On 4 mayo, 23:11, walt wrote:
On May 2, 8:09*pm, Cecil Moore wrote:



On May 2, 5:23*pm, Wimpie wrote:

I am very sorry Cecil, but I still don't see the point where the
discussed method may go wrong.

Everyone seems to be charging ahead, willy-nilly, without seeing the
point which is that there are other effects present besides
reflections.

Therefore carrying out a single-port measurement with a slightly off-
carrier frequency (to create non-coherence) under required output
conditions, will result in a meaningful output impedance.

Nope, it won't because virtual impedances don't cause reflections.
Only physical impedance discontinuities cause reflections. The rest of
the redistribution of RF energy is caused by the superposed
interaction between forward and reflected waves, i.e. interference
effects. Most hams do not understand the role of interference in the
redistribution of RF energy. Hope this helps.

http://micro.magnet.fsu.edu/primer/j...interference/w...

Please pay close attention to the last paragraph. "... when two waves
of equal amplitude and wavelength that are 180-degrees ... out of
phase with each other meet, they are not actually annihilated, ... All
of the photon energy present in these waves must somehow be recovered
or redistributed in a new direction, according to the law of energy
conservation ... Instead, upon meeting, the photons are redistributed
to regions that permit constructive interference, so the effect should
be considered as a redistribution of light waves and photon energy
rather than the spontaneous construction or destruction of light."

You guys are presuming that reflections are the only thing you are
seeing and that is just not true. You are also seeing interference
effects without realizing it so your conclusions are doomed to failure
unless you can differentiate between constructive/destructive
interference and reflected waves. Since there has been no mention of
interference effects, I am forced to conclude that you guys are
ignorant of such effects.
--
73, Cecil, w5dxp.com
"Halitosis is better than no breath at all.", Don, KE6AJH/SK

During 1991 Warren Bruene used the RPG method in which he believes he
measured the source resistance of an RF power amp, which he calls
'Rs'. I have never agreed that his method measures the source
impedance, or that his data has any relevance to anything.

Consequently, I am not impressed with the discussion going on here
concerning applying a signal back into an operating RF power amp to
determine the source impedance. Please define 'source impedance'--
where is it located in the amp? At the plate? At the output of the pi-
network?. And how do you know the data obtained using this method is
correct? Have you verified it by comparing it with data obtained using
another method?

I made a statement in an earlier post that when measuring the output
impedance using the 'load pull' method we're not concerned with the
absolute maximum power that can be delivered, but instead, limiting
the 'maximum' power delivered to that which can be delivered with a
specific level of grid drive, one which allows the power *to be
limited to that of a normal operating level. Tom disagrees with this
position, that it is really the ABSOLUTE MAXIMUM power delivery that
should be considered. As you can see, I don't agree with Tom.

*I don't know how many on this thread have actually reviewed the
portion of my Chapter 19 that presents the step-by-step procedure I
used in determining the output impedance of the Kenwood TS-830S tx,
which shows precisely the output impedance appearing at the output of
the pi-network.

To summarize the procedure that I maintain will provide an accurate
measurement of the output impedance appearing at the output terminals
of the pi-network is as follows:

1) Adjust the loading and tuning controls of the amp to deliver all
the available power to a complex load in the amount normally used in
operation with the setting of the grid-drive level required to obtain
that output power.

2).Measure the impedance of the complex load.

3) The output impedance, or 'source' impedance of the amp appearing at
the output terminals of the pi-network is the complex conjugate of the
load impedance.

Now, when you measure the source impedance using the externally-
injected signal, does the data from that measurement agree with that
of the load-measuring method? If it does, then I'll agree that the RPG
method is valid. If it doesn't I'll continue to have considerable
doubt as to its validity. But I'd still like to know where the
resistance measured by this method is located in the amp.

Walt

Hello Walt,

Except for bias or supply voltage change due to load change, load
pulling does give similar results as off-carrier signal injection. I
did this in simulation for various circuits (linear and non-linear).
You may remember that I put something in a document (discussion of
last year, it is still on my website).

You can get different results in case of soft power supply or bias
supplies. In case of manual load pulling, bias/supply voltage may
change (think of change in grid current due to change in RF plate
voltage). There is sufficient time for all voltages and currents to
settle. I added a section on the envelope response due to bias
current/voltage change.

When you use the injection method (for example with 130 Hz offset), it
is like load pulling where you switch the load 200 times/s (more
specifically you rotate the phase of the reflection coefficient as
seen by the PA). In such a situation bias and supply voltages will
settle to an average value resulting in (slightly) different results.

While not relevant for here, but nice to mention, stiffness of bias
supplies has influence on IMD also.


With kind regards,

Wim
PA3DJS
www.tetech.nl

On 5 mayo, 00:44, John KD5YI wrote:
On 5/4/2011 4:11 PM, walt wrote:



On May 2, 8:09 pm, Cecil *wrote:
On May 2, 5:23 pm, *wrote:

I am very sorry Cecil, but I still don't see the point where the
discussed method may go wrong.

Everyone seems to be charging ahead, willy-nilly, without seeing the
point which is that there are other effects present besides
reflections.

Therefore carrying out a single-port measurement with a slightly off-
carrier frequency (to create non-coherence) under required output
conditions, will result in a meaningful output impedance.

Nope, it won't because virtual impedances don't cause reflections.
Only physical impedance discontinuities cause reflections. The rest of
the redistribution of RF energy is caused by the superposed
interaction between forward and reflected waves, i.e. interference
effects. Most hams do not understand the role of interference in the
redistribution of RF energy. Hope this helps.

http://micro.magnet.fsu.edu/primer/j...interference/w....

Please pay close attention to the last paragraph. "... when two waves
of equal amplitude and wavelength that are 180-degrees ... out of
phase with each other meet, they are not actually annihilated, ... All
of the photon energy present in these waves must somehow be recovered
or redistributed in a new direction, according to the law of energy
conservation ... Instead, upon meeting, the photons are redistributed
to regions that permit constructive interference, so the effect should
be considered as a redistribution of light waves and photon energy
rather than the spontaneous construction or destruction of light."

You guys are presuming that reflections are the only thing you are
seeing and that is just not true. You are also seeing interference
effects without realizing it so your conclusions are doomed to failure
unless you can differentiate between constructive/destructive
interference and reflected waves. Since there has been no mention of
interference effects, I am forced to conclude that you guys are
ignorant of such effects.
--
73, Cecil, w5dxp.com
"Halitosis is better than no breath at all.", Don, KE6AJH/SK

During 1991 Warren Bruene used the RPG method in which he believes he
measured the source resistance of an RF power amp, which he calls
'Rs'. I have never agreed that his method measures the source
impedance, or that his data has any relevance to anything.

Consequently, I am not impressed with the discussion going on here
concerning applying a signal back into an operating RF power amp to
determine the source impedance. Please define 'source impedance'--
where is it located in the amp? At the plate? At the output of the pi-
network?. And how do you know the data obtained using this method is
correct? Have you verified it by comparing it with data obtained using
another method?

I made a statement in an earlier post that when measuring the output
impedance using the 'load pull' method we're not concerned with the
absolute maximum power that can be delivered, but instead, limiting
the 'maximum' power delivered to that which can be delivered with a
specific level of grid drive, one which allows the power *to be
limited to that of a normal operating level. Tom disagrees with this
position, that it is really the ABSOLUTE MAXIMUM power delivery that
should be considered. As you can see, I don't agree with Tom.

* I don't know how many on this thread have actually reviewed the
portion of my Chapter 19 that presents the step-by-step procedure I
used in determining the output impedance of the Kenwood TS-830S tx,
which shows precisely the output impedance appearing at the output of
the pi-network.

To summarize the procedure that I maintain will provide an accurate
measurement of the output impedance appearing at the output terminals
of the pi-network is as follows:

1) Adjust the loading and tuning controls of the amp to deliver all
the available power to a complex load in the amount normally used in
operation with the setting of the grid-drive level required to obtain
that output power.

2).Measure the impedance of the complex load.

3) The output impedance, or 'source' impedance of the amp appearing at
the output terminals of the pi-network is the complex conjugate of the
load impedance.

Now, when you measure the source impedance using the externally-
injected signal, does the data from that measurement agree with that
of the load-measuring method? If it does, then I'll agree that the RPG
method is valid. If it doesn't I'll continue to have considerable
doubt as to its validity. But I'd still like to know where the
resistance measured by this method is located in the amp.

Walt

It seems to me that we maybe should not be talking about source
impedance but maybe about 'regulation' or some other equivalent word.

For an open-loop source made from real components, I think one will
always find that there is a dE/dI number for output loading that we can
call source resistance.

It does not always mean that there is a physical resistance in the
circuit. It is simply a measure of the ability of the device to provide
an unvarying voltage under conditions of varying load.

Does this make any sense?

John

Hello John,

You are correct with respect to the physical resistor. If you tune a
real class-C amplifier for maximum output power (conjugated match),
you may not reach the highest efficiency, but at least significantly
above 60%. So this shows that there is no "physical" 50 Ohms in the PA
(assuming 50 Ohms load). If so, the efficiency would never exceed
50%.

When you change the load (after tuning for maximum output) from 50
Ohms to (for example) 30 Ohms and would carry out an impedance
measurement around that 30 Ohms load resistance, you will measure
something completely different. With "around 30 Ohms" I mean using 32
and 28 Ohms (for example).

With kind regards,


Wim
PA3DJS
www.tetech.nl

On 5/4/2011 11:34 AM, Jim Lux wrote:
John KD5YI wrote:


Transmitters don't have a clearly defined output impedance and,
whatever output impedance is there, doesn't necessarily mean anything.

Your main concern is to provide a 50 ohm load for the transmitter to see.

actually, to provide an *acceptable* load for the transmitter that
maximizes radiated RF power. That might not be 50 ohms.. and I don't
know that we actually care what it is, unless we're designing amplifiers.

Acceptable is what the manufacturer recommends for his gear. What does
this have to do with the device's output impedance?

John

John KD5YI wrote:

It seems to me that we maybe should not be talking about source
impedance but maybe about 'regulation' or some other equivalent word.

For an open-loop source made from real components, I think one will
always find that there is a dE/dI number for output loading that we can
call source resistance.

It does not always mean that there is a physical resistance in the
circuit. It is simply a measure of the ability of the device to provide
an unvarying voltage under conditions of varying load.

This is comparable to the concept of "dynamic resistance" and, as well,
to "negative resistance" (as exhibited by gas discharges for instance).

The local slope of the V/I curve is negative, but that doesn't mean that
it has negative power.



Does this make any sense?

John

John KD5YI wrote:
On 5/4/2011 11:34 AM, Jim Lux wrote:
John KD5YI wrote:


Transmitters don't have a clearly defined output impedance and,
whatever output impedance is there, doesn't necessarily mean anything.

Your main concern is to provide a 50 ohm load for the transmitter to
see.

actually, to provide an *acceptable* load for the transmitter that
maximizes radiated RF power. That might not be 50 ohms.. and I don't
know that we actually care what it is, unless we're designing amplifiers.

Acceptable is what the manufacturer recommends for his gear. What does
this have to do with the device's output impedance?


Absolutely nothing.. which is the point. Manufacturers have to make
their device work into some sort of loads, but can't test every possible
load, nor necessarily optimize the design for every possible load that
someone might use.

So we use, as a convention, that we'll measure the output power or DC to
RF conversion efficiency or Power Added Efficiency or whatever, with a
50 ohm resistive load hooked up to it.

That doesn't mean that we actually want to use a 50 ohm load in real
life. Maybe I have an amplifier designed to drive a loop antenna with a
resistive 10 ohm impedance. I might optimize it for that, but, because
people want to see a spec sheet, I'm going to have to measure the output
into 50 ohms and put that on the data sheet.

Or, perhaps, the active devices I use happen to have a "natural" output
impedance of, say, 8-10 ohms, so I put a 4:1 transformer on the output.
My box, if I were to measure the output Z, would show about 40 ohms.
I'd test it with a 50 ohm load, make sure I could put out 100W into that
load so I can sell it as a "100 W transmitter" and be done with it.

It might happen that you could hook up a 40 ohm load, or a 20 ohm load,
and it will work just fine, and may even put out more than 100W.

The fact of the matter is that ham manufacturers don't actually specify
the output Z: they just give you a range of outputs into which they'll
guarantee the transmitter won't fail. (e.g. say, 25 to 100 ohms..
perhaps specified a bit sloppily as VSWR 2:1 if driven from a 50 ohm source)

Interestingly, my IC7000 manual doesn't even give a output impedance, or
even a range of legal load impedances (page 150 of the manual).

Page 11 does say "Antenna Connector: Accepts a 50 ohm antenna with a
PL-259 connector"

And Page 15 does say "Use a well matched 50ohm antenna" and "An SWR of
1.5:1 or lower is recommended" doesn't say it won't work, doesn't even
say it only develops rated output power (on page 150) into that
particular load. There is a boxed note that says that if the SWR is
higher than approx 2.0:1, the transceiver's output power is reduced.


And, let's look at the service manual.. Spec page is identical to the
one in the user manual. The procedure calls out a whole bunch of
adjustments to set the "% power" to appropriate levels using a 50 ohm
load, but doesn't say what the power into some other load Z would be.

There is a test that when you hook up 100 ohms, the internal SWR meter
reads 2:1, and 50 ohms reads 1:1. But that's not a test of the
transmitter's output Z, more a calibration of the transmitter as RF
ohmmeter for an external component.


The ALC system actually measures the output VOLTAGE (not power) and
matches that up against the setpoint for the power (implying that if you
hook up a 40 ohm load, and you set the radio for 50W, you'll get a bit
more out).

the Automatic Power Limiting is a current sensor: if it exceeds 22A, it
reduces the drive. In fact, it kind of looks like the reflected HF power
(from the sampler at the filter output) isn't used to control the drive
at all.. it's just used to drive the SWR indicator. Kind of tough to
tell, the prose description is a bit unclear, and I haven't followed all
the signals through on the schematic.



but the summary is,

On 5/4/2011 7:54 PM, Jim Lux wrote:
John KD5YI wrote:

Acceptable is what the manufacturer recommends for his gear. What does
this have to do with the device's output impedance?


Absolutely nothing.. which is the point.


Are we arguing the same point?


but the summary is,


That it is a bag of worms? I'm waiting with baited breath...

On Wed, 04 May 2011 17:54:53 -0700, Jim Lux
wrote:

That doesn't mean that we actually want to use a 50 ohm load in real
life. Maybe I have an amplifier designed to drive a loop antenna with a
resistive 10 ohm impedance.

Interesting. For a community that is so tight-fisted with cash, and
so brag-hearty with power claims, absolutely no one has ever tossed
their hat into the ring that with lower than 50Ohm termination on
their rig (and I'm not talking about 45Ohms) or with a more than 50Ohm
termination on their rig (and this is certainly achievable with a
72Ohm Dipole and 70 Ohm coax) that they have then proclaimed they
substantially exceeded 100W radiated (or lost to heat for that matter)
by the same degree of offset from 50.

Or, perhaps, the active devices I use happen to have a "natural" output
impedance of, say, 8-10 ohms,

That must date to the 1950s vintage of solid state. Your own IC7000
certainly doesn't suffer that abysmal "natural" output impedance. The
RD70HHF1 for 97W into 50Ohms exhibits 0.77-j0.22 Ohms @ 30MHz, a far
cry from your supposition. Off by 1000%?

It might happen that you could hook up a 40 ohm load, or a 20 ohm load,
and it will work just fine, and may even put out more than 100W.

"May" is lazy, "Does" is more authoritative. Any reports of "Does?"

And Page 15 does say "Use a well matched 50ohm antenna" and "An SWR of
1.5:1 or lower is recommended" doesn't say it won't work, doesn't even
say it only develops rated output power (on page 150) into that
particular load. There is a boxed note that says that if the SWR is
higher than approx 2.0:1, the transceiver's output power is reduced.

Which is curious when the manufacturer of the RD70HHF1 power
transistors gives them a "No destroy" rating into a mismatch of 20:1.
Such is the inertia of 1950s design-think with modern components.

73's
Richard Clark, KB7QHC

Dear Wimpie: The content of the paragraph below may well rise above the
noise of this thread.
I expect to learn something of value from your comments about why twice the
(apparent) output Z of an amplifier was of importance and what you did to
have the amplifiers conform.

73, Mac N8TT
-------------------
"Wimpie" wrote in message
...

Hello Dave and John,

snip

Regarding the "academic discussion" I also agree. In my professional
career where I designed several RF PA's, only 2 times the output
impedance of the amplifier was of importance. In one of these cases I
couldn't meet the specs and had to insert attenuation (some waste of
power…).

snip

With kind regards,

Wim
PA3DJS
www.tetech.nl
Remove abc first before setting free the pigeon.
--------------

J. C. Mc Laughlin
Michigan U.S.A.
Home:

John KD5YI wrote:
On 5/4/2011 7:54 PM, Jim Lux wrote:
John KD5YI wrote:

Acceptable is what the manufacturer recommends for his gear. What does
this have to do with the device's output impedance?


Absolutely nothing.. which is the point.


Are we arguing the same point?


but the summary is,


That it is a bag of worms? I'm waiting with baited breath...

Exactly..


In fact, as interesting as it would be to measure the output impedance
of my radio, I started to think about what it would buy me, and came to
the conclusion, almost nothing (other than satisfying curiosity).

It *might* be interesting to look at (and write an article for QST/QEX
or something) "optimizing radiated power". Answering the question: do
you really want a 50 ohm match on your antenna analyzer, or do you want
maximum net power at the antenna feedpoint, and what that might mean for
typical 100W solid state rigs, antennas, etc.

(as a practical matter, this is what automatic antenna tuners actually
adjust for: either minimum reflected power, or maximum fwd-ref)

but it's possible that deliberately running a mismatch (as shown on your
rig's SWR meter) might actually result in more radiated power. e.g. if
at 1:1 you have 100W fwd and 0 rev, but at 2:1 you have 150W fwd and 15W
reflected, so you're actually net 135W vs 100W; assuming your rig
doesn't otherwise have any problems.

Richard Clark wrote:
On Wed, 04 May 2011 17:54:53 -0700, Jim Lux
wrote:

That doesn't mean that we actually want to use a 50 ohm load in real
life. Maybe I have an amplifier designed to drive a loop antenna with a
resistive 10 ohm impedance.

Interesting. For a community that is so tight-fisted with cash, and
so brag-hearty with power claims, absolutely no one has ever tossed
their hat into the ring that with lower than 50Ohm termination on
their rig (and I'm not talking about 45Ohms) or with a more than 50Ohm
termination on their rig (and this is certainly achievable with a
72Ohm Dipole and 70 Ohm coax) that they have then proclaimed they
substantially exceeded 100W radiated (or lost to heat for that matter)
by the same degree of offset from 50.


Maybe that's because of tradition.. it's not only hams who do that. In
the professional world, there's a strong tendency to make the
"interfaces" between subsystems 50 ohms, even if overall system
efficiency might be improved by, say, a 25 ohm impedance. I've worked
on systems where a DAC produced a balanced signal with a 200 ohm Z, and
the vector modulator at the other end had a balanced input with a 200
ohm Z, but everyone insisted we put a 4:1 transformer on each end, so we
could run 50 ohm coax for a 10cm run.

If you're a amp maker, you're probably going to sell a lot more amps if
you specify and design for 50 ohms (the 10 ohm PA for loop antennas
would be a niche market). Likewise, if you're an antenna builder,
designing a product for 50 ohms will sell more than, say, 35 ohms.


In the cellphone world, because the PA and antenna tend to be developed
as an integrated assembly, they're probably a bit more flexible on this.

There's also the issue of wanting to put a power meter in line (but I
think that if you're using a 10 ohm system, you're probably capable of
figuring out how to measure power to make sure that you're not busting
the FCC limits)



Or, perhaps, the active devices I use happen to have a "natural" output
impedance of, say, 8-10 ohms,

That must date to the 1950s vintage of solid state. Your own IC7000
certainly doesn't suffer that abysmal "natural" output impedance. The
RD70HHF1 for 97W into 50Ohms exhibits 0.77-j0.22 Ohms @ 30MHz, a far
cry from your supposition. Off by 1000%?

Just an example. I know that modern parts are MUCH lower. The point is
that if one knew you had a 3 ohm load, you could design an amplifier
that would directly drive it without a transformer or network, and
potentially get better overall efficiency.




It might happen that you could hook up a 40 ohm load, or a 20 ohm load,
and it will work just fine, and may even put out more than 100W.

"May" is lazy, "Does" is more authoritative. Any reports of "Does?"

The link from several days ago for the gentleman who measured the output
Z being around 40 ohms would be a "does"..



And Page 15 does say "Use a well matched 50ohm antenna" and "An SWR of
1.5:1 or lower is recommended" doesn't say it won't work, doesn't even
say it only develops rated output power (on page 150) into that
particular load. There is a boxed note that says that if the SWR is
higher than approx 2.0:1, the transceiver's output power is reduced.

Which is curious when the manufacturer of the RD70HHF1 power
transistors gives them a "No destroy" rating into a mismatch of 20:1.
Such is the inertia of 1950s design-think with modern components.


Yes, but the design limit might not be the transistor mismatch. It
might be a thermal dissipation limit, or some other component that's the
limiting factor.

(Or even, a desire to not have to answer 1000 customer service inquiries
to explain why it really doesn't matter what you hook up, or to explain,
that, yes, you DO need to actually connect an antenna of some sort)

Amateur radio, particularly in the mass market, is pretty slow to move.