Hi,

I have a matching circuit that matches 50 Ohms to 4500 Ohms in parallel
with 2.5pF at 45MHz.

The coil was measured at 147nH with a Q of 50 at 45 MHz.

The circuit is a C-Tap typology with following values.

Tap input to ground = 784pF
Tap to hot side of coil = 92pF
Par inductor = 147nH

My design calculations predict Loaded Q = 26 and BW = 1.7 MHz.

I used a 50p trimmer in par with 68pF cap for top of tap to fine tune.

A 4k7 resistor and 2p2 cap were placed across the output to simulate the
correct load.

The test results show BW = 1.7 MHz and Loaded Q = 24.7 (agree with
calculated values).

When installed on the IF front end, the expected sensitivity is achieved
when tuned. This would indicate most of the signal must be getting
through or sensitivity would be effected due to mismatch.

All seems to look ok but when I attach the MJF-269 Analyzer I get
impedance of 280 angle 35 degrees (240 -j 179) for the impedance, I get
SWR = 6.1 , return loss of 2.8dB and match efficiency of 48%.

I cannot understand why there is such a discrepency between the design
values and tested results (using sig gen, spectrum analyzer and power
meter) compared to the MJF-269.

Is there something I am doing wrong here or is it a mis-interpretation
of displayed values ?

Thanks for any help.

Have you considered that there may be non-linear effects in the test set up?

The MFJ uses several to ten milliwatts in it's circuitry [+10 dBm].
Sensitivity is generally measured at the pico-watt [~ -130 dBm] level.

AK

David wrote:

Hi,

I have a matching circuit that matches 50 Ohms to 4500 Ohms in parallel
with 2.5pF at 45MHz.

The coil was measured at 147nH with a Q of 50 at 45 MHz.

The circuit is a C-Tap typology with following values.

Tap input to ground = 784pF
Tap to hot side of coil = 92pF
Par inductor = 147nH

My design calculations predict Loaded Q = 26 and BW = 1.7 MHz.

I used a 50p trimmer in par with 68pF cap for top of tap to fine tune.

A 4k7 resistor and 2p2 cap were placed across the output to simulate the
correct load.

The test results show BW = 1.7 MHz and Loaded Q = 24.7 (agree with
calculated values).

When installed on the IF front end, the expected sensitivity is achieved
when tuned. This would indicate most of the signal must be getting
through or sensitivity would be effected due to mismatch.

All seems to look ok but when I attach the MJF-269 Analyzer I get
impedance of 280 angle 35 degrees (240 -j 179) for the impedance, I get
SWR = 6.1 , return loss of 2.8dB and match efficiency of 48%.

I cannot understand why there is such a discrepency between the design
values and tested results (using sig gen, spectrum analyzer and power
meter) compared to the MJF-269.

Is there something I am doing wrong here or is it a mis-interpretation
of displayed values ?

Thanks for any help.

"Amos Keag" wrote in message
...
Have you considered that there may be non-linear effects in the test set
up?

The MFJ uses several to ten milliwatts in it's circuitry [+10 dBm].
Sensitivity is generally measured at the pico-watt [~ -130 dBm] level.

AK

David wrote:

Hi,

I have a matching circuit that matches 50 Ohms to 4500 Ohms in parallel
with 2.5pF at 45MHz.

The coil was measured at 147nH with a Q of 50 at 45 MHz.

The circuit is a C-Tap typology with following values.

Tap input to ground = 784pF
Tap to hot side of coil = 92pF
Par inductor = 147nH


These seem like horrendously large capacitors to use at 45 MHz.


My design calculations predict Loaded Q = 26 and BW = 1.7 MHz.

I used a 50p trimmer in par with 68pF cap for top of tap to fine tune.

A 4k7 resistor and 2p2 cap were placed across the output to simulate the
correct load.

The test results show BW = 1.7 MHz and Loaded Q = 24.7 (agree with
calculated values).

For an unloaded Q of 50, I would shoot for a loaded Q of around 5. If you
plan to put any power through this, try getting the unloaded Q up to around
200, with a loaded Q of around 8.

My guess is that if you remove the 4.7K, the impedance at the other end of
the network will not change all that much.

Tam/WB2TT

Hi,

A little obvious perhaps but have you tried a 50-ohm resistor
across the output of the 269 to check its calibration. I have a
259B and will wind a similar inductor to see what results that
network yields. I did a SPICE run on your values and it came up
with figures within the correct range although phase resonance and
maximum impedance were off by a MHz or so.

Will report back - Joe

Amos,

The circuit is passive (R/L/C) only. I would not expect any non-linearity.

I suspect the previous post may have been correct. The match was over a
large ratio of impedance (50 Ohms to 4500 Ohms). I expect the small
value of inductor and reasonably low Q was limiting the performance of
the circuit.

I tried the match again from 50 Ohms to 800 Ohms as a test and found the
MJ-269 showed match efficiency of 90% with SWR dropping to 1.8. These
results are much better.

Regards

David

Amos Keag wrote:
Have you considered that there may be non-linear effects in the test set
up?

The MFJ uses several to ten milliwatts in it's circuitry [+10 dBm].
Sensitivity is generally measured at the pico-watt [~ -130 dBm] level.

AK

David wrote:

Hi,

I have a matching circuit that matches 50 Ohms to 4500 Ohms in
parallel with 2.5pF at 45MHz.

The coil was measured at 147nH with a Q of 50 at 45 MHz.

The circuit is a C-Tap typology with following values.

Tap input to ground = 784pF
Tap to hot side of coil = 92pF
Par inductor = 147nH

My design calculations predict Loaded Q = 26 and BW = 1.7 MHz.

I used a 50p trimmer in par with 68pF cap for top of tap to fine tune.

A 4k7 resistor and 2p2 cap were placed across the output to simulate
the correct load.

The test results show BW = 1.7 MHz and Loaded Q = 24.7 (agree with
calculated values).

When installed on the IF front end, the expected sensitivity is
achieved when tuned. This would indicate most of the signal must be
getting through or sensitivity would be effected due to mismatch.

All seems to look ok but when I attach the MJF-269 Analyzer I get
impedance of 280 angle 35 degrees (240 -j 179) for the impedance, I
get SWR = 6.1 , return loss of 2.8dB and match efficiency of 48%.

I cannot understand why there is such a discrepency between the design
values and tested results (using sig gen, spectrum analyzer and power
meter) compared to the MJF-269.

Is there something I am doing wrong here or is it a mis-interpretation
of displayed values ?

Thanks for any help.

Tam,

I believe you are correct. The 147nH inductor is the largest "can" that
I have. The inductor really needs to be larger to get the loaded Q of
the match down for this ration of impedance transform.

When trying 50 Ohms to 800 Ohms (instead of the original 4500 Ohms), the
match works. The actual loaded Q was about 8.

Seems a good rule of thumb would be to go for loaded Q around 10% of
unloaded Q.

Thanks

Tam/WB2TT wrote:
"Amos Keag" wrote in message
...

Have you considered that there may be non-linear effects in the test set
up?

The MFJ uses several to ten milliwatts in it's circuitry [+10 dBm].
Sensitivity is generally measured at the pico-watt [~ -130 dBm] level.

AK

David wrote:


Hi,

I have a matching circuit that matches 50 Ohms to 4500 Ohms in parallel
with 2.5pF at 45MHz.

The coil was measured at 147nH with a Q of 50 at 45 MHz.

The circuit is a C-Tap typology with following values.

Tap input to ground = 784pF
Tap to hot side of coil = 92pF
Par inductor = 147nH



These seem like horrendously large capacitors to use at 45 MHz.



My design calculations predict Loaded Q = 26 and BW = 1.7 MHz.

I used a 50p trimmer in par with 68pF cap for top of tap to fine tune.

A 4k7 resistor and 2p2 cap were placed across the output to simulate the
correct load.

The test results show BW = 1.7 MHz and Loaded Q = 24.7 (agree with
calculated values).


For an unloaded Q of 50, I would shoot for a loaded Q of around 5. If you
plan to put any power through this, try getting the unloaded Q up to around
200, with a loaded Q of around 8.

My guess is that if you remove the 4.7K, the impedance at the other end of
the network will not change all that much.

Tam/WB2TT

David,

With an MFJ-259B I just ran a test using an air-wound coil
resonating with a series combination of 820pF and 100pF at about
44MHz, all strapped across a 4.7K resistor. The figures came out
as follows -

Inductor (calculated) = 146nH

Real part of Zin (with 4.7k) = 14 ohms (tuned for X = 0)

With perfect components, and the capacitor tap chosen, this
should have been 55 ohms. Disconnecting the 4.7k (again tuned for
X = 0) the real part of Zin increased to 20 ohms. This, if my
maths are correct, means the inductor's equivalent shunt loss
resistance was about 1.6K giving an inductor Q of around 40.

Due to the similarity of component values used, these then are
the sort of figures you should have obtained. Obviously the
limiting factor here is the circuit Q but I supposed you had
intentionally chosen this to meet a specific bandwidth target.


Cheers - Joe

"David" wrote in message
...
Tam,

I believe you are correct. The 147nH inductor is the largest "can" that I
have. The inductor really needs to be larger to get the loaded Q of the
match down for this ration of impedance transform.

When trying 50 Ohms to 800 Ohms (instead of the original 4500 Ohms), the
match works. The actual loaded Q was about 8.

Seems a good rule of thumb would be to go for loaded Q around 10% of
unloaded Q.

Thanks

David,

The inductor loss is (LoadedQ)/(UnloadedQ). So for your numbers of 26 and
50, the loss will be 0.52, or about 3 DB. I think you will want to look at a
different configuration. Either something that puts a very small cap in
series with the 4500w||2.5PF, or go for something like a 2 uH inductor in
parallel with it. At 45 MHz you need an LC product of 12.5 to resonate; this
will then be inductive. Tune and match off a tap of the inductor with 2
capacitors.

Tam

David,

Be careful of that "Match Efficiency" number read out by the MFJ. I
don't remember what they are showing, but I do remember there was a caution
in their manual about just what it was and that it was a rather crude
indicator. Some over simplified readout.

73, Steve, K,9.D;C'I


"David" wrote in message
...
Amos,

The circuit is passive (R/L/C) only. I would not expect any non-linearity.

I suspect the previous post may have been correct. The match was over a
large ratio of impedance (50 Ohms to 4500 Ohms). I expect the small
value of inductor and reasonably low Q was limiting the performance of
the circuit.

I tried the match again from 50 Ohms to 800 Ohms as a test and found the
MJ-269 showed match efficiency of 90% with SWR dropping to 1.8. These
results are much better.

Regards

David

Amos Keag wrote:
Have you considered that there may be non-linear effects in the test set
up?

The MFJ uses several to ten milliwatts in it's circuitry [+10 dBm].
Sensitivity is generally measured at the pico-watt [~ -130 dBm] level.

AK

David wrote:

Hi,

I have a matching circuit that matches 50 Ohms to 4500 Ohms in
parallel with 2.5pF at 45MHz.

The coil was measured at 147nH with a Q of 50 at 45 MHz.

The circuit is a C-Tap typology with following values.

Tap input to ground = 784pF
Tap to hot side of coil = 92pF
Par inductor = 147nH

My design calculations predict Loaded Q = 26 and BW = 1.7 MHz.

I used a 50p trimmer in par with 68pF cap for top of tap to fine tune.

A 4k7 resistor and 2p2 cap were placed across the output to simulate
the correct load.

The test results show BW = 1.7 MHz and Loaded Q = 24.7 (agree with
calculated values).

When installed on the IF front end, the expected sensitivity is
achieved when tuned. This would indicate most of the signal must be
getting through or sensitivity would be effected due to mismatch.

All seems to look ok but when I attach the MJF-269 Analyzer I get
impedance of 280 angle 35 degrees (240 -j 179) for the impedance, I
get SWR = 6.1 , return loss of 2.8dB and match efficiency of 48%.

I cannot understand why there is such a discrepency between the design
values and tested results (using sig gen, spectrum analyzer and power
meter) compared to the MJF-269.

Is there something I am doing wrong here or is it a mis-interpretation
of displayed values ?

Thanks for any help.

Steve,

The MJK unit is now giving sensible results.

I changed the match from 4500 to 800 Ohms. I think the inductor Q at the
higher ratio
of impedance transform was not high enough. Matching to lower impedance
only required a Q of around 8.

Steve Nosko wrote:
David,

Be careful of that "Match Efficiency" number read out by the MFJ. I
don't remember what they are showing, but I do remember there was a caution
in their manual about just what it was and that it was a rather crude
indicator. Some over simplified readout.

73, Steve, K,9.D;C'I


"David" wrote in message
...

Amos,

The circuit is passive (R/L/C) only. I would not expect any non-linearity.

I suspect the previous post may have been correct. The match was over a
large ratio of impedance (50 Ohms to 4500 Ohms). I expect the small
value of inductor and reasonably low Q was limiting the performance of
the circuit.

I tried the match again from 50 Ohms to 800 Ohms as a test and found the
MJ-269 showed match efficiency of 90% with SWR dropping to 1.8. These
results are much better.

Regards

David

Amos Keag wrote:

Have you considered that there may be non-linear effects in the test set
up?

The MFJ uses several to ten milliwatts in it's circuitry [+10 dBm].
Sensitivity is generally measured at the pico-watt [~ -130 dBm] level.

AK

David wrote:


Hi,

I have a matching circuit that matches 50 Ohms to 4500 Ohms in
parallel with 2.5pF at 45MHz.

The coil was measured at 147nH with a Q of 50 at 45 MHz.

The circuit is a C-Tap typology with following values.

Tap input to ground = 784pF
Tap to hot side of coil = 92pF
Par inductor = 147nH

My design calculations predict Loaded Q = 26 and BW = 1.7 MHz.

I used a 50p trimmer in par with 68pF cap for top of tap to fine tune.

A 4k7 resistor and 2p2 cap were placed across the output to simulate
the correct load.

The test results show BW = 1.7 MHz and Loaded Q = 24.7 (agree with
calculated values).

When installed on the IF front end, the expected sensitivity is
achieved when tuned. This would indicate most of the signal must be
getting through or sensitivity would be effected due to mismatch.

All seems to look ok but when I attach the MJF-269 Analyzer I get
impedance of 280 angle 35 degrees (240 -j 179) for the impedance, I
get SWR = 6.1 , return loss of 2.8dB and match efficiency of 48%.

I cannot understand why there is such a discrepency between the design
values and tested results (using sig gen, spectrum analyzer and power
meter) compared to the MJF-269.

Is there something I am doing wrong here or is it a mis-interpretation
of displayed values ?

Thanks for any help.





--

Kind Regards

David Huisman
General Manager
-----------------------------------------------------------------------
ORBIT COMMUNICATIONS Pty Ltd - Wireless Solutions that Work
(Telemetry, Control, Monitoring, Security, HVAC ...)
A.C.N. 107 441 869


Website : http://www.orbitcoms.com
PO Box 4474 Lakehaven
NSW 2263, AUSTRALIA
Phone: 61-2-4393-3627
Fax : 61-2-4393-3685
Mobile: 61-413-715-986