1. Nanovna-Users
  2. Теми

Querry a tuners' settings for L&C with a NanoVNA

 

Is there any way to determine a tuners L and C settings/values such that a fixed PI network could be built to replace the tuner.

Case in point. I have a QCX and the 50w amp which exhibits a mismatch between the two. I have set my qrpguys tuner in between and tuned it. Now I want to build a tuned circuit with a toroid and caps to replace the tuner.

I'm sure there is a better way but ,,,

Bryan N0LUF
 

Hi Brian,

Indeed if you characterize the impedance you can arrive at a fixed matching network.

If we suppose the QCX requires a 50 Ω load, a reasonable assumption, and the QRP Guys tuner got it matched, you can try the following approach:

Disconnect the tuner and put a "dummy" load of 50 Ω on the _transmitter_ side;
Then measure with nanoVNA the impedance "seen" in the load side of the tuner, note its value.

The input impedance of the 50 W amp will be the conjugate of this measurement, so if you read, for example, 12 resistive and 40 inductive ohms, 12 +j40 Ω, then the input would be 12 -j40 Ω, i.e., 12 resistive and 40 capacitive.

From these numbers you can use the formulas appropriate for the match.

Depending on the values found maybe even a single L network could get the job done, as you intend to use a fixed one.

More details, as the frequency range of the intended match and the real numbers may steer the approach to other designs, but the road map is this.

HTH

--
Cesar, PY2CSH
 

Hi Bryan,Of course the real question here is possibly why is there a mismatch  between the output of the qcx and the input of the amp? They are presumably notionally both 50 ohms. Is it really significant? Is there a fault somewhere?But... Depending on the tuner's construction and topology you may be able to get at the individual  components and measure them in place but I suspect the calibration error due to difficult connection into the tuner would be quite large.A more holistic approach would be to simply measure the effective complex impedance of the adjusted tuner as a whole. One way would be to put a 50 ohm dummy load on the output of the tuner and measure the impedance R+jX at the input. This will tell you that the output impedance of the QCX is R-jX (the complex conjugate). Now you can use any of the online pi, L or T match calculators to design a matching network that is the equivalent of the tuner.Hth.Andrew Sent from my Galaxy

-------- Original message --------From: "Bryan Curl via groups.io" <bc3910@...> Date: 30/12/24 8:42 am (GMT+10:00) To: nanovna-users@groups.io Subject: [nanovna-users] Querry a tuners' settings for L&C with a NanoVNA Is there any way to determine a tuners L and C settings/values such that a fixed PI network could be built to replace the tuner.Case in point. I have a QCX and the 50w amp which exhibits a mismatch between the two. I have set my qrpguys tuner in between and tuned it. Now I want to build a tuned circuit with a toroid and caps to replace the tuner.I'm sure there is a better way but ,,,Bryan N0LUF
 

On Sun, Dec 29, 2024 at 05:57 PM, Andrew Love wrote:


Of course the real question here is possibly why is there a mismatch  between
the output of the qcx and the input of the amp?
Driving a non-class A amplifier is complicated. For one thing, the input impedance varies significantly with supply voltage and drive/power level, and to some extent the output tuning........
2:1 VSWR is a common spec limit. Consider the situation where the source is 100Ω and the load is 25Ω, there is a 4:1 mismatch yet both are "nominally" 50Ω. Power transfer will clearly not be optimum.

For another, the output impedance also varies and the output power and efficiency varies with the load, even more than just the mismatch loss would suggest.
Even the impedance at the 2nd, 3rd and possibly higher harmonics will have significant effects, at both input and output.

You might get more than power with 50Ω, you might get less, it depends on the amplifier and its tuning and bias conditions.
Many of these same effects will occur with the driving amplifier/transmitter.
Things get complicated when you are driving a dynamic, nonlinear load with a similarly imperfect source.

https://en.wikipedia.org/wiki/Load_pull https://www.everythingrf.com/community/what-are-load-pull-measurements
73, Don N2VGU
 

Of course, given that they’re not resistive 50 ohms on both sides, the fact that there’s a “mismatch” might not be a big deal - For instance, a DC power supply has a 0 ohm output impedance, and transfers as much power to the load as it will take (given the voltage).

The whole emphasis on matching is sort of a “design help” back in the day before good modeling tools existed, and when getting gain was difficult.

On Dec 29, 2024, at 17:18, Donald S Brant Jr via groups.io <dsbrantjr@...> wrote:

On Sun, Dec 29, 2024 at 05:57 PM, Andrew Love wrote:


Of course the real question here is possibly why is there a mismatch between
the output of the qcx and the input of the amp?
Driving a non-class A amplifier is complicated. For one thing, the input impedance varies significantly with supply voltage and drive/power level, and to some extent the output tuning........
2:1 VSWR is a common spec limit. Consider the situation where the source is 100Ω and the load is 25Ω, there is a 4:1 mismatch yet both are "nominally" 50Ω. Power transfer will clearly not be optimum.

For another, the output impedance also varies and the output power and efficiency varies with the load, even more than just the mismatch loss would suggest.
Even the impedance at the 2nd, 3rd and possibly higher harmonics will have significant effects, at both input and output.

You might get more than power with 50Ω, you might get less, it depends on the amplifier and its tuning and bias conditions.
Many of these same effects will occur with the driving amplifier/transmitter.
Things get complicated when you are driving a dynamic, nonlinear load with a similarly imperfect source.

https://en.wikipedia.org/wiki/Load_pull https://www.everythingrf.com/community/what-are-load-pull-measurements
73, Don N2VGU

1 - 5 з 5