Just How Constant Is the Dielectric Constant of PCB Materials?

One of the crucial material factors that we address in our classes is the dielectric constant or relative dielectric constant, e_r. This is sometimes called Dk by laminate suppliers. Sometimes, product developers are not clear as to the role the dielectric constant of PCB materials plays in a design, how to measure it, how to account for it, how it varies with frequency, and how to know if the dielectric constant data provided by the laminate manufacturer is accurate and reliable.

This article will address the foregoing topics and describe why the dielectric constant of PCB materials will play a significant role in determining the overall success of a given design. 

If you're scanning a dielectric constant table for quick comparisons, remember that PCB dielectric constant values depend on frequency, construction, and measurement method.

Overview: Dielectric Constant of PCB Materials

The dielectric constant of a vacuum is, by definition, 1. The dielectric constants of laminate materials, other than a vacuum, are compared to a vacuum. This comparison results in a relative dielectric constant, e_r, that expresses the effects of these materials on the capacitance of a structure such as a parallel plate capacitor as compared to a vacuum. Dielectrics also slow down electromagnetic fields traveling through them. Engineers often consult a table of dielectric constants to compare candidate laminates for a stackup.

Salient points to keep in mind include:

• The dielectric constant of a vacuum is 1. All laminate materials have dielectric constants higher than 1.
• The common method for measuring e_r is the parallel plate method at 1 MHz. As shown below, for transmission line design, a more useful method is to calculate the e_r by determining the signal velocity in the dielectric.
• The dielectric constant varies with frequency in all PCB materials.
• The dielectric constant in a laminate is a function of the glass-to-resin ratio. If there is more resin, the e_r goes down. If there is less resin, the e_r goes up. That’s why it’s important to know the glass-to-resin ratio as that will help determine the e_r
• The terms Dk and e_r are used interchangeably and they mean the same thing.
• The dielectric constant is really a complex number, and there is an imaginary part that is sometimes called the dissipation factor, or Df. Dk and Df together determine the loss tangent. For now, we'll focus only on Dk as it is normally a starting point for most designs.

Calculating and Measuring er

Equation 1 is the equation used for determining the e_r of a given material. You can use the velocity of a test signal and the speed of light in vacuum to calculate e_r:

Here, V is the velocity at a given frequency, e_r is the relative dielectric constant and C is the speed of light. Note that the square root of this quantity is the refractive index of the material (again, we've ignored Df here for simplicity), which most are probably familiar with from physics classes. So in other words, a measurement of signal speed in the dielectric gives you the dielectric constant.

There are multiple methods for measuring velocity and thus determining the dielectric constant. In practice, signal velocity cannot be measured directly and instead it must be calculated from some other measurement.

A simple method is to take a uniform transmission line and terminate it with a high reference impedance at one end. You can then use a TDR measurement to determine the round-trip time through the transmission line. A TDR instrument sends a pulse into one end of the transmission line, and then it detects a strong reflection off the large impedance mismatch at the other end of the transmission line. The time between signal injection and reflection is double the travel time for the pulse. Taking the line length and the one-way travel time gives the signal velocity; then using Equation 1 above gives the dielectric constant.

This gives you a measure of the propagation time and signal velocity for a wideband pulse, but not for a single frequency. In some ways, it is a more accurate representation for the velocity of a digital signal. To get the signal velocity and dielectric constant at a single frequency, you would need to source and measure the reflection of a sine wave, which is not normally possible in a TDR measurement. What the TDR measurement is actually giving you is the group velocity, or the velocity of the overall pulse due to superposition of its traveling Fourier components.

However, it is possible to use a VNA to get the S-parameters, then the propagation time can be determined from the phase of the S21 plot. By taking the phase data from of the S21 plot, a derivative can be calculated as a function of frequency, and this will give the propagation delay as defined in Equation 2. Read this article to see how to do this measurement/simulation for a via structure.

The propagation delay plot is given across the frequency range where the VNA measurement was taken. If you are doing the same measurement in simulation, then the same procedure is used. Once the propagation delay is found, the distance between the ports is used to get the wave velocity and the dielectric constant at each frequency in the measurement range.

A very important point to note is that the dielectric constant will depend on two factors:

• The copper roughness value, as described here and by designers like Bert Simonovich
• The structure used to define the transmission line; this is why microstrip, coplanar waveguide, grounded coplanar waveguide all have an effective dielectric constant measurement, while striplines give the actual dielectric constant for rough copper.

These are just two methods that give you dielectric constant measurements in either the time domain or the frequency domain, and they are worth mentioning because off-the-shelf equipment and simple lab setups can be used to perform these measurements on test coupons. There are more specialized methods which the material manufacturers will use and which are prescribed in IPC standards: 

• Ring Resonator
• Delta Line Pair
• Time-Interval for Impedance-Discontinuity Pair
• X-Band Clamped Stripline Resonator
• Parallel Plate Capacitor Method
• Bereskin Stripline Method

Dielectric Constants of PCB Materials

Dielectric constant table 1 shows dielectric constants of PCB materials and their corresponding wave velocities. Again, note that the wave velocities depend on the measurement structure and the roughness of the copper that was used to determine them. When interpreting dielectric constant PCB data in such tables, remember that structure and copper roughness strongly influence effective values.

Notice the qualifier at the bottom of this figure states that the dielectric constant is a function of the glass-to-resin ratio and the signal frequency. The measurements in this slide were made with a resin constant of 55% at 2 GHz (more about this below).

Figure 1 shows the e_r versus frequency of various laminates.

These are the classic four types of materials along with that somewhat confusing catchall thing called FR-4.  This chart shows that the dielectric constant goes down as the frequency goes up (note that this plot only extends up to 6 GHz). It should be noted that the thin lines represent a resin content of 42% (this is how all cheap materials are made). It’s from this measurement that the standard value e_r = 4.7 was determined because, at 1 MHz, the e_r is approximately 4.9. In reality, no real material has this dielectric constant.

As can be seen, with a 55% resin content, the e_r goes down. As noted below, 55% is no longer what would be designated as a high resin content. As can be seen in Figure 2 the dielectric constant versus frequency curve goes down with frequency and it flattens out at about 2 GHz.

A note of caution: If you use the e_r value at 1 MHz to calculate impedance, but your product will be operating at 2 GHz, you are starting out your design process with an error and that error will be propagated throughout the entire design process. It used to be a challenge to decide which frequency should be used for a particular design but the speed of modern edges is so fast now (2 GHz and higher) that this is no longer a factor of concern.

If a product developer uses the e_r calculations from a PCB fabrication facility, it’s important to know what frequency that fabricator is using for their stated dielectric constants. If that facility is not using 2 GHz and higher it’s wise not to have any confidence in their numbers. In order to ensure that a design will operate as specified, it’s mandatory that the fabricator provides specific frequency information along with the specific resin content for the cited laminates.

What Laminate Manufacturers Provide

All laminate manufacturers publish the  e_r for  laminate materials they produce. Figure 3 provides an example of the types of information, including e_r data, for FR408HR prepreg materials produced by Isola Group. Not all materials manufacturers offer this level of information; some will not have the information at all, or they may only take two frequency points (100 MHz and 10 GHz, for example) for the dielectric constant. Some companies will not tell you the test method and you will not know if the dielectric constant is corrected for roughness, resonance in the measurement structure, etc.

Figure 3. Prepreg laminate characteristics for a commonly used material from Isola (FR408HR).

Figure 3 is just an example of a high-performance FR4 laminate and it illustrates the typical laminate table that an engineer needs to have in order to create a good, workable stackup that results in an accurate impedance for a PCB under development. The information contained in this figure contains reliable data and it demonstrates how the e_r varies with frequency. Note that the Dk value is only provided at 3 different frequencies in this case. Also note that the e_r varies with laminate thickness because different thickness laminates have different glass-to-resin ratios.

It’s important to note that there is no value under 100 MHz in this table. Good laminate manufacturers know that data under that number is of no value. In fact, if the laminate manufacturer cites data that they designate as 1 MHz, it’s a good idea not to trust that information and it’s time to find a more reliable laminate supplier.

Another important point is that laminate manufacturers do not use a TDR trace to calculate the dielectric constant of PCB materials. You can certainly do this your self with a test coupon at a few frequencies, but this is not ideal. More sophisticated methods that are specified in IPC standards, and the value reported for the laminate's e_r value depends on the measurement method. Take a look at this podcast with Jon Coonrod to learn more about the Dk and Df values quoted in laminate datasheets.

Summary

Understanding the elements that factor into the dielectric constant of PCB materials is key to ensuring the correct laminate is selected for the product being designed. The data provided by laminate suppliers is a good place to start and it can be trusted as long as the frequency and resin content are correct. Keep a curated dielectric constant table for the laminates you use most, and verify PCB dielectric constant figures at your operating band.

When you need to calculate the effect of e_r on transmission line impedance in your next PCB, you can use Altium Designer available within Altium Develop and the integrated field solver from Simberian. This integrated field solver uses standard models to determine e_r and impedance in your layer stackup, and it helps you perfectly size your transmission lines to have required impedance.

Whether you need to build reliable power electronics or advanced digital systems, Altium Develop unites every discipline into one collaborative force. Free from silos. Free from limits. It’s where engineers, designers, and innovators work as one to co-create without constraints. Experience Altium Develop today!

References

• Ritchey, Lee W., and Zasio, John J., “Right The First Time, A Practical Handbook on High Speed PCB and System Design,” Volume 1.
• Ritchey, Lee W., Course Slides, “2-Day Signal Integrity and High Speed System Design,” training class.

Frequently Asked Questions

What is the dielectric constant (er or Dk), and why does it matter in PCB design?

The dielectric constant is a material property referenced to vacuum (which is 1) that increases capacitance and slows electromagnetic fields in a PCB. Because signal velocity scales as v = c / sqrt(er), Dk directly affects propagation delay and transmission line impedance. Choosing the right laminate Dk is therefore essential for accurate impedance control, timing, and overall signal integrity.

Is the dielectric constant really “constant,” and which frequency should I use for design?

No, Dk varies with frequency for all PCB materials and typically decreases as frequency increases, flattening around ~2 GHz for many common laminates. Using a 1 MHz Dk to design a 2 GHz system introduces error that propagates through the entire design. For modern fast edges (≈2 GHz and up), you should use Dk values characterized at 2 GHz or higher.

How does glass-to-resin ratio affect Dk, and why do thicknesses change the reported value?

Dk in a laminate rises with more glass and falls with more resin. Since different laminate thicknesses often imply different glass-to-resin ratios, the reported Dk changes with thickness. For example, low-cost FR-4 with ~42% resin can show higher Dk (e.g., the often-quoted ≈4.7 derived from ~1 MHz data), while higher resin content (e.g., ~55%) drives Dk lower, illustrating why a single “standard” Dk doesn’t represent real materials across frequencies and constructions.

How can I measure or extract Dk for my stackup?

You can infer Dk from signal velocity. A TDR can measure round-trip time on a known-length line with a high-impedance termination, yielding group velocity (and thus Dk) for a wideband pulse. For frequency-specific Dk, use a VNA: derive propagation delay from the phase of S21 as a function of frequency, then combine with the port spacing to get velocity and Dk across the band. Remember that results depend on copper roughness and the line structure: microstrip/CPW report an effective Dk, while stripline better reflects the bulk Dk (with roughness considerations). Material vendors typically use IPC-prescribed methods (e.g., ring resonator, clamped stripline) rather than TDR.

Why do datasheets (or different sources) give different Dk numbers for the same material, and what should I ask my fabricator or supplier?

Reported Dk depends on frequency, glass-to-resin ratio (and thus thickness), copper roughness, and the measurement method/structure. Some datasheets provide only a few frequency points and may omit the test method or roughness correction. Ask for: the exact frequency (preferably ≥2 GHz), the resin content (or specific glass style/laminate thickness), the measurement method, and whether the value reflects the intended structure (effective vs bulk Dk). Be wary of 1 MHz “parallel-plate” values for high-speed design; if a supplier cannot provide high-frequency, construction-specific data, treat their numbers with caution.