Issue link: https://iconnect007.uberflip.com/i/972396
APRIL 2018 I FLEX007 MAGAZINE 33 From the material choice, I'd like to see what the actual requirements are. A lot of times, and from the OEM side, they'll just spec in the best they can possibly get. If we look at polyimide for instance, as an example, if you go look at cast on copper products like single- sided copper-clad, you'll see these extremely awesome mechanical numbers. And then those can get transferred sometimes from the OEM level down to other flex circuits. It might be the dou - ble-sided copper clad with the single- sided copper-clad requirements down there and not knowing that mak- ing a double-sided clad is a lot different than making a single sided clad. The mate- rials are vastly different, too. These are the sort of things that provide us feedback, outside of when people call to complain that something is underperforming. So how do you bridge the gap? Maybe this could be a way to do that. Then you could say, "What are the actual require- ments for performance in the system?" Not "This is what we're using and it works, so those are the requirements." What does an antenna feed line need to be? What do the dielectric properties need to be? What does the flux performance need to be? What do digital interconnections need to be? Where can you get away with ED versus RA? Where can you get away with a lower per- forming polyimide or different system? Shaughnessy: Mark, what do you think? Finstad: Well, I'd like to address what Jona- than said. One of the biggest issues that a lot of people don't understand is that all of the flex dielectrics that have low dielectric constant and low loss require really high-temperature lamination. With the LCPs, DuPont DK, pretty much all of them, you have to laminate around 300°C. Not a lot of flex suppliers have presses that go up that high, so you really limit your vendor base. And I know that the first time that I looked at LCPs I was talking to a guy from Rogers and he specifically said, "Don't try to do this without having us involved because it's a real bear to work with. It's a thermoplas - tic so things don't stay put where they're sup- posed to and so there's a lot of challenges." And I don't know if the price has come down recently, but it used to be outrageously expen- sive and if you had multiple layers of the stuff you could have anywhere from $50- $100 or more per square foot in material costs before you start doing anything with the material. It really made it tough to make commercial appli- cations with that material because price-wise it was off the charts. That's a big prob- lem with the LCPs. Now I know that there's some new lower temp materials com- ing out that have a dielec- tric constant in the 2.8 range. I think when that comes out it would be good to stay in touch with the manufacturers to have the tri- als that verify that the end products are elec- trically and mechanically good but also that the materials are easier to process. Because having something that does everything you need electrically and mechanically on the final circuit doesn't do a whole lot of good if nobody can work with it to get to that final product. Weldon: I want to echo Mark on that point. Because you look at the LCPs for instance, right? You're really paying for that loss tangent number which is .001-.002 and it's a bear, or you can just sacrifice and go with an epoxy which can process easier, it's cheaper and much better to work with. And all you do is go from .001" and .002" up to .002" to .003". I'm going to go through all this trouble for that extra 0.001 loss tangent number? I think that's kind of the key, right? Are people looking at that? Mark Finstad