728pf at the antenna, so
that seem right.

All this happens with B+200V and 35 mA plate current.


LC = 25,330/(f * f)

so for 7 MHz, the LC constant is 516. Your 2.7 uH coil should resonate
with 191 uuf.

Older ARRL handbooks give typical values for pi network for 50 Ohm antenna
loads and my values are in range for the caps but my coil is too small.

The ouput voltage on my antenna measured with a scope is up to 75 volts peak
to peak, with a 50 Ohm load that would mean I get more out of the tx than I
put into it and I am not of the sort who says this might happen.

So my conclusion is, and tell me if this sounds right, that I have an
antenna which is far from 50 ohm resistive at 40m and that that makes
everything weird.

That's defintitely part of the problem. What antenna are you using?
Have you tried a resistor or lamp load?

The dips in plate current are nearly imperceptible and they are not aided by
my 250mA full scale meter. They may be 2 or 3 mA.
I tune with the help of my scope.

The meter tells more. You can use a pilot light (#47, 150 mA) instead
of a meter.

Sudden thought: Where is the meter connected? Are you reading plate
current, or plate-and-screen current combined?

Here's something else to try:

Often trouble of this sort is due to the RF choke used. What RFCs are
you suing, particularly in the plate circuit? Although the LC meter
may say they are a certain L, in real life they may have all sorts of
unwanted resonances.

To test this idea out, do the following:

- Remove the plate RFC
- Connect the antenna end of the plate coil to the B+ where the RFC
used to be connected. This point should already be bypassed to ground
through a disk capacitor of about .01 uF
- Disconnect the "loading" capacitor
- Remove the plate coupling capacitor.

What you will then have is the 200 volts being fed to the plate
through the coil, with one end of the coil going to the plate supply
and the other end connected directly to the plate of the 6V6. The
plate tuning capacitor is connected between the plate of the 6V6 and
ground.

End result is no plate RFC and a parallel resonant circuit. There's no
connection for an antenna yet, but that's not important right now.

Test out the rig and look for the plate current dip. It should be very
obvious because there is no load connected.

This is just a temporary setup to see if the RFC is OK.

73 es GL de Jim, N2EY

Jim, after my calculation of the coil were wrong I thought it was about time
to check everything and I did and to try and distinguish between radius and
diameter...

Using the L/C meter I wound a proper coil, I checked the calibration of my
plate current meter, I did a more thorough check of the grid current (it is
between 1 and 2 mA) and so on and so forth.
And I did connect a dummy load (even though they don't respond or send out
QSL cards when you tranmit into them).

None of the thing did make any real difference and the dip, the elusive dip,
was in the order of magnitude of maybe 2 mA, nearly impossible to see on my
meter.

Then I changed the circuit around as you suggested, testing the RFC and I
got a dip the likes of which I had never seen. The meter went slowly from
about 30 mA to 50 mA and then dropped to about 25 mA, I couldn't miss it.

But what does it mean. I gather my RFC is not ok. What is wrong??

I used a Series 4590 high current filter inductor I had around, it has the
Digi Key number DN 4528.


Happy about the dip but still not clear on the deeper reasons...

73 Uwe



Here's something else to try:

Often trouble of this sort is due to the RF choke used. What RFCs are
you suing, particularly in the plate circuit? Although the LC meter
may say they are a certain L, in real life they may have all sorts of
unwanted resonances.

To test this idea out, do the following:

- Remove the plate RFC
- Connect the antenna end of the plate coil to the B+ where the RFC
used to be connected. This point should already be bypassed to ground
through a disk capacitor of about .01 uF
- Disconnect the "loading" capacitor
- Remove the plate coupling capacitor.

What you will then have is the 200 volts being fed to the plate
through the coil, with one end of the coil going to the plate supply
and the other end connected directly to the plate of the 6V6. The
plate tuning capacitor is connected between the plate of the 6V6 and
ground.

End result is no plate RFC and a parallel resonant circuit. There's no
connection for an antenna yet, but that's not important right now.

Test out the rig and look for the plate current dip. It should be very
obvious because there is no load connected.

This is just a temporary setup to see if the RFC is OK.

73 es GL de Jim, N2EY

In article , Uwe
writes:

Using the L/C meter I wound a proper coil, I checked the calibration of my
plate current meter, I did a more thorough check of the grid current (it is
between 1 and 2 mA) and so on and so forth.

I think you mean "screen current".

And I did connect a dummy load (even though they don't respond or send out
QSL cards when you tranmit into them).

Yup!

None of the thing did make any real difference and the dip, the elusive dip,
was in the order of magnitude of maybe 2 mA, nearly impossible to see on my
meter.

Then I changed the circuit around as you suggested, testing the RFC and I
got a dip the likes of which I had never seen. The meter went slowly from
about 30 mA to 50 mA and then dropped to about 25 mA, I couldn't miss it.

But what does it mean. I gather my RFC is not ok. What is wrong??

I used a Series 4590 high current filter inductor I had around, it has the
Digi Key number DN 4528.

The RFC you're using is not meant for the appliucation. It's intended for much
lower frequencies. You can't tell that just by looking at it.

RF choke design is a matter of compromises. For example, the use of a powdered
iron or ferrite core will raise the inductance. But that same core does not
work at all frequencies, and may saturate from DC current in the core.

The biggest problem is called "distributed capacitance". In order to get lots
of inductance, you put on lots of turns, closely spaced. But each turn has a
small amount of capacitance to the turns next to it. All these small
capacitances add up, and as the frequency is increased they become significant
to the total reactance of the choke. At one or more frequencies the choke will
actually resonate - these are called "self resonant" frequencies. At some
frequencies the choke may act like an inductor of much lower inductance, or
even like a capacitor, because of the self resonances. Self resonance in a
choke can be found with a suitable dip meter.

RF chokes that are meant for applications like the AC-1 are designed to have
self-resonant frequencies far from the amateur bands.

Happy about the dip but still not clear on the deeper reasons...

Hope this helps.

73 de Jim, N2EY

in article , N2EY at
PAMNO wrote on 4/8/04 07:59:

In article , Uwe
writes:

Using the L/C meter I wound a proper coil, I checked the calibration of my
plate current meter, I did a more thorough check of the grid current (it is
between 1 and 2 mA) and so on and so forth.

I think you mean "screen current".

And I did connect a dummy load (even though they don't respond or send out
QSL cards when you tranmit into them).

Yup!

None of the thing did make any real difference and the dip, the elusive dip,
was in the order of magnitude of maybe 2 mA, nearly impossible to see on my
meter.

Then I changed the circuit around as you suggested, testing the RFC and I
got a dip the likes of which I had never seen. The meter went slowly from
about 30 mA to 50 mA and then dropped to about 25 mA, I couldn't miss it.

But what does it mean. I gather my RFC is not ok. What is wrong??

I used a Series 4590 high current filter inductor I had around, it has the
Digi Key number DN 4528.

The RFC you're using is not meant for the appliucation. It's intended for much
lower frequencies. You can't tell that just by looking at it.

RF choke design is a matter of compromises. For example, the use of a powdered
iron or ferrite core will raise the inductance. But that same core does not
work at all frequencies, and may saturate from DC current in the core.

The biggest problem is called "distributed capacitance". In order to get lots
of inductance, you put on lots of turns, closely spaced. But each turn has a
small amount of capacitance to the turns next to it. All these small
capacitances add up, and as the frequency is increased they become significant
to the total reactance of the choke. At one or more frequencies the choke will
actually resonate - these are called "self resonant" frequencies. At some
frequencies the choke may act like an inductor of much lower inductance, or
even like a capacitor, because of the self resonances. Self resonance in a
choke can be found with a suitable dip meter.

RF chokes that are meant for applications like the AC-1 are designed to have
self-resonant frequencies far from the amateur bands.

Happy about the dip but still not clear on the deeper reasons...

Hope this helps.

73 de Jim, N2EY

Yes this is all very helpful. Indeed I was thinking that it would have taken
me a very long time to figure this out by myself, if at all.

I put another choke in there, a Hammond 1535B, the self resonant min. frequ.
is 1.3Mhz.

I guess it does take some deeper inside what parameters to look for since
this one improves things slightly but not yet altogether (slightly more
pronounced dip).

Who carries the sort of chokes you were refering to?

73 Uwe