Measuring transmitter power, do and don't

[neazoi]

Hello, I usually see measurements of power of HF transmitters using just a common power meter (broadband coupler and diode inside). Please tell me if my next thoughts are correct:

1. Since these power meters are broadband, and they have just a needle analogue meter to measure power (non selective), they not only measure power of the carrier, but the total power of the carrier plus the transmitter harmonics. If the transmitter output is not filtered, false readings (higher) would result from such measurements.

2. Unless a selective volt meter or a spectrum analyzer is used, actual power of the carrier (thus ignoring harmonics power) cannot be made.

3. If an effective LPF is used at the transmitter output, then the total measured power from a common power meter will be close enough to the actual carrier power, since harmonics will be much attenuated.

4. One way to measure power with the scope (assumming a harmonics-free signal), from a 50R transmitter, is to terminate the transmitter with a 50R non-inductive dummy load and then measure the voltage accross this load, from the 1M scope port (not the 50R!). The power measurement then will be done like directly measuring with a a scope having an input of 50R.

Please tell me if my thoughts are correct on each of the above points.

Thanks

 

[jiripolivka]

Hello, I usually see measurements of power of HF transmitters using just a common power meter (broadband coupler and diode inside). Please tell me if my next thoughts are correct:

1. Since these power meters are broadband, and they have just a needle analogue meter to measure power (non selective), they not only measure power of the carrier, but the total power of the carrier plus the transmitter harmonics. If the transmitter output is not filtered, false readings (higher) would result from such measurements.

2. Unless a selective volt meter or a spectrum analyzer is used, actual power of the carrier (thus ignoring harmonics power) cannot be made.

3. If an effective LPF is used at the transmitter output, then the total measured power from a common power meter will be close enough to the actual carrier power, since harmonics will be much attenuated.

4. One way to measure power with the scope (assumming a harmonics-free signal), from a 50R transmitter, is to terminate the transmitter with a 50R non-inductive dummy load and then measure the voltage accross this load, from the 1M scope port (not the 50R!). The power measurement then will be done like directly measuring with a a scope having an input of 50R.

Please tell me if my thoughts are correct on each of the above points.

Thanks

Yes, you are right in all points. A good transmitter should have a LPF at its output to attenuate harmonics, then a power meter will indicate only the desired carrier power.
Using a power divider is also correct, and all its ports should be terminated in 50 Ohms. The oscilloscope input is often 1 MegaOhm, so a 50 Ohm termination should be attached.

The best method combines the power divider with a calibrated spectrum analyzer, so you can measure reliably the desired carrier power as well as the harmonics (and other possible spurs).
Care must be taken not to exceed SA input power as this may generate harmonics and spurs.
This combination is preferable as it allows you to monitor the correct PS adjustment and load matching to an antenna or another load.

 

[FvM]

You didn't talk about frequency range. A RF voltage measurement with a high impedance oscilloscope input might introduce all kinds of measurement errors, e.g. due to standing waves. Preferably, you'll use an attenuator with respective power rating and an oscilloscope with 50 ohm input. Lower SW band could work with high impedance input though.

To be legally operated, a transmitter requires harmonic filtering to at least -40 dBc. A wide band power measurement after the output filter should give accurate results.

 

[neazoi]

You didn't talk about frequency range. A RF voltage measurement with a high impedance oscilloscope input might introduce all kinds of measurement errors, e.g. due to standing waves. Preferably, you'll use an attenuator with respective power rating and an oscilloscope with 50 ohm input. Lower SW band could work with high impedance input though.

To be legally operated, a transmitter requires harmonic filtering to at least -40 dBc. A wide band power measurement after the output filter should give accurate results.

It is basically HF (0-30MHz). As said, I already terminate the transmitter with a 50R dummy load, so the 1M scope input parallel to it, will yeld to a very close to 50R input impedance. So I think we could roughly say that the measurements are done on 50R, like having a 50R scope input. Do you think that standing waves on HF with this conficuration, can cause a huge readings error?

What I want to emphasize with this pont is that the 1M scope port must be connected in parallel to the 50R dummy load. One should not connect the 50R scope port (if this is available) to the 50R dummy load, because this would yeld in an impedance of 25R. This pont is not so clear in some sites I have read.

In this above point it is assumed that a good LPF is at the TX output, if a scope is used for the measurement.

 

[betwixt]

What I want to emphasize with this pont is that the 1M scope port must be connected in parallel to the 50R dummy load. One should not connect the 50R scope port (if this is available) to the 50R dummy load, because this would yeld in an impedance of 25R. This pont is not so clear in some sites I have read.

Not only a 25 Ohm load but half the TX power is dissipated in the 'scope input attenuator! :-(

Brian.

 

[FvM]

Not only a 25 Ohm load but half the TX power is dissipated in the 'scope input attenuator!

Yes. People invented high power attenuators for this purpose.

I admit that things are less critical below 30 MHz. Even a (short) 50 ohm coax stub from a T-connector at the dummy load to a 1 Mohm oscilloscope input won't cause huge VSWR, although it's mismatched at both ends. You can also use a 10:1 or 100:1 passive probe to connect the RF.

 

[chuckey]

Just use a diode peak detector using something like a 1n916 straight on the feeder, it can be calibrated at low frequency, say 50 HZ, if the reservoir capacitor is made larger. It does rely on the load being known (50 ohms?), but you should be able to measure the voltage +- 1% which is +- 2% in power terms. With CRO, unless its a digital on, the absolute voltage is +- 5% and that's without any SWR on the coax lead. It is (very) wide band, one hopes but the harmonic content is low enough not to affect the measured value too much.
If you have two similar RF loads put your transmitter into one and DC into the other, adjust the DC until they both attain the same operating temperature. Swop over loads and repeat, the RF power is the mean of the two DC values.
If you are ambitous, try this :- https://en.wikipedia.org/wiki/Bolometer
Frank

 

[biff44]

if you had a lowpass filter already part of the transmit circuit, then obviously measure power after the filter. If there is no filter, then add a NON REFLECTIVE lowpass and measure the fundamental power after it, and then add the loss of the lowpass filter.