50Ω to 75Ω or 300Ω DIY Simple Transformer Plans

[NC1]

With all this 75Ω cable laying around here, I wanted to put it to good use. First I wanted to match it to 50Ω for radio use but could not find any decent (cheap) transformers.

A quick search turned up the following two sites with plans to make it myself, and I will never buy another transformer ever again!

Matching 50Ω to 75Ω

https://www.cv.nrao.edu/~demerson/twelfth/twelfth.htm

It does not get much simpler and easier than that, and I will have it today!

 

[prcguy]

The low loss solutions in your links are not that simple simple and easy, Matching with critical lengths of coax works fine but only over a narrow band of frequencies. You also need to calculate the lengths using the velocity factor of the coax and you never know if you did it all right without testing the assembly on a Scalar Network Analyzer or similar.

Similar problems with the capacitor/inductor matching, you need to calculate the values of the components then tune for a specific narrow band of frequencies.

The resistor matching is simple but very lossy and usually only done to match impedance when making measurements with test equipment. A 50 to 75 ohm matching pad will have about 5.7dB loss, or in other words will loose about 75% of your signal. A 50 to 300 restive matching pad has about 13dB loss and if you made one for high power and fed it with 10 watts you would only have a half watt left.
prcguy

With all this 75Ω cable laying around here, I wanted to put it to good use. First I wanted to match it to 50Ω for radio use but could not find any decent (cheap) transformers.

A quick search turned up the following two sites with plans to make it myself, and I will never buy another transformer ever again!

Matching 50Ω to 75Ω

https://www.cv.nrao.edu/~demerson/twelfth/twelfth.htm

It does not get much simpler and easier than that, and I will have it today!

 

[NC1]

If I use the minimum-loss pad, it's only 0.2dB, which is probably what you would lose with a commercially made product that you spent $35 for - if they even give you the correct loss at all.

Hey, it's Amateur Radio, experimentation is not only encouraged, it is almost a requirement.

I would not dismiss it just because theory says it only does this or that, there is a reason it is called "theory", lol

If Edison had listened to everybody, we would not have had success with the light bulb. What test equipment did he have?

Every now and then somebody has a breakthrough, or finds a new unexpected direction by whatever means it may be. Some things are pretty much set in stone, but I tend to think others are a little more fluid than we are led to believe.

 

[prcguy]

Go back and read your link, a 50 to 75 ohm minimal loss pad has 5.7dB of loss not .2dB. The 50 to 300 ohm version is 13.3dB loss. That's massive when put in a receive path and these types of impedance matching attenuator pads are normally used in test bench setups where the loss is accounted for in the measurement but otherwise doesn't matter.

I have a drawer full of minimal loss pads that I've used for decades in CATV and satellite broadcast measurements and I would never use one in a receive path because of the high loss.
prcguy

If I use the minimum-loss pad, it's only 0.2dB, which is probably what you would lose with a commercially made product that you spent $35 for - if they even give you the correct loss at all.

Hey, it's Amateur Radio, experimentation is not only encouraged, it is almost a requirement.

I would not dismiss it just because theory says it only does this or that, there is a reason it is called "theory", lol

If Edison had listened to everybody, we would not have had success with the light bulb. What test equipment did he have?

Every now and then somebody has a breakthrough, or finds a new unexpected direction by whatever means it may be. Some things are pretty much set in stone, but I tend to think others are a little more fluid than we are led to believe.

 

[NC1]

Well, I'm not going to run away because of one aspect, it takes a lot more to scare me off.
Besides, I have not even tried it my way yet, ha ha

Here is what the beginning of what one link has to say:

"Most signal generators have an output impedance of 50Ω. To align an FM tuner or measure its performance, it's best to match this to the tuner's 75Ω input impedance. Mismatch loss is only 0.2 dB, but a source impedance that differs from the design value may alter the RF input circuit bandwidth or resonant frequency. This can degrade front-end tracking and affect intermod or desensitization measurements."

Mismatch loss is only 0.2dB - so there is my starting point. Furthermore it states "Loss is 5.6 dB for the 5% values shown". OK, so it is for the values THEY show, in ONE example. That leaves a lot of variables they don't show.

I'm not convinced it's a terrible design, it's just that nobody knows how to make it a good design, not yet anyway.

 

[RFI-EMI-GUY]

If I use the minimum-loss pad, it's only 0.2dB, which is probably what you would lose with a commercially made product that you spent $35 for - if they even give you the correct loss at all.

(snip)

where do you find a minimum-loss pad, (with) only 0.2dB? More like 5.7 dB if I recall.

There are BALUN transformers that will readily convert 75/50. 50/75 . 300/50 or 75 Ohms and other ratios with broad band characteristics and low loss. Check Minicircuits for inexpensive low power stuff. For high power, the QST magazine, ARRL Handbook and other suppliers to amateur radio offer many options.

Resources:

Transformers - Mini Circuits

All Baluns - 1:1 Feedline Isolation - Balun Designs

Palomar Engineers® | Why a Balun by Kurt N Sterba

Personally, for receive only applications that are not impedance critical I will use CATV grade RG6 cable without any conversion. But if for example you have to insert filters the impedance again becomes a concern.

 

[prcguy]

A restive or minimal loss pad provides the wanted impedance termination at each end of the circuit and 5.7dB is the minimal loss you can incur with the proper design. You can play with different values of resistors in an L or T or Pi configuration and reach the same impedance goals, but the 5.7dB loss version is the least lossy you can make, that's why its called a minimal loss pad. And pad is part of the term "attenuator pad", which is exactly what it is, a big attenuator.

These weren't designed last week and have been around for almost 100yrs. Its not that someone has not figured out a way to make them less lossy, it simply can't be done with a resistor design.

Sometimes additional loss over theoretical is listed as part of a component spec. For example, a perfect two way divider has a theoretical loss of 3.01029996dB but in reality there are additional losses. If you look in a MiniCircuits catalog for two way dividers it will sometimes say the divider has .2dB loss or similar. That's on top of the 3dB theoretical for a total not to exceed 3.2dB in that case.

Same holds true for minimal loss pads, the mfrs assume you know how much loss they are supposed to have and will sometimes only list the additional loss on top of theoretical. That's probably where you are seeing a .2dB loss spec. In that case you will have a 5.9dB attenuator in line with that specific minimal loss pad.

Not trying to scare you off, just trying to provide information so you don't spend time and money going in a direction then wonder why it has so much loss.
prcguy

Well, I'm not going to run away because of one aspect, it takes a lot more to scare me off.
Besides, I have not even tried it my way yet, ha ha

Here is what the beginning of what one link has to say:

"Most signal generators have an output impedance of 50Ω. To align an FM tuner or measure its performance, it's best to match this to the tuner's 75Ω input impedance. Mismatch loss is only 0.2 dB, but a source impedance that differs from the design value may alter the RF input circuit bandwidth or resonant frequency. This can degrade front-end tracking and affect intermod or desensitization measurements."

Mismatch loss is only 0.2dB - so there is my starting point. Furthermore it states "Loss is 5.6 dB for the 5% values shown". OK, so it is for the values THEY show, in ONE example. That leaves a lot of variables they don't show.

I'm not convinced it's a terrible design, it's just that nobody knows how to make it a good design, not yet anyway.

 

[Ed_Seedhouse]

I don't believe there is any point trying to "match" 75 ohm cable to a 50 ohm output. The "mismatch" is only 1.5:1 and that's considered "matched" for all practical purposes. There will be worse mismatches in the antenna itself.

For a 300 ohm twin line of course you need an impedance transformer (commonly called a "balun").

 

[SCPD]

Maybe this is an over simplification to throw into this technical discussion, but I have never liked using any form of turner to match anything to anything. We've run 72 ohm cables into and out of 50 ohm systems all the time-- Yes, I know know, it isn't always a perfect match, but usually under 1.5:1 SWR. I would rather deal with that known quantity than calculate and live with the other possible loses, which are usually much higher with matching networks.
.
NC1-- I like your thinking.... never buy another RF transformer again ! .....
.
.......................................CF

 

[WA8ZTZ]

From a practical standpoint, the mismatch when using, say, 75 ohm RG-6 with a 50 ohm input will be unnoticeable on receive and probably a non-issue on transmit.

However, if you absolutely, positively desire a pefect match, then the twelfth wave matching section really does work.

 

[NC1]

From a practical standpoint, the mismatch when using, say, 75 ohm RG-6 with a 50 ohm input will be unnoticeable on receive and probably a non-issue on transmit.

However, if you absolutely, positively desire a perfect match, then the twelfth wave matching section really does work.

That is mostly what I thought too.
Make up any loss with gain from the antenna.

They want far too much $$$ for a commercially produced transformer. I'll make one out of a few pennies of scrap I already have, and a few minutes of my time.

 

[n5ims]

One thing I didn't see is what type of antenna are you wanting to use that 75 ohm coax on? The standard HF dipole is often closer to 75 ohms than it is to 50 ohms. Most commercially made antennas are designed with a 50 ohm impedance, basically so it will appear to match the typical radio's spec. That said, if your radio (like many current generation HF radios) have a built-in antenna tuner, the use of either 50 or 75 ohm coax should be fine. For VHF or UHF radios though, you should probably stick to 50 ohm coax since those radios typically don't have built-in antenna tuners.

 

[Karl-NVW]

Note too that if the T-line is carrying power, all components of the matching device must be rated to carry that power continuously. Otherwise you can let the magic radio smoke out, which is followed by going out to buy more matching network components. And possibly a very costly set of output transistors for your rig!

 

[ko6jw_2]

Have you considered just hooking up the coax directly. The losses from some of the impedance transforming schemes outlined here will be greater than doing nothing. If you are receiving only, there should be no problem.