Designing an L-Network by Hand: Q From the Resistance Ratio
For an L-network matching two resistances, Q = sqrt(R_high/R_low − 1), so 50 to 200 ohm gives Q = 1.732. The series reactance is Q × R_low = 86.6 ohm and the shunt reactance is R_high/Q = 115.5 ohm, which at 1 GHz means 13.78 nH and 1.378 pF.
Engineer’s note: Hand-calculating an L-network, the first thing that falls out is Q, and the bigger the impedance ratio, the higher Q climbs. A high Q means the match is needle-narrow, and a 2 to 3% part tolerance can send production yield down. If Q comes out too high, drop the single L-network and look at other structures.
For practitioners
An L-network matches two unequal resistances at one frequency with one series reactance and one shunt reactance. With only two elements there is no free parameter. Once the two resistances are fixed, the network Q is fixed, and every component value follows from it. That makes the L-match the one matching network you can design completely on paper in a few lines. It's worth doing by hand at least once before you trust a calculator.
Where the Q comes from
The whole design rests on series-to-parallel conversion. At a single frequency, a resistance R_s in series with a reactance X_s behaves exactly like a resistance R_p in parallel with a reactance X_p, where:
So putting a series reactance on a resistance raises its parallel-equivalent resistance by a factor of (1 + Q²). An L-network uses exactly this. The series element turns the low resistance into a parallel equivalent equal to the high resistance plus an unwanted parallel reactance. The shunt element then cancels that reactance.
Set R_s = R_low and R_p = R_high, then solve for Q:
The same Q describes both branches. The series branch has Q = |X_s|/R_low and the shunt branch has Q = R_high/|X_p|, and both equal the value above. This is the loaded Q of the match. A larger resistance ratio forces a higher Q, which means a narrower bandwidth and more sensitivity to component tolerance. You don't get to trade that away with an L-network.
Step 1: Q from the resistance ratio
Take the worked example of matching 50 ohm to 200 ohm.
- Resistance ratio: R_high/R_low = 200/50 = 4
- Q = sqrt(4 − 1) = sqrt(3) = 1.732
That is the only real design decision, and the resistances make it for you.
Step 2: From Q to reactances, and which side gets which
The two reactance magnitudes come straight from the definitions of Q:
For 50 to 200 ohm:
- |X_s| = 1.732 × 50 = 86.6 ohm, in series, in line with the 50 ohm termination
- |X_p| = 200 / 1.732 = 115.5 ohm, in shunt, across the 200 ohm termination
The placement is not optional. A series element raises the parallel-equivalent resistance of whatever it is attached to, and a shunt element lowers the series-equivalent resistance. To step 50 ohm up to 200 ohm, the series element must work on the 50 ohm side and the shunt element must sit across the 200 ohm side.
The two reactances must also have opposite sign, one inductor and one capacitor. The shunt element exists to cancel the reactive part left behind by the series element, and only an opposite-sign reactance can do that.
Step 3: Component values at the design frequency
Convert reactance to components with the usual expressions:
At f = 1 GHz, with a series inductor and a shunt capacitor:
- L = 86.6 / (2π × 10⁹) = 13.78 nH
- C = 1 / (2π × 10⁹ × 115.5) = 1.378 pF
If you exchange the element types and use a series capacitor with a shunt inductor, the reactance magnitudes stay at 86.6 ohm and 115.5 ohm, so the values become:
- Series C = 1 / (2π × 10⁹ × 86.6) = 1.838 pF
- Shunt L = 115.5 / (2π × 10⁹) = 18.38 nH
Checking the match by hand
A quick check catches sign and placement errors. Work from the 200 ohm load toward the 50 ohm port using the low-pass solution.
Start with the shunt capacitor across 200 ohm. Its reactance is −j115.5 ohm, so its susceptance is +j(1/115.5) = +j0.00866 S. The admittance of the load plus capacitor is:
Y = 0.005 + j0.00866 S
Invert it. The magnitude squared is 0.005² + 0.00866² = 0.000025 + 0.000075 = 0.0001, so:
Z = (0.005 − j0.00866) / 0.0001 = 50 − j86.6 ohm
The resistance is already 50 ohm. What's left is −j86.6 ohm of capacitive reactance, and the series inductor adds +j86.6 ohm to cancel it, leaving exactly 50 ohm. The numbers close cleanly because Q² = 3 makes 1 + Q² = 4, the same as the resistance ratio.
Two solutions: low-pass or high-pass
Both solutions match 50 ohm to 200 ohm at 1 GHz with the same Q and the same reactance magnitudes. They differ away from the design frequency. The series-L, shunt-C version passes DC and attenuates harmonics. The series-C, shunt-L version blocks DC on the series path and provides a DC path to ground through the shunt inductor. Which one you pick depends on biasing, harmonic content and which values are practical. The trade-offs are covered in L-Network: Low-Pass or High-Pass?.
To check your hand results, the L-Network Matching Calculator gives the same four values. If the fixed Q of an L-network is too high or too low for your bandwidth, you need a third element. See RF Matching Networks: L, Pi and T.
In plain words · for students and beginners
The L-network in plain words
A ramp between two heights
Picture the 50 ohm side as a short step and the 200 ohm side as a step four times taller. Energy moves easily from one step to the other only if they are at the same height. An L-network works like a ramp between them. It tilts the 50 ohm resistance up until it looks like 200 ohm. (Resistance here means how strongly a circuit resists current. Matching means making both sides look the same, so power isn't reflected back.)
The ramp has two parts. The coil (an inductor, a part that resists fast-changing current) sits in line with the low side. That's the series part. The capacitor (a part that stores charge between two plates) sits across the high side. That's the shunt part, meaning it connects from the signal line to ground. The ramp also works in the other direction: seen from the 200 ohm side, the same network tilts 200 ohm down to 50 ohm.
The analogy breaks down in one way. A real ramp works for anything that rolls along it. This ramp is only the right shape at one frequency. Move away from 1 GHz and the coil and capacitor change their behaviour, so the match slowly gets worse.
The same example, step by step
- Divide the big resistance by the small one: 200 / 50 = 4.
- Subtract 1 and take the square root: sqrt(3) = 1.732. This number is Q. It tells you how steep the ramp has to be.
- Series part: Q × 50 = 86.6 ohm. Shunt part: 200 / Q = 115.5 ohm. These are reactances, the "ohms" that coils and capacitors show at a given frequency.
- Turn the ohms into parts at 1 GHz: a 13.78 nH inductor in series and a 1.378 pF capacitor in shunt.
If you swap the inductor and capacitor, you get a second working design with the same Q and the same ohm values.
Mistakes beginners make
- Putting the shunt part on the wrong side. It must go across the higher resistance. On the low side the network can't match at all.
- Using two inductors or two capacitors. One must be a coil and one a capacitor, because each one cancels the other's leftover effect.
- Forgetting the frequency. 13.78 nH and 1.378 pF are right only at 1 GHz. At another frequency, redo step 4.
- Expecting to choose Q. With two parts, the resistance ratio sets Q. To change it, you need a three-part network.
Remember
- Q = sqrt(R_high / R_low − 1). For 50 to 200 ohm, that is 1.732.
- Series reactance = Q × R_low. Shunt reactance = R_high / Q, placed across the high side.
- Swapping L and C gives the second solution with the same Q.
Related pages on RF View
Frequently asked questions
How do you calculate the Q of an L-network?
Divide the higher resistance by the lower one, subtract 1 and take the square root: Q = sqrt(R_high/R_low − 1). For 50 ohm to 200 ohm the ratio is 4, so Q = sqrt(3) = 1.732.
Which side of an L-network gets the shunt element?
The shunt element goes across the higher-resistance termination, and the series element sits in line with the lower-resistance termination. If you swap them, no component values will give a match between two pure resistances.
Can I choose the Q of an L-network myself?
No. With two elements and two fixed resistances, Q is set by the resistance ratio. If you need a different Q, for example a narrower or wider match, use a pi or T network, which adds a third element.
What is the difference between the low-pass and high-pass L-network?
The low-pass version uses a series inductor and a shunt capacitor. The high-pass version swaps them for a series capacitor and a shunt inductor. Both have the same Q and the same reactance magnitudes.