LAMINATES: DK, DF AND TRACE WIDTH
What are the Dk and Df of AGC/Nelco ELL 102?
AGC/Nelco ELL 102 has a dielectric constant (Dk) of 3.03 and a loss factor (Df) of 0.0009 at 10 GHz, as published by its maker. A 50 ohm surface microstrip over a 0.2 mm dielectric comes out about 0.466 mm wide on it, against 0.353 mm on standard FR-4.
- The nearest grade in the catalogue by Dk is Isola Tachyon 100G, at Dk 3.02 and Df 0.0021 at 10 GHz.
- Dielectric loss on that microstrip is about 0.028 dB per inch at 10 GHz, and the signal travels at about 130 ps per inch.
- Buried as a stripline between planes 0.4 mm apart, 50 ohms needs a trace about 0.189 mm wide.
50 ohm surface microstrip, 1 oz copper
| Dielectric height | Trace width on ELL 102 | On standard FR-4 |
|---|---|---|
| 0.1 mm | 0.222 mm | 0.166 mm |
| 0.2 mm | 0.466 mm | 0.353 mm |
| 0.5 mm | 1.21 mm | 0.920 mm |
| 1.6 mm | 3.95 mm | 3.02 mm |
As published
| Property | Value |
|---|---|
| Dk @ 10GHz | 3.03 |
| Df @ 10GHz | 0.0009 |
| Tg (°C) | 190 |
| Halogen-Free | No |
| Target Data Rate | 224G+ |
Nearest grades in the catalogue by Dk at 10 GHz
| Laminate | Dk | Df | Tg |
|---|---|---|---|
| Isola Tachyon 100G | 3.02 | 0.0021 | 215 (DSC) |
| Rogers RO3003 | 3 | 0.001 | not published |
| Shengyi mmWave77 | 3 | 0.001 | not published |
How this was counted
Dk and Df are the maker's published values at 10 GHz, from the MPBxChange material catalogue, which keeps each source's test frequency and method. Where a maker publishes a range, the midpoint is used. Widths are solved with the Hammerstad and Jensen microstrip equations and Wadell's stripline form, for 1 oz copper, and are a starting point: Dk moves with frequency, resin content and glass style, and a fabricator's field solver is the authority. Standard FR-4 here is the median of 22 FR-4 grades in the catalogue (Dk 4.4).