Impedance Matching of the RF Amplifier Output

[froteros]

Hi All. I hope this is the appropriate forum to ask this. I needed some basic advice about impedance matching. I have an RF amplifier with 50 ohm output impedance, capable of generating 400W output power at 500kHz. I am trying to deliver this to a 12 ohm load.

I looked online and figured out that I need to match the impedance of the amplifier to the load. My question is what is the best way to achieve this? Would it be possible to simply use a capacitor and inductor and put an L section to the output of the amplifier? I saw that digikey has very cheap capacitor and inductors that is rated for very high voltages.

If that is not going to work is it possible to buy any transformers that would do that? For instance I saw that there are many cheap 220v-110v voltage converters, which should have 2:1 coil ratio. So in theory that should convert 50 ohm to 12.5 ohm and matches to my load. Will these work at 500 kHz as required for my application? Or are there other type of transformers that I can use for this purpose (there are many many types of transformers and I got confused which ones are suitable for impedance matching)?

Thanks a lot for the help.

 

[THINKER_KAM_MAN]

just a suggestion

use you first method
or use this equation TURNS RATIO = Sq Root(Rin/RL) = 2.04:1 in your case
then as you frequency is 500Hz, use a ferrite core ring.
T400, may be. or a yellow instead of red core (refer to the ferrite core manufacturer Data sheet for the best choice)
and wind you self a matching transformer.
you can use thick electrical wire to wind it
or say twisted strands of smaller wire
space the coils turns evenly around the ring
you will need a SWR meter to test it.
if you can't be bothered with all that then what you need to
buy is an High Power ATU.

 

[FvM]

A transformer might turn out as the best option for the relative low frequency, but it has to be specially designed for the application. Mains frequency transformers are completely unsuitable.

Fortunately the transformer can be made with a small ferrite core, but it needs a special winding design, copper foil or litz wire.

Alternatively a tuned LC circuit can be used, preferably a CLC "pi" filter.

 

[BigBoss]

The standard transformers don't work at 500kHz but your may find ferrite core transformers (or you can do it yourself too) that would work at that frequency.Don't consider L-C matching etc. circuit because power is really high, small components ( except special ones) can not resist to this power level.
The optimum way is to use a ferrite core transformer with impedance ratio 1:2

 

[Borber]

For intermezzo you can calculate pi LC matching network to see how does it look like and what value of inductances and capacitors are involved:
https://home.sandiego.edu/~ekim/e194rfs01/jwmatcher/matcher2.html

 

[froteros]

Thanks a lot for the replies. Antenna tuner unit is an interesting idea. I looked around and they are selling many used ones in the $100-200 range. However, they are usually designed for AM bands and rated for (1.8-30Mhz). So, do you think they will work at 500 kHz? As an example I am looking to MFJ-989C right now.

They seem to have three knobs, transmitter, antenna and inductor. I am guessing this makes a matching circuit with two capacitor and one inductor. Since they are designed to match to antennas, I understand that they assume that they are going to increase the source impedance whereas I require to decrease the source impedance from 50 to 12 ohm. So, will they operate in this range?

I am guessing I can answer these question if I knew what the actual values for the knobs are in terms of capacitance and inductance but that is not known as far as I tell.

 

[FvM]

The standard SW match boxes will not work for 500 kHz because the available capacitor values are two small. You may be able to modify an existing matcher by adding parallel capacitance.

A variable matching network woulb be particularly useful if you're not sure about the actual load impedance and it can be complex. Fixed 4:1 matching can be better done with a transformer.

 

[THINKER_KAM_MAN]

just a suggestion


The impedance of an antenna can be any value, high or low.

if the ATU has 2 knobs its called a long wire tuner
one knob adjust capacitance, the other adjust inductance
if it he ATU has 3 knobs ore more then its more complex and its a pi match or other
it has one knob labelled band or just letters and it adjust the inductance, the other two knobs adjust the output capacitance. and
the load cap. so if you try to use an ham ATU and you can’t get a match then you
will need to add more inductance in series with the existing inductance on the lowest band. This inductance can be made using a ferrite ring core. And also extra fixed caps. Across C and CL,



SEE: **broken link removed** FOR MORE OPTIONS

You could make you own by referring to the above page and calculate the ballpark values: modifying an ATU is useful because you can tune the match because as said
The 12R load could be any thing.

 

[chuckey]

Are you sure that the 12 ohm load is purely resistive? I have just had a quick go at a PI matching section based on a 100µH, and I reckon that the output Cap is 4.8 NF and the input cap is 1.2 NF. The reason I chose 100µH as hi power RF chokes are available as the current in the output is over 5 A. If you increase the inductance then smaller caps can be used. What sort of bandwidth are you after?
Frank

 

[froteros]

@chuckey: Thank you for the calculation. I will be using pretty much single frequency of 500 kHz, it won't be modulated or anything. Also the duration of the operation will be at most 1-2 seconds, so it will not be a continuous output.

After doing more search I come up with the following simple transformer design:
https://obrazki.elektroda.pl/2414866000_1410565017.png

Basically it has ferrite core coil former, such as this:
https://www.digikey.com/product-detail/en/B65714K1020T1/495-5189-ND/3913978

Then I wire 10 coils using a thick litz wire. Then I tap to the tenth coil from the input and to the fifth coil from the output. Can you comment whether you think this would work or not? Also do you think I should combine all the grounds like the way it is drawn there?

 

[BigBoss]

https://obrazki.elektroda.pl/2414866000_1410565017.png[/url]

Basically it has ferrite core coil former, such as this:
https://www.digikey.com/product-detail/en/B65714K1020T1/495-5189-ND/3913978


Then I wire 10 coils using a thick litz wire. Then I tap to the tenth coil from the input and to the fifth coil from the output. Can you comment whether .
you think this would work or not? Also do you think I should combine all the grounds like the way it is drawn there?

You cannot define this transformer so simple.You should check different ferrite components ( permeability,max. magnetic flux density,power handling capacity ect.) then you can design your transformer by using all these datas regarding to your goal.Because you talk about 400W RF power and it should be taken into account.
Read something about ferrite core transformers and ferrite materials and then go forward.

 

[FvM]

After doing more search I come up with the following simple transformer design.

I'm not sure which ferrite core shape is addressed in your idea. If you mean a magnetical open reel core with "coil former", that' won't be a good idea. The transformer should implement a closed magnetical path without air gap. It can be e.g. a torroid, E or pot core. For the 2:1 windings ratio, a bifilar design is a simple way to achieve best coupling (= low leakage inductance). And yes, it should use litz wire.

 

[E Kafeman]

Maybe you already have a fitting core and windings. A PC PSU have pretty much similar power and frequency range.
With some luck can existing windings be reused with minor adjustments. 5V and 12V windings can be in fitting current range.
Current at 12 Ohm will be rather high. A too small core can be saturated which makes its transformation ratio nonlinear and will building up heat very quick.
Wire serial resistance will also cause heating of the core which limits max current.
If you have core data, can max allowed current be calculated like this: https://www.mag-inc.com/design/design-guides/Inductor-Design-with-Magnetics-Ferrite-Cores

 

[froteros]

I'm not sure which ferrite core shape is addressed in your idea. If you mean a magnetical open reel core with "coil former", that' won't be a good idea. The transformer should implement a closed magnetical path without air gap. It can be e.g. a torroid, E or pot core. For the 2:1 windings ratio, a bifilar design is a simple way to achieve best coupling (= low leakage inductance). And yes, it should use litz wire.

This was the design I inspired from:
https://w5jgv.com/600_meter_ferrite_transformer/index.htm

I was hoping to use a similar toroid such as this one:
https://www.digikey.com/product-detail/en/B65714K1020T1/495-5189-ND/3913978

But you are right I should search more about the power capacity, winding etc.