# Impedance matching for amplifier output

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. 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 online stores have very cheap capacitor and inductors that are 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)?

• at 500kHz the wavelength is some 600m, I think that you will be fine without any matching. – Vladimir Cravero Sep 11 '14 at 13:39
• But then the power delivered in the load will be 20% of the 'output' power instead of the 50% (with an optimal match). And the RF amplifier probably won't like the lower load at all. – Wouter van Ooijen Sep 11 '14 at 13:44
• @froteros Out of curiosity, what application has got 12 Ω load impedance and requires this frequency and power? – Nick Alexeev Sep 11 '14 at 17:25
• Thanks a lot for the replies. I found that there are antenna matching units that have two knobs to adjust capacitors and another knob for inductor. Do you think I can use those? They are spec'd to work for > 1.6MHz and higher, so I am not sure whether they will work at 500 kHz as well. – froteros Sep 11 '14 at 17:40
• @Nick My application is rather biological. We will be using it to do electrical ablation and the tissues have around this impedance at 500 kHz. You can't use too high or too low frequencies due to some physiological constraints. So, 100-1000kHz is the most suitable band. – froteros Sep 11 '14 at 17:43

## 4 Answers

Yes, you can use an L-C network to match your load. It will be very frequency-specific, but as long as you're operating at a fixed freqeuncy, that shouldn't matter too much. But it does mean that it will need to be "tuned" for your application.

A 2:1 transformer would be a more broadband solution, but you definitely cannot use a power transformer designed for 50/60 Hz power, which has a laminated iron core that would be far too lossy at 500 kHz. A large toriodal ferrite core would be much more appropriate.

It might be more appropriate to ask this on Amateur Radio.SE — some hams work at these frequency and power ranges.

Since this is for a single frequency you can use an L-Network to match the impedance.

The Q you need is $Q= \sqrt{\frac{R_S}{R_A} -1}$

$X_L = Q\cdot R_A$

$X_C = \frac{R_S}{Q}$

Where $R_S$ is the amplifier output impedance and $R_A$ is the antenna impedance

Inserting your numbers of f = 5E5 Hz, $R_S = 50$, $R_A = 12$, we get L1 = 6.79uH, C1 = 11.3nF. Q = 1.78 so it's not horrifically sensitive to frequency, within reason (say 5-10% of nominal).

At 400W out, the peak voltage across C1 will be about 200V (possibly more during transient conditions) so use at least a 600V capacitor.

Using a transformer is a good solution provided your load is resonant at 500kHz.

Here's how OM Ralph tackled it.

http://w5jgv.com/600_meter_ferrite_transformer/index.htm

• Thanks a lot for the reply. Inspiring from that design I cam up with the following transformer: obrazki.elektroda.pl/2414866000_1410565017.png Basically it has ferrite core coil former, such as this: 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? – froteros Sep 12 '14 at 23:50
• You got it! Grounding concept is okay. But grounding to a metal enclosure should be close to the transformer and each connector. – user52238 Sep 13 '14 at 4:21

Definitely check out the link in Nandu's answer... I up voted it since you can follow his build and have something that you're looking for!

(Sources: L Network Article, Capacitor Voltages, also The ARRL Antenna Book 22nd ed. an honest-to-goodness print book if you can believe it)

An L network works for unbalanced loads, what's the nature of your antenna?

Here's a few things you need to be aware of:

The net impedance isn't enough, you need to know if you have any capacitive or inductive reactance components that need matching as well.

Voltages developed across capacitor plates can be pretty stunning, several kV is typical and at 400W I imagine that you could see several 10kV... so RF capacitor selection will matter.

The 220-110 V transformers will be rated for use at about 50-60Hz, and you'll want to verify that they'll hold up to use at 500KHz.

• Wow... 220/110 transformer with iron cores? they are constructed using layers of iron plates to avoid eddy currents and heating up at 50/60Hz. I'm quite sure that at 500 kHz you'll get large amounts of heat! (and little output) – jcoppens Sep 13 '14 at 18:22