Suppliers Should Know Better
I’ve just been looking around the internet recently and I saw some head-scratchingly poor advice and, ah, dodgy tech onfo.
Seriously, I expected better from what I thought was a professional supplier of baluns.
Here are some claims and my refutations.
You Can't Get Good Efficiency With QRP Designs
My tiny tiny little Hybrid UnUn is around 90% efficient. That beats many if not most auto tuners.
My not much larger classic 49:1 50 watt UnUn which I regularly push 100W SSB though is over 92% efficient on 40m. Tested. The cores weigh 12.8 grams total. A not much larger again version is even better.
Now. The testing is really quite simple. It takes a couple of hours maximum and a calculator. I bond a temperature sensor to the cores. Then I put digital power into the UnUn for a period of time. Then I measure temperature rise. Then it is a simple mathematical formula to calculate efficiency.
So the claim of low efficiency by virtue of being QRP is easily proven top be just plain crap.
Higher Rating UnUns are More Efficient
BZZZANG – Wrong again.
Well, if you (the supplier I looking at) produce a crap QRP design, then a higher power design may well be better.
Most likely they copied a larger core design directly onto a small core. From my testing it will probably be worse. That just means the design has not been adjusted accordingly.
Quite frankly the most popular design I’ve seen on the internet has quite low efficiency so the bar is not set high. What is low? It is subjective.
I personally feel anything below 80% I would call low and below 60% which is easily available from some dsigns is appalling.
What is good? Again it is subjective. I feel if you can get to 90% that seems to be pretty good. If you can get to 95% then wow – well done and I’d feel suspicious it is too high. But for heavens sake, document how the efficiency was calculated. There are 4 methods of testing – and a blog post is in production on that topic.
There was some mention of larger core sizes being more efficient. Nope. They’re not. It will probably be that their magnetic properties better suit the design they are testing with – and a “design” has not been done properly for QRP and or vice versa. In general larger cores can have a longer magnetic path and that brings loss. Try reading Owen Duffy’s blog on core geometry.
I’ve found through testing and measurement that getting past about 95% efficient, which is a very small improvement on a simple QRP design that involved testing and measurement also, is pretty much impossible. For me that is – and in testing in real life using heating tests – so this is based on actual test and measurement.
Saturation And Core Failure
The claim is that core saturation is the prime cause of core failure.
Nope. It is not the prime cause.
You see, at amateur radio typical power levels saturation should not be a real cause of failure particularly with the larger and sensibly sized cores. Look at the data sheet. It takes a LOT of power to lead to core saturation.
The caveat is simple though. If you use a half inch diameter core and put 100W digital in it – stand back – it will get really hot and really fast. Is it saturating? Maybe. Loss from poor efficiency generating heat will be a huge factor even ignoring possible saturation. What would you expect from a dinky little core?
My smallest QRP core, similar in length to my pinkie finger nail takes 50W of SSB with ease. No saturation.
HEAT is the killer. Heat comes from loss. Loss comes from a number of places. Core loss is a consideration. More typically it is the way the system is operating and what you’re trying to do.
For a 1:1 balun if you have a high common mode current and use digital modes, then the heat it is required to dissipate can be far too much for just about any core.
Also, as a core gets hotter, and rightly mentioned, I am sure that efficiency will drop off gradually and hence more heat will be generated. It will be a rapid vicious cycle leading to over heating.
One thing that so many people forget. If you use a “tuner” sure you match your transceiver to the coax. But at the antenna the SWR can be 5:1 or worse on some antennas (ie operating an OCF out of design band). What is happening to a balun at the feed point? The answer will be the impedance will be hugely different for what the balun is designed for and hence loss will be – guess what – large!
I’ve made a lot of 4:1 current baluns. I spent months on this. At 200 ohm resistive loads they come in at 98.5% efficient. I make this claim because I went to a lot of trouble to get a 100 ohm transmission line happening with Teflon sleeved wire. I TESTED THEM by measuring heat rise and doing the calculations. BUT at 135 ohms which is typical of a lower OCF antenna (mine actually!), efficiency drops to 85%. AGAIN I measured it, and with a 135 ohm, dummy load.
Sure. Your numbers may well be different. But hell, I measured it. I did the calcs.
If you run that same “highly efficient” 4:1 at a high SWR at the feed point I can just about guarantee it will get very very very hot.
Just because you use a tuner, it does not mean there isn’t a hell of a lot of loss leading to core heating AND WORST OF ALL it is hiding the mismatch at the antenna and in the coax.
Slowly Rising SWR Should Be a Hint
The article mentioned, and I saw internet references, to an SWR that was rising slowly during a number of overs, or with digital modes like FT8 after a period of time.
This is a real hint.
If it were saturation then the SWR should change very quickly, like in a few seconds at most.
If it is a heating problem then the core will get more and more warm over time and hence SWR will slowly rise.
The supplier wants to sell you 3kW and 5kW baluns. They cost more.
Along the way they make a lot of invalid claims and lead the casual reader to wrong conclusions.
I’ve built a lot of UnUns and Baluns. I’ve tested them by 4 methods. I’ve run the numbers. I’ve measured in real life use with antennas.
But the big thing, they are wrong in so much that they claim. That leads me to suspect so many of their other claims, notably about power rating, are wrong.
What does a 3kW rating mean? Do you have a license exemption for 3kW? What common mode current would it handle? What core heat rise would be expected with say 1 amp of common mode current? Is there a published specification for how much power the balun can dissipate before failure? Where are the numbers?
I Expected Better From A Major Supplier
I’m just one Amateur Radio operator in nowhere’s-ville Australia.
I’m doing some simple tests here in Australia, not America where my natural bias is to believe they can do more and better.
I constantly see the FT240-43 cores with the “2:14” winding ratio. My “Test-#1” (2 turn link and some calcs) shows this to have 70% efficiency on 80m. I’ll be doing some tests on these, the full “4 efficiency tests” in a subsequent blog post.
It is just me here. Not a team. I’ve spent limited time looking at cores, selecting and optimizing, and actually measuring and testing what I’ve built. Any I’ve built maybe 50 different designs. Getting to 90% is achievable, and better as well. I have one design showing over 95% efficiency I’ll be using a temperature sensor to verify (yes, it is too good to be true – I’m not believing it yet).
It’s not hard. It is a bit of work. It does bring a better product.
Maybe I shouldn’t blog when I’m grumpy!