CB and Ham Radio Techs Love Their Bird Wattmeters

[EE4all]

Valid point about 5 watts. I am not using them for CB, so for myself, it wasn't a concern. For the thread scope though, also including CB, yes certainly a concern. I could run some tests if there is interest.

[joeqsmith]

I guess the hams have their QRP radios which are also limited to 5W.    Someone was recently asking about Wattmeters that I think would work below 100mW.    There was also a ham who was asking about measuring their SWR out three places beyond the decimal point. 

It would be interesting to measure yours if you have a way to measure it.   

[EE4all]

Ironically, they have a LP-1 sensor that is listed as "20 milliwatts – 60 watts" for HF, but then all their other sensors say 0 to whatever their maximum is.

[joeqsmith]

Ironically, they have a LP-1 sensor that is listed as "20 milliwatts – 60 watts" for HF, but then all their other sensors say 0 to whatever their maximum is.

Maybe the 60/2^12 or 14.6mW per count.  Still though just the quantization error would make no sense.    The 5W min SWR was with the 500W sensor.  So 122mW per count.  5W hardly seems like enough.  I wonder if they have a higher gain stage for the reverse. 

With the log amps, it's no problem as they will automatically adjust the gain.   Problem I ran into was not wanting to do anything beyond a simple point slope correction and getting the coupler linear over that range.   No idea how linear these couplers are as they never publish the data.   Odd when you consider they provide a recommended interface circuit if you wanted to use them with your own meter. 

[EE4all]

They have this table of voltage reference points published.

https://wavenodedevelop.com/wp-content/uploads/2017/12/power_cal_3.pdf

[joeqsmith]

I guess you could try to correct for it.  A better option may be to pull the detector... Gets back to buying the right tools.   I think if we wanted to measure down to 25mW in the HF bands, I would start with a better coupler.   

[xrunner]

Speaking of couplers ... I ordered another one of the 25 to 400 MHz Putnam couplers before they're sold out. 

Been working on my VSWR tests with the one I have now using more power (up to +37 dBm) than my Agilent sig gen can output (+13 dBm). Should have some results by tomorrow. From what I'm seeing, I should be good-to-go on the next steps in the project. 

[EE4all]

I guess you could try to correct for it.  A better option may be to pull the detector... Gets back to buying the right tools.   I think if we wanted to measure down to 25mW in the HF bands, I would start with a better coupler.

If talking about that low power Wavenode sensor, to use for QRP 25mW ham, one wouldn't want to pick that one anyway, because it is at the bottom of its range. One wants to work mostly in the most linear middle range of a sensor.

I do have an in calibration E4418B and in calibration 8482A sensor here. If you wanted me to test something specifically, then I can try to get to it.

[joeqsmith]

I guess you could try to correct for it.  A better option may be to pull the detector... Gets back to buying the right tools.   I think if we wanted to measure down to 25mW in the HF bands, I would start with a better coupler.

If talking about that low power Wavenode sensor, to use for QRP 25mW ham, one wouldn't want to pick that one anyway, because it is at the bottom of its range. One wants to work mostly in the most linear middle range of a sensor.

I do have an in calibration E4418B and in calibration 8482A sensor here. If you wanted me to test something specifically, then I can try to get to it.

That was the data for the one you mentioned that was advertised to 25mW.   Sound like we agree that it is a poor choice.  It makes me wonder why they would advertise it this way rather than spec it in its linear region.   

I think the data they provided is good enough to get a feel for how the detector / couplers work.  I wonder if you had that coupler with one of their meters and swept from 25mW on up, how well the system performs.  Does their meter try to correct for it?   

[joeqsmith]

... before they're sold out. 

I swear it's not me buying them up...   

Curious if they have a detector built in.  Guessing with the SMA, it's AC made to drive 50 ohms.   Will wait for your report.

[EE4all]

It looks like purely dual directional couplers to me.

[Wavenode]

Hello All-- I can share a bit of information about the Wavenode Wattmeters since I did the R.F. design for the meters.  The primary source of sensor error tolerances is with R.F. detectors and their inherent nonlinearity.  This is factored into the meter calibration scales and also by applying as much R.F. to the diodes as possible.  To measure real-time SWR of modulated signals (such as SSB), it is essential that forward and reflected powers are measured synchronously on sensor detectors.  Also, if the gain relationship between sensors is desired, then all sensors must run synchronously.  To do this, all Wavenode meters used track-following peak detectors with a common reset for all four sensors.  A single A/D is capable of acquiring both the forward and reflected power data for all four sensors running synchronously. 
We have provided a document of our meters and sensors that we use for Club presentations.  It provides much detail about our sensors and overall meter architecture.  You can find that here:
https://wavenodedevelop.com/wp-content/uploads/2022/08/Wavenode-Station-Presentation_rev2.pdf
    We are always happy to provide information and help for inquiries regarding our products.  You can reach us at contact@wavenode.com.
73
Wavenode Inc.

[joeqsmith]

...    We are always happy to provide information and help for inquiries regarding our products.  You can reach us at contact@wavenode.com.

I took the time to read your PDF but it did not address my question about the LP-1 sensor that was mentioned.   

Ironically, they have a LP-1 sensor that is listed as "20 milliwatts – 60 watts" for HF, but then all their other sensors say 0 to whatever their maximum is.

With it connected to your meter as a system if you plot the linearity, what does it look like?  Does your meter attempt to correct for the change in slope at the lower levels?    If not, why does your company list the part as 20 mW? 

[Wavenode]

     The Wavenode LP-1 sensor will indicate power down to 25 milliwatts, but not within +/- 5% accuracy.  Most users would not use this sensor at such low power levels and SWR measurement would be more of concern in normal operation.  All Wavenode meters use a 3rd order polynomial to account for detector non-linearity at very low power levels. 
     All Wavenode products are sold in Europe, and need CE certification for susceptibility to static and ESD, as well as radiated/conducted emissions from the meters.  Internal clocks were selected expressly to avoid any harmonic radiation in Amateur Radio bands, a consideration in a long list of engineering do's/don't  to survive and operate correctly in the high R.F. fields around a typical Amateur radio antenna, as well as public-service communication towers.   These products have matured over a 20 year timespan and survived some brutal conditions in remote communication mountain-top systems.  You can find some info about clock selection in our tech document here:   https://wavenodedevelop.com/wp-content/uploads/2022/08/Technote-120-R.F.I.-Interface.pdf

73
Wavenode Inc.

[joeqsmith]

     The Wavenode LP-1 sensor will indicate power down to 25 milliwatts, but not within +/- 5% accuracy.  Most users would not use this sensor at such low power levels and SWR measurement would be more of concern in normal operation.  All Wavenode meters use a 3rd order polynomial to account for detector non-linearity at very low power levels. 

So no data showing the system error over the entire range?   Do you have some hard limit where below nWatts, you switch to the polynomial fit or are you using that 3rd order for the entire specified range?   

I had to use a fifth order for my display graphics but the actual is point slope only.   Took some effort to get the coupler as good as it is but I don't have to deal with that diode detector.   

[Wavenode]

   Operating from memory here since these sensors were devel0ped 20+ years ago.  They will remain +/- 5% over 2-60 watt range and +/- 10% down to 1 watt.  The polynomial fit is for levels <2 watts, but there is a piecewise linear fit above 50% of the range, which is about 30 watts for the LP-1.  All of our sensors are plug/play and not individually calibrated for a particular control box.  Other production tolerances come into play to contribute to the +/- 5% spec.  A user can purchase our wattmeter today, but add sensors at any time in the future and expect the rated tolerances.   A mass produced device can never economically produce the precision  results of an HP437B and 8482A sensor that have been calibrated together at a cal lab.  The HP units we use for calibration and engineering will always be superior instruments for product development.
73
Wavenode Inc.

[xrunner]

Curious if they have a detector built in.  Guessing with the SMA, it's AC made to drive 50 ohms.   Will wait for your report.

Yea just RF out @ -50dB

Refer to pic for the testing setup I used to check the measurements using a higher power input from my little test transmitter (Baofeng UV-5R).

The actual measured output powers of the UV-5R checked with a hp 436B (no coupler inserted):

VHF low/high Watts: 1.8, 5.1
UHF low/high Watts: 1.4, 4.0

Some years ago I made an SWR test load for calibrating the SWR / power meters on HF transceivers - meaning up to 29 MHz. It uses three 50 ohm RF load resistors and, using a switch, can present SWR 2 or 3 to the transmitter. I used this to check the SWR on 146 and 445 MHz. Given the fact I never designed it for use with VHF and UHF bands it still gave me results that tell me the coupler will work in the application I'm working towards. I need to design a better SWR test load for these higher bands (better shielding and layout for higher frequencies).

Using this SWR test load, I measured the FWD & REV port powers under load conditions that should cause a VSWR of 2 and 3 and so the coupler outputs (FWD & REV) should give us (or a microcontroller) the power levels to calculate it.

Corrected port power used below in dBm (actual port power -50 dB less)

VSWR 2

VHF 146.0 MHz
F 36.4, R 28.4, calculated VSWR 2.4

UHF 445.0 MHz
F 36.7, R 26.8, calculated VSWR 1.9

VSWR 3

VHF 146.0 MHz
F 35.6, R 30.0, calculated VSWR 3.2

UHF 445.0 MHz
F 35.0, R 29.0, calculated VSWR 3.0

In addition to the wideband plot of coupled FWD power output provided in earlier post, I took precise measurements of the advertised -50 dB coupling for each of four ham bands, for use in the software to come later:

6m band (52 MHz): -50.8
2m band ( 146 MHz): -50.8
1.25m band (222 MHz): -51.2
70cm band (445 MHz): -50.4

At this point I'm pretty satisfied that the Putnam dual directional coupler will be satisfactory for this project, so I'll move forward and start collecting some parts I'll need to check it out with a microcontroller.

[EE4all]

Nice. Did you happen to characterize the directivity on the Putnam?

The directivity graph will highlight the quality or lack of it on the coupler. I would suggest that before starting those SWR tests.

[xrunner]

Nice. Did you happen to characterize the directivity on the Putnam?

The directivity graph will highlight the quality or lack of it on the coupler. I would suggest that before starting those SWR tests.

My testing this morning indicates a directivity of 25.5 dB. This is in-line with other quality dual-directional couplers such as the ones made by Mini-Circuits here which have a directivity range of 23 to 27 dB -

https://www.minicircuits.com/WebStore/Couplers.html

[EE4all]

It would be good to check it across the entire range of frequencies. It will not be flat of course.

[joeqsmith]

My testing this morning indicates a directivity of 25.5 dB.

An article from Bird I had posted previously.   

The directivity of the coupler that forms the heart of a directional power meter is critical to the accuracy of these measurements. Errors due to directivity can be significant and may influence the decisions you make based upon your test results.

https://www.electronicdesign.com/home/article/21201015/directional-power-measurements-and-the-effects-of-directivity

[xrunner]

An article from Bird I had posted previously.   

The directivity of the coupler that forms the heart of a directional power meter is critical to the accuracy of these measurements. Errors due to directivity can be significant and may influence the decisions you make based upon your test results.

https://www.electronicdesign.com/home/article/21201015/directional-power-measurements-and-the-effects-of-directivity

Yep, every meter made is going to have some degree of error, just a matter of knowing what it could be. I'm working on a plot for the Putnam as we speak. I do believe this: The coupler I'm using will be much better than almost any of the low end SWR / Power meters sold.

The article states:

"The directivity of the coupler that forms the heart of a directional power meter is critical to the accuracy of these measurements. Errors due to directivity can be significant and may influence the decisions you make based upon your test results."

I shudder to think about the "heart" of one of the cheap ham meters like the Daiwa and its little tin coupler box with components that didn't even match the schematic, probably fine tuned as best they could by some poor shmuck after the design was "finished". What could the errors be, you know they don't exactly publish the directivity specs of those things ... 

And all the equations in the article to calculate the correct powers even if the directivity was known - what inside the Daiwa is performing the calculations?

Yet hams will believe the readings of them more than many things their own family members tell them. The errors in most of those cheap devices due to very poor couplers & directivity, according to the graphs, would probably make them nearly useless but nobody cares. 

[xrunner]

Putnam coupler directivity plot. Markers at four ham bands of interest.

[joeqsmith]

I keep thinking it would be fun to do some sort of shootout.  I've forked out this amount of cash to test handheld multi-meters to failure.  Not being a ham or CB radio hobbyist, I don't have that same level of interest.   Maybe we could reach out to these companies to see if they would be interested in it.  It's a lot of time to invest as well.  No way I would make a ham review where I just unbox them, blab for a half hour how great they are, and give them all five stars.      

I swore I had posted data for the couplers but searching this section, looks like I did not.   The video shows the directivity printed on the one coupler as 30dB @25MHz and 50dB@1MHz.   

When I made the one that was actually used during the video, it measured 65@1M and 40@30MHz.  (see attached)  Isolation was better than 50 and the flatness was <0.7 total.   Good enough that I chose not to correct for it but I think if we did a shoot out, I would go ahead and add correction to shave some of that error out. 

[joeqsmith]

This was some of the data I collected for the 19dB coupler.   Directivity isn't very good and of course it dissipates more heat.