PCB News - Sometimes one sentence can improve the channel loss of high-speed signals?

Let me give you a scenario to think about: When the PCB design is completed, the laminate is confirmed, and the board has been put into the state, the simulation evaluation found that the high-speed signal channel margin may not be safe. Is there still a chance to improve?

Of course, the probability of this problem will not be very high, because since the simulation evaluation is to be carried out, the simulation verification work will definitely be put in the front to determine in advance whether the loss of the channel is ok, and to determine whether to use a better board or walk. Wider traces to reduce losses. If you really meet an SI engineer who starts the simulation to evaluate the loss after the design, Mr. Gao may feel sad for you!

But if you are really unlucky and meet a simulation engineer who tells you that the loss margin of the high-speed signal may be insufficient when you finish the design and prepare for the board, what else can you do? At this time, the laminate and board have been determined long ago. The device layout and wiring have been completed. Maybe the probability that you need to tear down and redesign has exceeded 95%. At this time, if you read this article, you may just Grasping the remaining 5% of the possibility, a simple sentence can improve the loss that is not enough!

What kind of method does this article want to talk about? Let's sell it first. First of all, let's take a look at which parts determine the loss of our PCB traces!

Let's first introduce the classification of PCB losses from the general direction. There are three types of loss: conductor loss, dielectric loss and radiation loss. You may hear that there are more of the first two, but the actual loss of the PCB is mainly the first two. As for the radiation loss, it is also related to the dielectric constant DK. In addition, the radiation loss basically only exists in the microstrip line, and with proper design, it can be reduced to a relatively low level, which is the proportion of the total loss. It's very small, so I won't introduce it here.

Among them, the dielectric loss is mainly caused by the polarization phenomenon of the dipole. In order to start from the habit of listening to too many theories, we will make the theory short, just like the figure below. The higher the frequency of the applied voltage, the larger the current., The greater the number of dipoles that swing in the material, the greater the amount of dipole movement under the action of the electric field, and the greater the volume resistivity, the higher the power loss in the medium. In order to describe the material properties that measure the law of dipole motion, the concept of DF came into being.

Let me talk about another part, the principle of conductor loss. First of all, we must know that there is an important concept in the high-speed theory called the skin effect. At higher frequencies, the current will run along the surface of the conductor, that is, at high frequencies, the size of our resistance depends on the current flowing through. The size of the cross-section, the smaller the cross-section through which the current flows, the greater the resistance, so the conductor loss also gradually increases with the increase in frequency.

Sorry, Mr. Gao has tried his best to compress the theoretical knowledge, and some fans may think that it is unnecessary, but this is still very helpful for everyone to analyze the loss of PCB traces!

To sum up the loss factors mentioned above, the sheet material mainly determines the dielectric loss, that is, the size of the sheet DF that we often say has the greatest impact on the loss, and it is for this reason that we distinguish between different grades of sheets . In addition, the line width and copper thickness of the trace affect the conductor loss. The above mentioned is very consistent with the scenario we summarized. The board is fixed, the dielectric loss is basically fixed, the stacking and design are fixed, the wiring structure is fixed, and the conductor loss is basically fixed. So if we still want to improve the loss in this case, we must see if there are any factors that can affect our loss.

In fact, fans who read our Mr. Express article or our new book will know that in addition to the above factors, we also introduced the influence of copper foil roughness. The surface of the copper foil is relatively rough (in order to increase the adhesion between the copper foil and PP), so the roughness of the copper foil needs to be considered at high speed, and the roughness of the copper foil will also affect the loss of the trace.

In fact, this loss can also be counted in the conductor loss. The principle is roughly like this. Due to the skin effect, current will be transmitted on the copper teeth. When passing through the protruding copper teeth, compared to the smooth copper surface, the current transmission The path becomes longer, so it will further increase the DC and AC resistance at the same time, thereby increasing the conductor loss.

We are familiar with several types of copper foils with different roughness grades, including ordinary STD copper foil, RTF reversal copper foil and HVLP ultra-low profile copper foil. Of course, there are now HVLP2 and HVLP3 copper that continue to be optimized on the basis of HVLP copper foil. Foil. But the reality is like this. Many friends know that copper foil with different roughness will affect the loss, but they don't know how much influence it has. Can there be quantitative data to give it. Everyone actually thinks that it is not easy. The not easy point is that the loss is composed of several large parts, mainly including the sheet DF, the trace width and copper thickness, the reference layer thickness and the copper foil we are talking about now. Roughness. If you want to extract the effects of different types of copper foil roughness separately, you must ensure that the factors are consistent before you can simply extract them. The specific point is to ensure that the board is the same, the wiring structure is the same, and the thickness of the wiring reference is also the same. Only when the roughness of the copper foil is different can the influence of the copper foil roughness be known separately. Do you think it is possible to do it?

Of course, since Mr. Expressway asked, it means that Mr. Expressway must have done it! Mr. Gao also specially made a test board for this purpose, in order to compare the difference between the popular RTF and HVLP copper foils. Yes, it is simply the difference brought about by the difference in copper foil!

The above is an introduction on how to improve the channel loss of high-speed signals. Ipcb is also provided to PCB manufacturers and PCB manufacturing technology.