Do You Need Low Dk PCB Materials? 20221202

 

您需要低介電常數 PCB 材料嗎？20221202

 
Hey everybody, welcome to Altium Academy. I'm your host Zach Peterson and of course, I am your local technical consultant for Altium and today we're going to be talking about an important class of materials that often appears in high speed and high frequency boards and that is PTFE material. These materials are probably best known by some of their brand names like Rogers and Arlon and they are very useful in high frequency boards, but they are sometimes recommended for use at high speed boards where they might not always be necessary. So, you'll always need to use a PTFE laminate in a high speed piece of meat. That's why we're going to break down this video. Let's go ahead and get started. So, one of the common recommendations in high speed PCB design is to use a low decay dielectric. I think people misunderstand why these should be used and particularly why they aren't always needed in a high speed PCB. So, one of the most common low decay materials that's out there that is recommended for use in high speed PCBs is poly tetra flora ethylene or PTFE better known by the name Teflon.
大家好，歡迎來到 Altium Academy。我是主持人 Zach Peterson，同時也是 Altium 的在地技術顧問。今天我們要討論的是高速與高頻電路板中經常使用的重要材料——PTFE 材料。這類材料最廣為人知的是像 Rogers 和 Arlon 這樣的品牌名稱，它們在高頻電路板中非常有用，但有時也會被推薦用於高速電路板，儘管在這些應用中未必總是必要。因此，在高速電路板中，你總是會需要使用 PTFE 層壓板。這就是我們製作這部影片的原因。讓我們開始吧。 在高速 PCB 設計中，常見的建議之一是使用低衰減介電材料。我認為人們對為何需要使用這些材料，尤其是為何在高速 PCB 中並非總是必要存在誤解。目前最常被推薦用於高速 PCB 的低衰減材料之一是聚四氟乙烯（poly tetra flora ethylene），簡稱 PTFE，更廣為人知的名稱是鐵氟龍（Teflon）。


So, this is one class of materials that is often used in radio frequency PCBs operating at very high frequencies, and sometimes its recommendation gets put forth for high speed PCBs. So, there are two typical reasons that someone will recommend the use of a PTFE or similar laminate with a low decay value in a high speed PCB, and I think they're both not totally correct. So, one of the more common high speed PCB design recommendations that I sometimes see out there on the internet is the recommendation to use a low decay dielectric as the material that supports higher speed interconnects. So, I think it's a valid question to ask, do you actually need to use one of these materials and why would you use one of these materials in the first place? So, let's just take a look at an example of some of these recommendations. So, here I'm on kinics and I'm just going to search for a common material PTFE or poly tetra flora ethylene.
這是一類常用於極高頻率射頻電路板的材料，有時也會被推薦用於高速電路板。關於在高速電路板中使用低衰減值的 PTFE 或類似基板材料，通常有兩種常見的推薦理由，但我認為這兩種說法都不完全正確。其中一個我在網路上較常見到的高速電路板設計建議，就是推薦使用低衰減介電材料來支持更高速的互連傳輸。我認為這確實值得探討：究竟是否真的需要使用這類材料？又為何會優先考慮採用這類材料？現在讓我們來看看一些具體的推薦實例。這裡我打開了 Kinics 網站，準備搜尋常見材料 PTFE（聚四氟乙烯）。

Here in this blog, they recommend for high speed and high frequency applications enhanced epoxy or PTFE materials may be a better choice. I think this is a pretty arbitrary statement. There's no explanation as to why here. Also, they say the dielectric constants of FR4 may be insufficient to hold the impedance values required, which is just not true at all. There are plenty of high speed boards that are built on FR4 laminates and they function just fine. Let's go over here to EPEC. Now, these guys are very knowledgeable, and I like a lot of their material that they have on their website. It's all really valuable. But even they recommend their numerous materials for your high speed signals such as PTFE based laminates, spread glass laminates or other specialized material systems. Here, they're not explicitly telling you all the time to use PTFE, but they are recommending PTFE as one of the laminate materials that you can use for high speed signals.
在這篇部落格中，他們建議對於高速和高頻應用，增強型環氧樹脂或 PTFE 材料可能是更好的選擇。我認為這是相當武斷的說法，完全沒有解釋原因。此外，他們聲稱 FR4 的介電常數可能不足以維持所需的阻抗值，這根本不符合事實。有許多高速電路板都是使用 FR4 基板製成，而且運作完全正常。讓我們來看看 EPEC 的資料。這些專家非常專業，我很喜歡他們網站上的許多資料，都非常有價值。但就連他們也推薦多種適用於高速信號的材料，例如 PTFE 基板、分散玻璃纖維基板或其他特殊材料系統。雖然他們沒有明確要求必須使用 PTFE，但確實將 PTFE 列為可用於高速信號的基板材料選項之一。

So, this is a little bit more appropriate because they're just recommending it as an option, but they're not telling you that it's going to be more appropriate for high speed than for other applications. Versus other materials. And then, let's take a look over here at Caden's Real Quick. So, they do the same kind of thing. In fact, they actually just kind of copied the kind of statement over here, writing for high speed and high frequency applications, enhanced epoxy or PTFE materials may be a better choice. You actually see that stated almost exactly right here in this bullet point. So, they apparently just copied it. So, PTFE is the one material that everyone seems to bring up when they start talking about low decay materials being used for high speed. Now, it is true. PTFE materials can be very useful for certain high speed boards, and there are some reasons why, as I'll show in just a moment. Now, they're normally used in RF boards. And one of the reasons that they are used in RF boards is the same reason that someone might recommend that you use them in a high speed board, which is their lower losses.
所以這樣說會更恰當些，因為他們只是把這當作一個選項來推薦，但並沒有告訴你它比其他材料更適合高速應用。接著，我們來看看凱登的快速指南。他們做了類似的事情。事實上，他們幾乎就是把這邊的說明直接複製貼上，寫著「針對高速與高頻應用，增強型環氧樹脂或 PTFE 材料可能是更好的選擇」。你可以在這個要點處看到幾乎完全相同的敘述。顯然他們就是直接照搬了這段話。 每當人們開始討論將低損耗材料用於高速應用時，PTFE 總是那個被反覆提及的材料。確實如此，PTFE 材料對某些高速電路板非常有用，稍後我會說明具體原因。這類材料通常用於射頻電路板，而它們被應用在射頻板的原因之一——也正是有人會推薦你在高速板中使用的原因——在於其較低的損耗特性。

However, they also have a lower decay material, and I'll bring up why that might be important here in just a moment when we do a demo in all-time design. First, let's take a look at a common PTFE based material, which would be a Rogers Laminate. Rogers Laminates are one class of different Laminates that are compatible with standard epoxy glass Blaminates. So, standard epoxy glass Laminates would be essentially a resin system reinforced with fiberglass, and then that is bonded to another layer of Lamin ate material to build up your stack-up. These Rogers materials are compatible in that you can use them to build a hybrid stack-up. So, Rogers materials are often brought up because here, if you look on this webpage, you'll see that they do have a low dielectric constant. And here , this particular Rogers material RO4835 has a decay value rated of 3.48. That's probably at about 10 gigahertz. You'll also notice here that they have a dissipation factor cited of 0.0037.
然而，它們也具有較低的衰減材料特性，稍後在我們進行全時序設計示範時，我會說明這點為何重要。首先，讓我們來看看常見的 PTFE 基材——羅傑斯層壓板。羅傑斯層壓板是眾多層壓板類型中的一種，能與標準環氧樹脂玻璃層壓板相容。標準環氧樹脂玻璃層壓板本質上是以玻璃纖維強化的樹脂系統，再透過黏合其他層壓材料來構建堆疊結構。這些羅傑斯材料的相容性在於，您可以用它們來建構混合堆疊結構。之所以經常提及羅傑斯材料，是因為在這個網頁上可以看到它們確實具有低介電常數。此處展示的特定羅傑斯材料 RO4835，其衰減值標示為 3.48（推測測試頻率約為 10GHz）。同時您會注意到，這裡列出的損耗因數為 0.0037。

So, if you compare that to the lost tangent of FR4, which is about 0.02, that's about a factor 5 reduction. Another Rogers material that is often brought up as kind of the classic PTFE material is Rogers 303. So, Rogers 303, again, is a useful Laminate, and it does also have a very low decay value. In fact, decay value right here is cited on the webpage is 3. And then the dissipation factor is also very low 0.001. So, that's about a factor 20 reduction compared to FR4, with remember its lost tangent of about 0.02. These are some of the classes of materials that are often brought up when someone is talking about a PTFE material, as you see in some of these other web pages. So, another group of materials that offers slightly lower decay value than a typical FR4 Laminate, and can also offer better losses, is a group of materials manufactured by iTech. So, iTEQ, this company offers a range of different Laminates. There are some companies that I work with that actually use these Laminates in their standard stackups. And they give you that option to find a somewhat lower DF material with a moderate to low decay value, as well as higher than normal glass transition temperature values. So, here you see one of their options. It has a DKF 3.9 DF of 0.008. Here's a couple more. Here you can see this one is actually a high TG Laminate. So, you can see the TG value gets up to 185 degrees Celsius, and then it also has much lower DF value, as well as DK value that is somewhat lower than a standard FR4 Laminate. And you can go through and find all of their data sheets online, and you'll be able to find several different options where you can have the freedom to choose different Laminates and different DK and DF values. And then there's also Isola. Again, we're not endorsing any of these companies, and I know that I bring up Isola a lot, but Isola does also have a set of Laminates that have somewhat lower than normal DK values, but then kind of typical DF values.
因此，若將其與 FR4 的損耗因數（約 0.02）相比，這大約降低了 5 倍。另一種常被提及的經典 PTFE 材料是羅傑斯 303。羅傑斯 303 同樣是實用的基板材料，其介電常數值也非常低——根據網頁資料顯示，此處標註的介電常數值為 3，而損耗因數同樣極低，僅 0.001。相較於 FR4 的損耗因數 0.02，這相當於降低了約 20 倍。這些都是討論 PTFE 材料時經常被提及的類別，正如您在其他網頁所見。 另一類由 iTech 公司生產的材料，能提供比典型 FR4 基板稍低的介電常數值，同時具備更優異的損耗表現。iTEQ 這家公司提供多種不同基板材料，我合作的某些企業實際會將這些材料納入標準疊構中。這些材料讓您有機會選擇介電常數中低、損耗因數較低，且玻璃轉移溫度高於常規值的方案。 所以，你現在看到的是他們其中一個選項，其 DK 值為 3.9，DF 值為 0.008。這裡還有幾個例子。你可以看到這個其實是一種高 TG 層壓板，TG 值最高可達攝氏 185 度，同時它的 DF 值也低得多，DK 值也比標準 FR4 層壓板稍低一些。你可以上網查閱他們所有的數據表，會發現有幾種不同的選擇，讓你能自由挑選各種層壓板及不同的 DK 和 DF 值。另外還有 Isola 這家公司。再次強調，我們並非為任何公司背書，雖然我經常提到 Isola，但 Isola 確實也提供一系列 DK 值略低於常規、但 DF 值相對典型的層壓板。

So, there's a lot of different options out there, and you don't just have to default to PTFE when you're looking for a somewhat lower DK or DF value, then the standard grade FR4 Laminate that's you're going to find out there on the market. So, why do people continue to recommend PTFE as the standard material set that you should use for high speed PCBs? Well, we just saw one of those reasons as we were looking through some of these material sets, and one of those reasons is lower loss. However, let's not forget, you can still get lower losses without using a PTFE material. One reason that isn't ever brought up, but that is a true reason, is the fact that you can get unreenforced PTFE. So, some PTFE Laminates come with a fiberglass reinforcement, which makes them more rigid and easy to work with and manufacturing. Other PTFE Laminates are unreenforced, and there are sometimes called wet noodles, meaning that they flop around a little bit when you're actually trying to put them into a PCB stack up, and that can create misregisteration during the drilling and alignment process. So, that's actually problematic if you need to produce at high volume, and I've actually had fabricators tell me that they recommend that you go with a reinforced versus an unreenforced PTFE Laminate. Another reason that isn't often brought up is actually the wet target that you can hit for impedance controlled line for various Laminate thicknesses , and then there's one other reason which is sometimes brought up, which is the propagation delay. So, the propagation delay will be different when the DK is lower, and that is sometimes brought up in the context of impedance control, however, if you're the smart high-speed PCB designer, you won 't worry about that particular point. So, let's take a look at all team designer and we'll get into the impedance tool, and we'll be able to see why some of these different recommendations are important. So, here I'm in the side all team designer, I'm just going to create a new PCB in this project, and we'll actually just pull it out of the project, and then we'll go over here to the layer stack manager so we can access the impedance calculator. Okay, so now that the layer stack is up, let me open up the impedance calculator, and then we'll add an impedance profile, so we can see some of these important results with these calculations in the term, and why it might be important to use a lower DK value. So, here, just as an example, we're just going to consider a 10-mil thick dielectric, and we're going to consider a DK of 4 just to get started. And so you'll notice here, then we have a target impedance of 50 ohms, and we have a 10-mil thick dielectric, we can then get to a width, or for mic rostrips, of 18.9 mills. So, that's a little wide, if you were to go to a thinner laminate here, let's say 5 mills, and we still kept that DK of 4. We see that the width goes down now to about 8.98 mills, so about 9 mills, so it's still satisfying that approximately 2x relation, that's pretty common with DK 4 materials. So, now as we start to get much thinner, and let's say we went down to, I don't know, 2 mill thickness. So, now with a 2-mil thickness, you can see that the width of this microstripline is actually 3 mills wide, so is 2-mil 2 thin? Well, that's going to depend on your application, and 2-mil thick laminates are possible, because just look through this list by isola, you're going to see here that you can actually get down to 2-mil thick dielectrics with an open weave that have somewhat low dielectric constant. Now, the lower dielectric constant becomes beneficial in this case, because let's say we go down to a DK of 3. Now, you can see that my width for my transmission line has increased by about 30 percent, so it's gone from 3 mills to 4 mills. So, this is actually going to be much easier to manufacture at 4 mills versus 3 mills. Now, if you could get even lower than 3, you could actually get this a little wider, and then you won't have to use as precise of a process in order to fabricate that particular board. Now, let's just go back here to DK of 4 for just a moment, and then if we scroll down here, you can see right here one of the results from this calculation is the propagation delay. So, when we have a DK of 4, it's about 150 picoseconds per inch. If we take this we change it to a DK of 3, you can see that the propagation delay in time per length goes down. That means the signal is moving faster. So, in the DK value is lower, the signal is moving faster, and that means the electrical length of that interconnect is going to be somewhat longer. And that means you could route over a longer distance without impedance control before you have an issue with reflections at the receiver end of that interconnect. But if you're a smart high-speed PCB designer, you're just going to design to the target impedance regardless, and that's what you should be doing, especially if you're using Altium because it has this convenient tool built into it to help you out. So, this is one of the reasons why you would actually want to use a lower DK dielectric in a high-speed PCB. It's because when you have to hit an impedance target, you can actually use somewhat wider line widths, as those laminate thicknesses go down, and you try to pack more layers into the board. It's going to require a less precise manufacturing process. It's going to make it easier for you to produce the board anywhere, and of course it's going to lower your cost.
市場上確實有許多不同的選擇，當你需要比標準等級 FR4 層壓板更低的 DK 或 DF 值時，不一定非得直接選用 PTFE。那麼，為什麼大家仍持續推薦 PTFE 作為高速 PCB 的標準材料呢？我們剛才在檢視這些材料組時已經看到其中一個原因——較低的損耗。但別忘了，即使不使用 PTFE 材料，你仍然可以實現更低的損耗。有個鮮少被提及卻真實存在的原因，就是你可以取得「未增強型 PTFE」。部分 PTFE 層壓板會加入玻璃纖維增強材，使其更堅硬且易於加工製造；而其他未增強型的 PTFE 層壓板有時被戲稱為「濕麵條」，因為在實際組裝 PCB 疊構時，這些材料會有些軟塌晃動，可能導致鑽孔與對位過程出現偏移誤差。 因此，若需要進行大規模生產，這確實會產生問題。實際上，我曾聽製造商建議，應選擇強化型而非非強化型的 PTFE 層壓板。另一個較少被提及的原因，是針對不同層壓板厚度進行阻抗控制線路時可達到的濕蝕刻目標值。此外還有偶爾被討論的傳播延遲問題——當 DK 值較低時，傳播延遲會產生變化，這個議題有時會在阻抗控制的範疇中被提出。但如果您是聰明的高速 PCB 設計師，其實無需過度擔憂這點。 現在讓我們開啟 Altium Designer，進入阻抗工具來解析這些不同建議的重要性。目前我已在 Altium Designer 介面中，準備在此專案創建新的 PCB 檔案。我們先將它從專案中獨立取出，接著切換到層堆疊管理員，即可開啟阻抗計算器進行後續操作。 好的，現在層疊結構已經設定好了，讓我開啟阻抗計算器，接著我們來加入阻抗分析，這樣就能看到這些計算結果中的關鍵數據，以及為什麼使用較低的 DK 值可能很重要。這裡我們先以 10 密耳厚的介電層為例，並假設 DK 值為 4 作為起始。你會注意到，當目標阻抗設為 50 歐姆，介電層厚度為 10 密耳時，微帶線的寬度計算結果是 18.9 密耳。這個寬度稍微有點大，如果改用更薄的層壓板，比如 5 密耳，同時保持 DK 值為 4，你會看到寬度降至約 8.98 密耳，也就是大約 9 密耳，這仍然符合 DK 值為 4 材料常見的約 2 倍關係。那麼，當我們進一步縮減厚度，假設降到 2 密耳，這時你會發現微帶線的寬度實際上只有 3 密耳——所以 2 密耳會不會太薄呢？ 這取決於您的應用需求，而 2 密耳厚的層壓板確實是可行的。只要瀏覽 Isola 的這份列表，您會發現實際上可以獲得介電常數略低、採用開放式編織結構的 2 密耳厚介電材料。在這種情況下，較低的介電常數會帶來優勢，假設我們將 DK 值降至 3，您會發現傳輸線的寬度增加了約 30%，從 3 密耳變成 4 密耳。這意味著生產 4 密耳寬度實際上會比 3 密耳容易得多。如果能把 DK 值降到比 3 更低，線路寬度還能再增加，這樣就不需要採用過於精密的製程來生產特定電路板。 現在讓我們暫時回到 DK 值為 4 的情況，當我們向下滾動頁面時，可以看到此計算結果中的傳播延遲數據。當 DK 值為 4 時，每英寸延遲約為 150 皮秒；若將 DK 值改為 3，會發現單位長度的傳播延遲時間隨之降低，這代表信號傳輸速度變得更快。 因此，當介電常數（DK）值較低時，訊號傳輸速度會更快，這意味著互連線路的電氣長度會稍微增加。也就是說，在阻抗控制之前，您可以在更長的距離上佈線，而不必擔心互連線路接收端出現反射問題。但如果您是一位聰明的高速 PCB 設計師，無論如何都會按照目標阻抗進行設計，這正是您應該做的，特別是當您使用 Altium 時，因為它內建了方便的工具來協助您。這就是為什麼在高速 PCB 中，您實際上會想要使用較低 DK 介電材料的原因之一。因為當您需要達到阻抗目標時，隨著層壓板厚度減薄，並且試圖在電路板上塞入更多層時，您實際上可以使用稍寬的線寬。這將對製造工藝的精確度要求較低。 這將使您在任何地方生產電路板都更加容易，當然，這也會降低您的成本。

Now, as we saw, if we made this laminate thickness larger, so let's say we go back to 10 mills for a moment, that's actually going to require still a wider line width in order to hit that 50 ohm impedance target. Now, this is going to require that you bring two traces in a differential pair closer together in order to bring that width down, so that it's actually manageable, and you can route into fine pitch components, or it's going to require you to use a larger DK value or both. So, if you're on thicker laminates, you may not actually want to use a lower DK material unless your interconnects are going to be very long. So, when your boards are very large, and those routes are going to be very long, and you may have to go through, say, a couple of connectors, then it actually makes sense to use a lower loss material and possibly even a PTFE material. That issue with the insertion loss along the length of that interconnect, and getting that down as low as possible, to ensure that you can actually recover the signal at the receiving end of that interconnect. On shorter interconnects, this is not going to be so important, you could actually use a somewhat higher DF laminate, and you'll still be able to get that signal to the receiver, as long as everything is impedance matched. So, just to summarize, don't be afraid to use a DK value around 4, even if you're operating with a high-speed PCB design. So, there are plenty of DK 4 laminates or DK approximately 4 laminates that still offer somewhat lower DF values. You don't need to go all the way up to a PTFE laminate in order to hit your loss targets, and in order to make sure that your board operates correctly. So, I've designed plenty of high-speed boards that are on like an i-tech laminate, and have never gone up to PTFE simply because we didn't need to use PTFE. We didn't need that low of a loss, and we wanted to make sure that we could produce the board anywhere, and so, of course, we just went with one of the standard i-tech laminates. Alright, that's all I've got for you today folks. Thanks for watching, make sure to hit that subscribe button, leave your comments and questions in the comments section. Let us know your favorite laminates, and of course, hit that like button and share it on social media. And last but not least, don't forget to call your fabricator, folks.
現在，正如我們所見，如果我們將層壓板厚度加大，假設暫時回到 10 密耳，要達到 50 歐姆阻抗目標實際上仍需要更寬的線寬。這將要求您把差動對中的兩條走線靠得更近，以縮減寬度使其實際可操作，讓您能佈線到細間距元件，或者需要您使用更大的 DK 值，或兩者兼施。因此，若使用較厚的層壓板，除非您的互連線路非常長，否則可能實際上不會想選用較低 DK 的材料。所以當電路板尺寸很大，且走線路徑非常長，甚至可能需要穿過幾個連接器時，選用低損耗材料甚至 PTFE 材料就確實合理。這是為了降低沿互連線長度的插入損耗，並盡可能將其減至最低，以確保能在互連的接收端實際恢復信號。 在較短的互連線路上，這點並不那麼重要，您實際上可以使用介電損耗因子稍高的基板材料，只要所有元件都實現阻抗匹配，信號仍能順利傳送至接收端。總結來說，即使您正在進行高速 PCB 設計，也無需畏懼使用介電常數值約為 4 的材料。現今有許多介電常數為 4 或接近 4 的基板材料，仍能提供相對較低的介電損耗因子。要達成損耗目標並確保電路板正常運作，您不一定非得選用 PTFE 材質的基板。我設計過許多採用類似 i-tech 基板的高速電路板，從未升級到 PTFE 材質，因為我們根本不需要使用 PTFE——我們不需要那麼低的損耗值，同時希望確保電路板能在任何地方生產，因此自然選擇了標準的 i-tech 系列基板。好了，今天就和各位分享到這裡。感謝觀看，別忘了按下訂閱按鈕，並在留言區留下您的意見與問題。 請告訴我們你最喜歡的基板材質，當然也別忘了按讚並在社群媒體上分享。最後同樣重要的是，各位別忘了打電話給你的電路板製造商。