Testing EFHW UnUns

testing in the lab

EFHW UnUn Testing - The Tales It Tells

The saying, “You can’t test quality into a product” was said to me quite a few decades ago now and to this day I still like it.

With respect to EFHW UnUns it is actually rather prophetic. I’ve seen EFHW designs and kits that have efficiencies around 60% that make me wonder if they have ever been tested adequately. Some are even worse and made me think they would be like fire-crackers.

The old “works great” is my guess as to how these were tested which probably means  a simple listening test which just isn’t adequate.

The other amusing and actually perplexing thing is that so many designs look the same yet when I test them, their efficiency is appalling. Then when I make one small change, the loss is halved.

Maybe it’s my country upbringing, but I’ve always tried to do the best I can, to make the best I can, and for OZ-Tenna to make sure what we do meets design criteria. There is a fair amount of pride in that for me.

Conclusion – I’ll take pity on you – You don’t have to read all of this!

The  test I like the most is to connect the the device under test to a transceiver and use it the way it is intended to be used.

I like to push the device well past its stated limits to ensure we have plenty of “leeway” to take into account real life variations and uncertainties. For example, we pushed 2.5 times the rated power into one device for an hour of SSB – and it passed with ease.

The hardest test for any UnUn is with digital modes.  If an UnUn has an efficiency issue, this till show it up in the form of core heating.

For example, I put 50W of digital into the Handy-100 and measured the temperature rise in degrees C. Knowing the time, weight, specific heat of ferrite we can calculate loss.

Possible Test Methods

Here are a list of test methods you can use to test an UnUn.

What are you testing for? My own preference is efficiency –  how much power that is put into the UnUn that is then delivered to the load – the antenna. To me that makes the most sense – when I run QRP I want all my watts in the antenna, not wasted heating up ferrite cores.

Undoubtedly there are more test methods; we just don’t know them.

SWR Test. This test must be understood and used only within its limits.

It is the worst test to tell you about efficiency – meaning the amount of power that reaches your antenna – as a percentage of course.

When you measure SWR, remember that a dummy load has a great SWR. Your UnUn could be soaking up a lot of your power in heating making the SWR look great.

One thing that makes me smile. People publish sweeps showing their antenna and UnUn has a fantastic SWR on all bands. It tells you nothing about how the antenna is radiating or the efficiency of the UnUn.

With that in mind, SWR is valuable as a “precursor” test – if the SWR is nowhere near what you expect then clearly there is an issue somewhere.

VNA – Rp Measure and estimate efficiency.    This one is rather simple and can be a great help in eliminating designs early.

Method is simple; put a primary only on the core. Measure Rp with a VNA, then use the formula Efficiency = 1-50/Rp as a percent.

This works remarkably well for the low end of HF, but does tend to collapse at the top end.

Back To Back Transformer.    This “feels” like a great test as the thought process is that you are using the transformers into ideal loads, themselves.

Method is to build 2 transformers, connect the secondaries together, then connect one side to a VNA P1 and the other transformer primary to P2. Then measure the through loss and divide by 2. You can read the loss right off the screen.

The reality is however that the test is flawed. The secondaries of the transformers are highly reactive and do not present an ideal load, far from it.

My own experience is that this test shows results about 4 to 5% lower than other tests.

Resistor Load Termination.    This test is easier, you only need one UnUn and that is an advantage.

Method is to build just one UnUn and terminate the secondary in its intended impedance. Then with a VNA measure the through loss and add the ratio of the primary to secondary impedance decibels. If the ratio is 49:1 then the value to add to the reading from the VNA is 10 Log 49 = 16.9dB.

If you read -17.0dB as S21, then the loss is -17.0dB + 16.9dB = -0.1dB which is your through loss.

My concern with this test is that no antenna in real life will present this impedance, purely resistive and a perfect match, to the UnUn. Also, and to be highlighted, you must carefully calibrate your VNA as reading values to 0.1 dB resolution and accuracy is an ask for nanoVNA type instruments.

My experience is that this test shows better results than are achieved with a real antenna – maybe by a few percent at least.

If you are a ‘sharpie’ at marketing, then this test gives the best results – never mind the reality. One of our products shows 97% efficiency with this test – this is GREAT and like all things that look too good, it must be not well connected to reality.

The real benefit of this test is simplicity – you can test and compare designs relatively quickly. To be honest – it can also stroke the ego a little.

Core Heating Test. This is my favorite test as it is a real life and under actual use conditions test. Unfortunately it is also, for me at least, the most difficult. It shouldn’t be, maybe it is me being lazy.

Method.  Connect an antenna to the UnUn then bond a temperature sensor to the core.

Measure the initial temperature.

Then put a known amount of power into it over a known amount of time and then read the temperature rise. Make multiple readings 10 seconds apart and find the maximum temperature rise – it can be delayed.

From this, and knowing the weight of the core, power in and power lost as joules as heating and the specific heat of ferrite, an efficiency percentage can be calculated.

I really like this test. It tells you in real life how the UnUn behaves, how it handles power with an antenna and what you can expect.

It really is important to know temperature rise as heat is the killer for UnUns. If you get ferrite much past 100 degrees Celsius and maybe less, your cores start to become lossy and behave differently – the SWR may suddenly rise alarmingly. If you don’t stop transmitting the cores may well shatter.

I tested this – I put 50 watts of digital into our QRP Hybrid UnUn  DO NOT DO THIS. after 30 seconds the SWR rose alarmingly and the core was too hot to touch. This was of course no surprise.

One thing; I put 50 watts of SSB into the same Hybrid UnUn and for an hour’s QSO – it ran cool, maybe only rising a few degrees at most. Bear in mind I was transmitting for maybe only 50% of the time. This highlights the difference between SSB and digital modes. An old rule of thumb is that SSB is about 5% to 10% of the rated power as an average – so my 50W SSB was 2.5 to 5 watts of average power.

Second Conclusion

I’ve presented some methods for testing cores in isolation, and for testing complete UnUns.

I’ve described the benefits and shortcomings of each method and presented some real world scenarios.

For me, I want the OZ-Tenna Hybrid UnUns to be the best we can make and hence testing, verifying, measuring is the only way to have some assurance of that.

Without measurement you don’t know.

73 – Play radios. Be happy!

Share This post

Share on facebook
Share on twitter
Share on whatsapp
Share on linkedin
Share on pinterest