Panasonic Meeting Market Needs with Higher-Performance Megtron 7
I-Connect Technical Editor Pete Starkey sat down with Panasonic’s Tony Senese and Tomoyuki Abe at IPC APEX EXPO 2015 in San Diego–high-speed digital materials, particularly those with very low-loss characteristics, was the focus of their discussion. Also noted were the positive responses from chemical process suppliers, and the realistic length of a product development cycle.
Pete Starkey: Panasonic Megtron 6 has become established as the industry standard, but I understand there is a new product coming out of development and in the early stages of production, Megtron 7, which is the lowest-loss material that is currently available from Panasonic. Can you gentlemen give me some details of the characteristics and performance of the material, and the sort of applications of the industry sectors where this material is going to be of substantial benefit?
Tony Senese: We have a technology development roadmap that we update on a regular basis at least a couple times a year, and we have several market segments that we track our products in.
Tomoyuki Abe: Right now, Megtron 6 is the standard of the high-end server router equipment material, but the market is requesting the more high-speed material. Megtron 6 is quite good, but the market needs higher-performance material.
Starkey: To clarify, when you talk about high-speed material, what sort of data rates are you aiming at?
Senese: In the high-speed digital area right now, products are operating in the 10 Gbps to 20 Gbps range enabled by Megtron 6. However, at 20 Gbps other things have to be done to the circuit boards to make them meet the loss budget. Megtron 6 enabled the major IT infrastructure suppliers to use very similar design techniques to what they were using at lower speeds—say, 6–10 Gbs. So now, that is kind of the standard rate in that high-speed digital arena.
Starkey: So Megtron 7 effectively is an enabler that opens up the possibilities for them to develop into higher data rates?
Senese: That’s right. Basically, at the 25 Gbps data range, which is actually already in place, Megtron 7 removes a few roadblocks that make the fabrication difficult. At the 25 Gbps data range, especially on the large format back planes, and some of the line cards, the loss budget is too close to the mark, even with Megtron 6. It is a goal of all of these infrastructure companies to increase the amount of data they can transmit because of things like streaming video that have started to become ubiquitous, so they need to have their core routers to be much faster.
So 25 Gbps is where they're designing right now, and there are people who are designing certain pieces of equipment at the 50 Gbps range. So that meansthat really Megtron 6 isn't enough to build those core routers, as is.
There are connectors being developed that are fast enough and can be used that are early in production, and the people that are building what I would call proof-of-concept hardware are really just too close to the edge. They can build it once, but they can't build it in high volume, or even in normal quantities. But when it comes to loss, there is quite the difference from Megtron 6.
Abe: With Megtron 7 the loss is maybe 20-30% improved from the Megtron 6.
Starkey: Can I just ask, going away from the performance characteristics and the enabling characteristics, but speaking as a fabricator, does it present any problems in fabrication, or any particular special precautions needed?
Senese: Let’s address this by going through our targets for Megtron 7, which include the product characteristics, but also the kinds of things you're talking about. We know from experience that even products that have very good characteristics are unsuccessful if they don't fit into a certain window in terms of how the fabricators can use them.
Abe: Okay, our first target is of course the DF/DK properties. It's most important for us. Our goal was half of the loss of Megtron 6. That's our most important target. Then Megtron 6 is, electrically, very high end, but it also has a very good thermal and reliability performance, so Megtron 7 should be the same. So that is our second target.
Starkey: If I patiently fabricate and I have established a process procedure for building boards with Megtron 6, can I just use that established procedure for building boards on Megtron 7?
Abe: Yes, it’s very close.
Senese:One of the things that Megtron 6 has that is unique compared to previous materials that have been used in high-speed, beyond just better electrical performance, is the rheology of the system that allows material to be laminated in a way where the dielectrics are very consistent across any panel size. That means if somebody has to back drill for getting rid of stubs, they can do it very consistently with Megtron 6. A lot of the designs for the back planes and daughter cards that are being used now exploited that property because the thickness of the edge, the center, and from board-to-board was so consistent that they could easily set up to maintain a very tight back-drilling to get the stubs on any layer within 2–3 mils of the next interconnect, which is very tight.
That was something that, when Megtron 6 was designed, nobody planned on, but it has become a de facto standard. Luckily for Panasonic, other materials that look good on paper didn't always succeed in the market because that was something that had been exploited by the fabricators. So this is something that was added to our list of the things we have to do ,because we're designing this material to replace Megtron 6, and if they can't do this with this material, our own material will fail.
Starkey: Exactly.
Senese: So that's one thing that's similar with Megtron 7, the lamination cycle, the lamination thickness consistency, and the ability to exploit that, to make those designs is still there. One of the things that most of these high-end materials have in common, especially the thermally robust ones, is that drill wear is always a question. Some materials in the past have actually just failed because the drill wear went from a product like Megtron 6 where you could do a thousand hits on almost any size, down to sub-500 hit level. The drills for these materials, as you can imagine when you have a board that's almost half an inch thick, are very expensive. Another thing that happened is that people said, "Well, does it drill okay? Is it as good at least as Megtron 6?" Actually, in the wear studies that we've done with Megtron 7, it is a little better than Megtron 6.
Starkey: What sort of foils are you using or are recommended for use with the materials? What sort of bonding treatments for the foils, with signal integrity in the lines?
Abe: Yes, actually, as Tony said, the Megtron series has a unique resin system, so we need to find a custom adhesion treatment on the foil. We cannot use every foil for the Megtron series, but we already have some technology developed around Megtron 6 that is a similar resin, so we can use that technique for the Megtron 7. Now we have a very low profile
Senese: The use of the lowest profile foil available at the time we developed Megtron 6 became the de facto standard and led the way for all these products in the market today, and it has spawned a family of new products by our copper foil suppliers around the world who try to exploit that.
Starkey: I think the chemical process suppliers responded as well by producing bonding treatments that give a very low profile surface.
Senese: Right, this is the next part. The newest part or the part of developing Megtron 7 that you've asked about is this idea that we were able to quantify over the last seven or eight years that the bonding treatment has a significant effect. Maybe half the effect of the actual roughness of the copper foil, but a measurable, significant effect, especially as we go to the higher and higher frequencies. And so for the first time ever, part of Abe's development requirements became to go out at the time that the product comes to market and evaluate any and all of the most promising inner layer adhesion promoters that are available, so that we can make a very strong recommendation about which of those products we believe will give the customers the lowest insertion loss and the lowest transmission loss.
Starkey: So it's certainly an area that the chemical process suppliers are very conscious of at the moment.
Senese: Yes, now we have the chemical suppliers coming to us, to make sure the products they develop work with the Megtron series.
Starkey: They know what need to achieve and they're going to do as much as they can to help you. Guys, I think we've covered an awful lot of ground. Is there anything else in particular that you want to mention at this stage?
Senese: I think one of the things that is misunderstood most about this segment of the industry is how long it takes for a product to be developed and then submitted to the OEMs for their evaluations and then come to market. If nothing else, Megtron 6 has taught us patience, and so our perspective on how these development cycles go, I think, is a lot different than some of the rest of the people in the industry. We know that for a product to get into the market, to be qualified for these high-level designs, and then to actually become commercially successful takes years—many years in some cases.
Our goals for this product are different than they are for many of our other commercial products. We look at this product to be the leading product in the business to be evaluated by consortia, by OEMs, and by our customers over the next few years, so that maybe we start receiving production-level orders in 2016, perhaps the end of 2016, and that we actually don't even come close to the level of business that we have on our other products for several years.
Starkey: I completely understand. I'm sure as well that you're already working on Megtron 8.
Senese: I could tell you, but I'd have to shoot you (laughs).
Starkey: Gentlemen, thanks very much indeed for your time.
