Calibration of Fwd/Ref (SWR) meters.

Discussion in 'Antennas, Feedlines, Towers & Rotors' started by KY8D, Nov 21, 2021.

ad: L-HROutlet
ad: l-rl
ad: abrind-2
ad: Left-2
ad: L-Geochron
ad: L-MFJ
ad: Left-3
  1. KY8D

    KY8D Subscriber QRZ Page

    Which is the most common practice for calibrating meter readouts: Peak, Average, or RMS power?

    I ask because I have recently built a meter according to G3YNH's Duoma bridge schematic (this for inclusion into a home-brew link-coupled FRI-Match antenna tuner still being constructed).

    I have two MFJ antenna tuners, an MFJ-984 purchased new in 1981, and an MFJ-989C got used from a hamfest. These do not agree with each other. Both work equally well for tuning antennas, but now I'm wanting a reference for the meter I'm building. A quandary, therefor.

    The MFJ-984 shows my IC-745 outputting 140W, while the MFJ-989C reports 100W. On the one hand, as IC-745 is rated for 100W, I might be inclined to trust to that. Except for the fact that some years ago I had sent the IC-745 out for service to a technician. It then came back with all new caps, the 60m band added, and immediately showing higher output on the same MFJ-984 (then my one and only meter).

    A further quandary. I have also built a home-brew dummy load, into which I included a diode and taps for measuring power (as per a design found on-line). With a 0.01uF cap across the power tap terminals, and doing the V^2/R formula, I get different answers entirely, both higher and lower than either MFJ meter depending on whether dividing V by 1.414.

    Also...so I find...that diode and 0.01uF cap on the dummy load now strikes me as a not such a good way after all for measuring power. It raises SWR all by itself. And the worse, of course, the higher in F.

    Just wanting to do this thing as well as I can without shelling out for a Bird meter. For instruments I have a Fluke DVM and a Hitachi 60MHz scope. Recommendations?
     
  2. K7TRF

    K7TRF Premium Subscriber QRZ Page

    You've got the classic man with two watches never knows the time situation especially as you introduce home-brew gear.

    I'd probably start by testing the dummy load with an ohm meter and perhaps temporarily remove the built in RF probe (that cap and diode circuit) to see if it reads close to 50 ohms at DC. Then I'd use the scope to measure the peak voltage on a lower frequency HF band using a 10x scope probe. Then calculate power based on peak voltage as:

    P = Vpk^2/(2R)

    IOW, a 100 watt signal into a 50 ohm dummy load should measure 100 volts peak on the oscilloscope.

    That removes a few variables like the RF probe and the conversion of peak voltage to RMS voltage prior to the power calculation. The accuracy of this method depends on how close to 50 ohms your dummy load is at the operating frequency and how accurate your scope is as well as how accurately you can read peak voltage from your scope. As you go higher in frequency the accuracy will likely drop depending on the dummy load construction methods and how close you get to the 60 MHz 3dB bandwidth of your scope.

    Also remember even a good commercial power meter like a Bird is only accurate to approximately +/-5% of the full scale meter reading so don't expect pin point accuracy from common ham SWR/power meters.
     
    Last edited: Nov 21, 2021
    SWL37632, KY8D and WN1MB like this.
  3. SM0AOM

    SM0AOM Ham Member QRZ Page

    The only way to measure RF power with high accuracy
    is by using a calibrated attenuator in front of a thermal power sensor. Properly done, better than +/- 2 % of reading accuracy can be obtained.

    Such an arrangement always reads average power, but is quite expensive, especially the attenuator.

    The next best is a diode rectifier over a dummy load of known impedance operated at a sufficient high voltage so the contact potential and diode non-linearities can be neglected, for a semiconductor diode at least 15 - 20 V.

    A vacuum-tube diode such as the one in the HP410C probe can work up to 300 V or so, and may also read peak power depending on the time-constant of the indicator.

    Using diode detectors, an accuracy of +/- 5% of full scale is what to be expected.

    Finally, there is no such thing in the radio world as "RMS power", which is an abomination that has crept in from the audiophools.

    73/
    Karl-Arne
    SM0AOM
     
    KD5YI, SWL37632, KY8D and 1 other person like this.
  4. W1VT

    W1VT Ham Member QRZ Page

    There is a hidden cost with thermal power sensors. They often burn up when someone miscalculates the attenuation in front of it. The repair cost is very high.

    I used 1N277M germanium diodes to peak detect the voltage across a dummy load when I was a young ham. I measured the rectified voltage with a Fluke 73 DMM.

    Zak W1VT
     
    KY8D likes this.
  5. KY8D

    KY8D Subscriber QRZ Page

    Ten minutes ago, I had a very bad thought. I had built my home-brew dummy load some months ago from an array of twenty discrete 2W resistors in parallel, these under mineral oil inside a 1 pint paint can.

    I went with 2% precision metal film resistors, with the result that R = 50.1 Ohms at the BNC. But...I have since learned that metal film resistors are spiral wound. A few turns only of wide-banded film. But still inductive. A big problem in the microwave range. This is an issue I read of in G3YNH's wonderful website.

    I have just now tested for L with my LCR meter, getting 0.069 uH. Surely that can't be enough at HF to be any significant problem. Or do I err?
     
    Last edited: Nov 21, 2021
  6. K7TRF

    K7TRF Premium Subscriber QRZ Page

    Only a minor issue at 3.5 MHz at roughly 1.5 ohms of inductive reactance but a bit bigger deal at 29 MHz with over 12 ohms of inductive reactance on a nominally 50 ohm load.

    From an SWR standpoint that's not a big deal at HF but the SWR near the top of the 10m band would climb to roughly 1.3:1 which isn't a show stopper but also not ideal. Practically speaking your rig will load into that just fine but for test and measurement purposes a dummy load with a smaller inductive component would be better.
     
    KY8D likes this.
  7. W1VT

    W1VT Ham Member QRZ Page

    I used metal oxide resistors to make dummy loads and attenuators at VHF and 70cm. You may want to try those if you want a better SWR. They are quite resistant of abuse.
     
    KY8D likes this.
  8. KY8D

    KY8D Subscriber QRZ Page

    Adding that rebuild to the end of my to-do list.
     
  9. KY8D

    KY8D Subscriber QRZ Page

    A bit of research informs that SMD metal film resistors are laser etched in a zig-zag, rather than spiral. So there is that option also.
     
  10. K7TRF

    K7TRF Premium Subscriber QRZ Page

    I'm a big fan of home brewing ham gear but if you do find yourself needing a dummy load flat to much higher frequencies I've bought a few of these for use on my lab bench and it's a hard deal to beat. I've swept mine on professional test equipment and they're a very good 50 ohm match up past 1 GHz: https://www.ebay.com/itm/265351908998?hash=item3dc834ce86:g:EtMAAOSwNWxaS~Bu
     
    KY8D likes this.

Share This Page