**Discussion on the Impact of Right-Angle Routing in PCB Design**

The debate over the impact of right-angle routing on signals in PCB design has been raging online for years—seriously, how many years has it been? Frankly, I never intended to discuss this topic because no one in the industry who knows their stuff really cares about it. However, many outsiders or those with superficial understanding keep spreading misinformation, stating it with absolute certainty, leading to countless beginners constantly asking me about this.

Just look at some of the ridiculous statements popular online. For example: "Current is like water flow; right-angle routing is like a right-angle river bend, where water easily rushes out of the channel and erodes the banks over time."

Uh... your imagination is quite vivid, and your ability to fabricate is impressive. Tomato Novel's fantasy section desperately needs talents like you with such wild creativity. Discussing low-level stuff like PCBs is a waste of your talents.

Back to the topic. The essence of right-angle routing in PCBs is that it creates an impedance discontinuity at the corner. But how significant is this discontinuity? Let's calculate it.

The following calculation is based on a JLCPCB 4-layer board, model 7268, with 1 oz outer layers, 0.5 oz inner layers, a board thickness of 1.6 mm, and outer layer single-ended impedance.

First, we draw a straight line 13.7 mil wide on the top layer. The calculated theoretical single-ended impedance is 50.0829 Ω (I emphasize "theoretical" because the actual impedance of a manufactured PCB is not the same as what the impedance calculator gives).

We then create 135-degree and 90-degree bends using this 13.7 mil trace and measure the width at the widest point of the corner, as shown in the figure below.

[Figure showing trace width measurement at bends]

The measurement shows:

*   For the 135-degree bend, the widest point is 14.3 mil. The theoretical impedance at this point is 48.9742 Ω, representing a 2.3% discontinuity compared to 50.0829 Ω.

*   For the 90-degree bend, the widest point is 16.5 mil. The theoretical impedance here is 45.3267 Ω, representing a 9.5% discontinuity.

This might seem like a large percentage, but let's put it in perspective. Even with JLCPCB's impedance control service (for an additional fee), they only guarantee within 10%. So, at what speed would this discontinuity become noticeably impactful? In reality, even for PCIe 5.0 with a Nyquist frequency of 16 GHz, this single-point 9.5% discontinuity is almost negligible. This is because the impedance change is localized at a single point. Compared to the discontinuities introduced by vias, BGA fanouts, or connector pads, it's insignificant—trivial.

**Conclusion 1:** Right-angle routing does have an impact on high-frequency signals. However, this impact is negligible within the context of overall signal path losses. Therefore, it's essentially irrelevant.

After hearing this, you might think, "Should I route everything at right angles then?" Of course not. This leads to the second point: **Is there any benefit to right-angle routing?**

[Figure showing three routing patterns between two points: free-form (direct), 135-degree angles, and 90-degree angles]

Look at the three routing patterns in the figure above. All three connect the same two points, only the routing style differs.

1.  **Free-form routing:** The shortest distance between two points is a straight line—basic knowledge. Length: 24.513 mm.

2.  **135-degree routing:** Total length: 26.4 mm, an increase of **7.7%**.

3.  **Right-angle (90-degree) routing:** Total length: 33.19 mm, an increase of **35.4%**.

The answer is clear. Right-angle routing isn't "forbidden"; it's just that using it offers no advantage—it simply makes traces longer. While free-form routing is the shortest, PCBs also need to accommodate components. Free-form routing can increase layout complexity. Both 135-degree and 90-degree routing facilitate easier trace adjustment, and components can also be placed at 45-degree angles. Therefore, the primary reason for using 135-degree (or 45-degree) routing in PCBs is that it represents a compromise between layout difficulty and trace length.

**In summary:** Right-angle routing has no fundamental impact on signals, regardless of how high the speed is. There is no universal rule in PCB design that says "right-angle routing is forbidden." The reason most people avoid it isn't because of signal integrity concerns; it's simply because right-angle routing offers no benefit and often just increases trace length.

Any talk about it significantly affecting signals is mostly drawing the target after shooting the arrow—creating a problem to fit a preconceived solution. Simplifying the complex is a mark of competence! Making the simple seem complex is mystification. No matter what tutorials you've read or what summarized experiences you've heard, if the subject seems to become more mysterious the more you learn, you're being misled. Things seem difficult only when you don't understand them; once you do, they're not difficult at all!