Harmonics from K3

[Poul Erik Karlshøj (PKA)]

I have been measuring the level of harmonic spurious from my K3 #3674 and came up with some interesting results.
 
I measured with the K3 connected to a dummy load and to my antennas (loop for 80-40-30 and 17m dipole for higher bands, both antennas fed by 450 ohm feeder and Elecraft 4:1 baluns). Since I have one of the older baluns with just one ferrite core and one with two cores, I measured on each antenna with each of the baluns.

In all measurements I used the K3 ATU and the SWR was "1 bar segment"

I used a Rohde&Schwarz FSH8 portable Spectrum Analyzer using a short wire in the input connector center pin  to pick up the signal from the antennas. The analyzer was in my shack, next to the rig. But since I also used the pick-up antenna to measure the leaking signal from the dummy load (and the harmonics were all very low) the higher level harmonics measured when connected to the antenna supposedly did not come from stray signals in the shack.

So, what did I see with K3 transmitting 100W?
1,8 >38 dB
3,5 37
7 >55
10 43
14 40 (2nd harm)
18 44
21 53
24 49
28 59
This seems pretty good to me.

With K3 transmitting 100W and 2-core balun on loop (160-40) and 1-core balun on dipole (20-10)
1,8 38 dB
3,5 40
7 55
10 28
14 16 (2nd harm), 3rd harm at 30 dB
18 29
21 28
24 28
28 37
Especially the 14 MHz value was a surprise, but generally much higher harmonics on 30-10m than on the dummy load


With K3 transmitting 100W and 1-core balun on loop (160-40) and 2-core balun on dipole (20-10)
1,8 16 dB
3,5 45
7 31
10 33
14 17 (2nd harm)
18 35
21 45
24 36
28 42
The 14 MHz value still pretty bad and the high bands much better now with 2-core balun. Low bands some worse and some better with 1-core balun!

Finally - assuming that some of the harmonics might origin from non-linearities in the balun core - I measured on 14 MHz with the 2-core balun at different power levels:
14 MHz, 2-core balun and dipole antenna
100W 18 dB
80 17
60 16
40 16
20 18
10 21
5 21
It seems that the power value has very little influence on the harmonic level, so my theory of harmonics being created in the balun may not be correct.
When the ATU shows SWR ~1 does it mean that the current waveform in the PA transistors are independent of whether the load is a dummy load (and the ATU practically bypasses) or the load is a complex impedance requiring some L and C to reach 50 ohms?

Now I have the following questions:
1) is my K3#3674 specifically having a problem concerning harmonics on 14 MHz or would I measure the same on any other K3
2) what is the reason that I measure much higher harmonic levels when connected to the antenna than on dummy load?
3) when the ATU shows SWR ~1 does it mean that the current waveform in the PA transistors is independent of whether the load is a dummy load (and the ATU practically bypasses) or the load is a complex impedance requiring some L and C to reach 50 ohms?
4) have I overlooked some fundamental measurement error in using a small piece (3 cm) of wire to pick up the RF signals?

73 de
Poul-Erik
OZ4UN
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Jack Smith-6]

Poul-Erik:

You have an implicit assumption that the spectrum analyzer to transmit
antenna has a flat response with respect to frequency.

That's unlikely to be the case.

Another assumption is that the spectrum analyzer antenna is in the "far
field" of the antenna. This is also unlikely to be the case.

And another is that the K3 itself does not radiate due to imperfect
shielding which may be a problem when the pickup antenna is close to the K3.

The normal method to measure harmonic suppression is to use either a
power attenuator, such as 30 dB attenuation rated at 100 watts, so that
the spectrum analyzer only detects signals from the transmitter. Or, if
a suitable power attenuator is not available, a standard dummy load and
a resistive "tap" or a directional coupler of known frequency
characteristics can be used to obtain a low level signal sample that
does not vary with frequency.

With respect to measuring radiated harmonics, the most accurate method
is to locate the measuring equipment sufficiently far from the transmit
antenna such that the signal is far field. 10x the transmitting
wavelength is a safe number, but a somewhat shorter distance may be OK.  
The spectrum analyzer or field strength meter uses an antenna with a
known antenna factor at the fundamental frequency and all the harmonics
so that the actual radiated field strength can be computed and from that
the harmonic level. Since the transmitting antenna almost certainly has
directional characteristics -- particularly at harmonic frequencies --
several sets of measurements should be made at varying azimuths from the
transmitting antenna.

So, to make these measurements accurately in the case of radiated
harmonics is a non-trivial undertaking.

Jack K8ZOA


On 10/27/2011 7:52 AM, Poul Erik Karlshøj (PKA) wrote:

> I have been measuring the level of harmonic spurious from my K3 #3674 and came up with some interesting results.
>
> I measured with the K3 connected to a dummy load and to my antennas (loop for 80-40-30 and 17m dipole for higher bands, both antennas fed by 450 ohm feeder and Elecraft 4:1 baluns). Since I have one of the older baluns with just one ferrite core and one with two cores, I measured on each antenna with each of the baluns.
>
> In all measurements I used the K3 ATU and the SWR was "1 bar segment"
>
> I used a Rohde&Schwarz FSH8 portable Spectrum Analyzer using a short wire in the input connector center pin  to pick up the signal from the antennas. The analyzer was in my shack, next to the rig. But since I also used the pick-up antenna to measure the leaking signal from the dummy load (and the harmonics were all very low) the higher level harmonics measured when connected to the antenna supposedly did not come from stray signals in the shack.
>
> So, what did I see with K3 transmitting 100W?
> 1,8 >38 dB
> 3,5 37
> 7 >55
> 10 43
> 14 40 (2nd harm)
> 18 44
> 21 53
> 24 49
> 28 59
> This seems pretty good to me.
>
> With K3 transmitting 100W and 2-core balun on loop (160-40) and 1-core balun on dipole (20-10)
> 1,8 38 dB
> 3,5 40
> 7 55
> 10 28
> 14 16 (2nd harm), 3rd harm at 30 dB
> 18 29
> 21 28
> 24 28
> 28 37
> Especially the 14 MHz value was a surprise, but generally much higher harmonics on 30-10m than on the dummy load
>
>
> With K3 transmitting 100W and 1-core balun on loop (160-40) and 2-core balun on dipole (20-10)
> 1,8 16 dB
> 3,5 45
> 7 31
> 10 33
> 14 17 (2nd harm)
> 18 35
> 21 45
> 24 36
> 28 42
> The 14 MHz value still pretty bad and the high bands much better now with 2-core balun. Low bands some worse and some better with 1-core balun!
>
> Finally - assuming that some of the harmonics might origin from non-linearities in the balun core - I measured on 14 MHz with the 2-core balun at different power levels:
> 14 MHz, 2-core balun and dipole antenna
> 100W 18 dB
> 80 17
> 60 16
> 40 16
> 20 18
> 10 21
> 5 21
> It seems that the power value has very little influence on the harmonic level, so my theory of harmonics being created in the balun may not be correct.
> When the ATU shows SWR ~1 does it mean that the current waveform in the PA transistors are independent of whether the load is a dummy load (and the ATU practically bypasses) or the load is a complex impedance requiring some L and C to reach 50 ohms?
>
> Now I have the following questions:
> 1) is my K3#3674 specifically having a problem concerning harmonics on 14 MHz or would I measure the same on any other K3
> 2) what is the reason that I measure much higher harmonic levels when connected to the antenna than on dummy load?
> 3) when the ATU shows SWR ~1 does it mean that the current waveform in the PA transistors is independent of whether the load is a dummy load (and the ATU practically bypasses) or the load is a complex impedance requiring some L and C to reach 50 ohms?
> 4) have I overlooked some fundamental measurement error in using a small piece (3 cm) of wire to pick up the RF signals?
>
> 73 de
> Poul-Erik
> OZ4UN
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Dave Johnson]

"I used a Rohde&Schwarz FSH8 portable Spectrum Analyzer using a short
wire in the input connector center pin  to pick up the signal from the
antennas"

Using such a wire pick up will exaggerate the higher frequencies and
not give a flat response. You need to use a resistive tap or power
attenuator in order to obtain the correct measurement.

A -40 dB resistive tap can be easily built from the article in June
2001 QST. The article is called "Simple RF Power Measurement" by Wes
Hayward and available from the ARRL members section of their web site.

Measurements on my own K3 (serial 80) give the worst case harmonic as
-53 dB on 80m and less than I can measure on 14 MHz. These
measurements were on a factory calibrated analyser using a 50 Ohm
load.

73 Dave
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Robert Friess]

Hello Poul-Erik,

If your K3 is putting out 100 watts, it is very unlikely that it is
generating the harmonics that you are measuring, especially in the case of
the even harmonics.  Your measured results may be due to some non-linear
element in the near field such as a corroded or loose connection or the
difference in path loss, antenna pattern and gain (loss) of both the loop
and the your short probe on the spectrum analyzer.  Also, it appears that
the balun is a significant contributor as shown by the difference in
measured results when using one balun verses the other.

You can verify that the K3 is operating properly by connecting a suitable
attenuator between the K3 and the spectrum analyzer and making a conducted
harmonic measurement.  If you do this, be careful that the signal level into
the spectrum analyzer is low enough so that harmonics are not generated in
the spectrum analyzer itself.  The harmonics from the K3 will be in the
range of -60dBc if operating normally.

73,
Bob, N6CM

On Thu, Oct 27, 2011 at 4:52 AM, Poul Erik Karlshøj (PKA) <[hidden email]>wrote:

> I have been measuring the level of harmonic spurious from my K3 #3674 and
> came up with some interesting results.
>
> I measured with the K3 connected to a dummy load and to my antennas (loop
> for 80-40-30 and 17m dipole for higher bands, both antennas fed by 450 ohm
> feeder and Elecraft 4:1 baluns). Since I have one of the older baluns with
> just one ferrite core and one with two cores, I measured on each antenna
> with each of the baluns.
>
> In all measurements I used the K3 ATU and the SWR was "1 bar segment"
>
> I used a Rohde&Schwarz FSH8 portable Spectrum Analyzer using a short wire
> in the input connector center pin  to pick up the signal from the antennas.
> The analyzer was in my shack, next to the rig. But since I also used the
> pick-up antenna to measure the leaking signal from the dummy load (and the
> harmonics were all very low) the higher level harmonics measured when
> connected to the antenna supposedly did not come from stray signals in the
> shack.
>
> So, what did I see with K3 transmitting 100W?
> 1,8     >38 dB
> 3,5     37
> 7       >55
> 10      43
> 14      40 (2nd harm)
> 18      44
> 21      53
> 24      49
> 28      59
> This seems pretty good to me.
>
> With K3 transmitting 100W and 2-core balun on loop (160-40) and 1-core
> balun on dipole (20-10)
> 1,8     38 dB
> 3,5     40
> 7       55
> 10      28
> 14      16 (2nd harm), 3rd harm at 30 dB
> 18      29
> 21      28
> 24      28
> 28      37
> Especially the 14 MHz value was a surprise, but generally much higher
> harmonics on 30-10m than on the dummy load
>
>
> With K3 transmitting 100W and 1-core balun on loop (160-40) and 2-core
> balun on dipole (20-10)
> 1,8     16 dB
> 3,5     45
> 7       31
> 10      33
> 14      17 (2nd harm)
> 18      35
> 21      45
> 24      36
> 28      42
> The 14 MHz value still pretty bad and the high bands much better now with
> 2-core balun. Low bands some worse and some better with 1-core balun!
>
> Finally - assuming that some of the harmonics might origin from
> non-linearities in the balun core - I measured on 14 MHz with the 2-core
> balun at different power levels:
> 14 MHz, 2-core balun and dipole antenna
> 100W    18 dB
> 80      17
> 60      16
> 40      16
> 20      18
> 10      21
> 5       21
> It seems that the power value has very little influence on the harmonic
> level, so my theory of harmonics being created in the balun may not be
> correct.
> When the ATU shows SWR ~1 does it mean that the current waveform in the PA
> transistors are independent of whether the load is a dummy load (and the ATU
> practically bypasses) or the load is a complex impedance requiring some L
> and C to reach 50 ohms?
>
> Now I have the following questions:
> 1) is my K3#3674 specifically having a problem concerning harmonics on 14
> MHz or would I measure the same on any other K3
> 2) what is the reason that I measure much higher harmonic levels when
> connected to the antenna than on dummy load?
> 3) when the ATU shows SWR ~1 does it mean that the current waveform in the
> PA transistors is independent of whether the load is a dummy load (and the
> ATU practically bypasses) or the load is a complex impedance requiring some
> L and C to reach 50 ohms?
> 4) have I overlooked some fundamental measurement error in using a small
> piece (3 cm) of wire to pick up the RF signals?
>
> 73 de
> Poul-Erik
> OZ4UN
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Poul Erik Karlshøj (PKA)]

 
Hi Jack
Thanks a lot for your valuable feed-back.
Yes you are correct that there are some assumptions.
The Spectrum Analyzer is very flat over the spectrum I have been checking - thats a fact.
I assumed that a small piece of wire would be fairly omnidirectional - and thus have the same gain - on these low frequencies. I still think that is true, butI have however not taken into account that the short wire will have very different impedance on the fundamental and harmonic frequencies. This may be an important error source.

Yes I am indeed in the near field of the antenna, and this may be a source of error too. This I can verify, because the Spectrum Analyzer may run on batteries and I can operate the K3 with my iPhone (or a small laptop) so I can easily measure in the far field. I need to think about how to make a proper measurement antenna though.

I dont think pick-up directly from the K3 due to imperfect shielding is the cause, since it then also should be present on the measurements on the dummy load.

But, al this said, I don't think it explains why the measurements are so markedly different on 14 MHz compared to the other bands. Thatr is my main worry.
The question remains whether this is specific to my K3 or to something else locally.

Poul-Erik OZ4UN

-----Oprindelig meddelelse-----
Fra: [hidden email] [mailto:[hidden email]] På vegne af Jack Smith
Sendt: 27. oktober 2011 15:00
Til: Poul Erik Karlshøj (PKA)
Cc: [hidden email]
Emne: Re: [Elecraft] Harmonics from K3

Poul-Erik:

You have an implicit assumption that the spectrum analyzer to transmit antenna has a flat response with respect to frequency.

That's unlikely to be the case.

Another assumption is that the spectrum analyzer antenna is in the "far field" of the antenna. This is also unlikely to be the case.

And another is that the K3 itself does not radiate due to imperfect shielding which may be a problem when the pickup antenna is close to the K3.

The normal method to measure harmonic suppression is to use either a power attenuator, such as 30 dB attenuation rated at 100 watts, so that the spectrum analyzer only detects signals from the transmitter. Or, if a suitable power attenuator is not available, a standard dummy load and a resistive "tap" or a directional coupler of known frequency characteristics can be used to obtain a low level signal sample that does not vary with frequency.

With respect to measuring radiated harmonics, the most accurate method is to locate the measuring equipment sufficiently far from the transmit antenna such that the signal is far field. 10x the transmitting wavelength is a safe number, but a somewhat shorter distance may be OK.  
The spectrum analyzer or field strength meter uses an antenna with a known antenna factor at the fundamental frequency and all the harmonics so that the actual radiated field strength can be computed and from that the harmonic level. Since the transmitting antenna almost certainly has directional characteristics -- particularly at harmonic frequencies -- several sets of measurements should be made at varying azimuths from the transmitting antenna.

So, to make these measurements accurately in the case of radiated harmonics is a non-trivial undertaking.

Jack K8ZOA


On 10/27/2011 7:52 AM, Poul Erik Karlshøj (PKA) wrote:

> I have been measuring the level of harmonic spurious from my K3 #3674 and came up with some interesting results.
>
> I measured with the K3 connected to a dummy load and to my antennas (loop for 80-40-30 and 17m dipole for higher bands, both antennas fed by 450 ohm feeder and Elecraft 4:1 baluns). Since I have one of the older baluns with just one ferrite core and one with two cores, I measured on each antenna with each of the baluns.
>
> In all measurements I used the K3 ATU and the SWR was "1 bar segment"
>
> I used a Rohde&Schwarz FSH8 portable Spectrum Analyzer using a short wire in the input connector center pin  to pick up the signal from the antennas. The analyzer was in my shack, next to the rig. But since I also used the pick-up antenna to measure the leaking signal from the dummy load (and the harmonics were all very low) the higher level harmonics measured when connected to the antenna supposedly did not come from stray signals in the shack.
>
> So, what did I see with K3 transmitting 100W?
> 1,8 >38 dB
> 3,5 37
> 7 >55
> 10 43
> 14 40 (2nd harm)
> 18 44
> 21 53
> 24 49
> 28 59
> This seems pretty good to me.
>
> With K3 transmitting 100W and 2-core balun on loop (160-40) and 1-core balun on dipole (20-10)
> 1,8 38 dB
> 3,5 40
> 7 55
> 10 28
> 14 16 (2nd harm), 3rd harm at 30 dB
> 18 29
> 21 28
> 24 28
> 28 37
> Especially the 14 MHz value was a surprise, but generally much higher
> harmonics on 30-10m than on the dummy load
>
>
> With K3 transmitting 100W and 1-core balun on loop (160-40) and 2-core balun on dipole (20-10)
> 1,8 16 dB
> 3,5 45
> 7 31
> 10 33
> 14 17 (2nd harm)
> 18 35
> 21 45
> 24 36
> 28 42
> The 14 MHz value still pretty bad and the high bands much better now with 2-core balun. Low bands some worse and some better with 1-core balun!
>
> Finally - assuming that some of the harmonics might origin from non-linearities in the balun core - I measured on 14 MHz with the 2-core balun at different power levels:
> 14 MHz, 2-core balun and dipole antenna
> 100W 18 dB
> 80 17
> 60 16
> 40 16
> 20 18
> 10 21
> 5 21
> It seems that the power value has very little influence on the harmonic level, so my theory of harmonics being created in the balun may not be correct.
> When the ATU shows SWR ~1 does it mean that the current waveform in the PA transistors are independent of whether the load is a dummy load (and the ATU practically bypasses) or the load is a complex impedance requiring some L and C to reach 50 ohms?
>
> Now I have the following questions:
> 1) is my K3#3674 specifically having a problem concerning harmonics on
> 14 MHz or would I measure the same on any other K3
> 2) what is the reason that I measure much higher harmonic levels when connected to the antenna than on dummy load?
> 3) when the ATU shows SWR ~1 does it mean that the current waveform in the PA transistors is independent of whether the load is a dummy load (and the ATU practically bypasses) or the load is a complex impedance requiring some L and C to reach 50 ohms?
> 4) have I overlooked some fundamental measurement error in using a small piece (3 cm) of wire to pick up the RF signals?
>
> 73 de
> Poul-Erik
> OZ4UN
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net Please help support this email
> list: http://www.qsl.net/donate.html
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html