There was prior art, ZBC inventor says: 'distributed' capacitance covered by now-expired patent.

MINNEAPOLIS--In a stunning sequence of events, the inventor of the predominant material for buried capacitance revealed at a major technical conference his finding of several pieces of prior art that, in his words, "were not considered" when the U.S. Patent and Trademark Office was undergoing the process of validating his patent. In doing so, Howard raised several questions over the use of his invention, ZBC, the buried capacitance material now property of Sanmina-SCI, and the possibility of an open architecture for designing and manufacturing embedded technologies.

On Jan. 7, 1992, the USPTO assigned a patent for a buried Capa citance material to Zycon Corp., and the engineers who conceptualized it, James Howard and Greg Lucas. Patent no. 5,079,069 described a capacitor laminate as an intermediate layer for use in a PCB (or assembly) as means to provide a "bypass capacitive function for large numbers of devices motmted or formed" on the board. The patent further described the capacitor laminate as formed from sheets of conductive material and an intermediate sheet of dielectric material. The material was henceforth known as ZBC--Zycon Buried Capacitance.

In retrospect, what the USPTO missed, Howard told an audience of about 25 assembled at an IPC technical conference on Sept. 29, were citations to senior patents--other works that predate 5,079,069. "It's pretty obvious that the patent examiners didn't have any of these references to check when they made their decisions," he told PCD&M. "The fact is, these citations, on the front of every patent, don't list the other patents. They were not considered." Indeed, one of the senior patents, U.S. no. 4,494,172, was issued on Jan. 15, 198.5, seven years earlier than the Howard patent. Howard also noted that John Sider of Unisys patented "exactly the same thing and he did it before I did." Sisler's patent, for capacitive planes using epoxy material of about 0.015" thick, was issued April 30, 1991. Howard's research also uncovered at least two other works that cover similar ground.

The big fish, however, is 4,494,172. Issued on Jan. 1.5, 1985, to Burt Leafy and Shaun (Dennis) Silverio of what was then Mupac Corp., it teaches distributed capacitance through an epoxy layer. By using the Leary-describcd plane with a planar resistor technology, a complete embedded passive PCB may be fabricated, Howard says. "What I am saying is, if you are an OEM, [and] just using a distributed capacitance layer, and not shared or borrowed capacitance, you are practicing under Leary."

Zycon was purchased in 1996 by Hadco Corp., which in turn merged with Sanmina in 2000. The rights to the ZBC patent were transferred each time. George Dudnikov, senior VP and CTO of Sanmina-SCI's PCB division, said any claims that Sanmina-SCl's patents are invalid are "absolutely false. There are nine patents in Sanmina's portfolio. All are valid."

U.S. law protects patented works for 17 years from the dare of issue. The Leary patent expired Jan. 15, 2002. Accord ingly, anyone is free to use it. And the point isn't whether the ZBC patent now held by Sanmina-SCI is valid, Howard says. Rather, it's that the principle put forth in Leary is the "best and clearest use" of how to build a distributed capacitance layer. "And," says Howard, "if the principle you are using is Leary, use it. It's expired."

Sanmina-SCl asserts its patent for ZBC is valid. The "Leary patent is not a revelation," Dudnikov wrote in an email to PCD&M. "It has come up before and not had an influence .... Read the patent claims carefully and you will see [the Leary patent] addresses wire wrap panel cards with three layer structures and dielectric thickness from 0.005-0.009". People can use capacitive planes with 0.005-0.009" dielectrics. Our patents cover a different space."

Under the United States Code of Patent Laws and Regulations, Section 301, "any person at any time" may cite to the USPTO prior art--patents or printed publications--deemed to have a bearing on an issued patent. Per Section 302, a request for reexamination may be filed. The USPTO Director is then obligated to make a determination of the claim's legitimacy within three months. Per the USPTO's Presumption of Validity, "all issued U.S. patents are presumed to be valid." In asserting Sanmina-SCI's rights, Dudnikov cited this fact to PCD&M. According to Smith and Hopen, a Washington, DC-based intellectual property law firm, "The presumption of validity means that a patentee does not have to prove in court that the patent is valid. The accused infringer bears the burden of proving, by clear and convincing evidence, that the PTO erred in awarding the patent." U.S.C. sections 314-316 set forth the process for the USPTO to declare a patent invalid.

Howard told PCD&M he happened upon the purported prior art during the course of research undertaken during the past 12 months at the behest of an OEM. "I have heard from OEMs and consultants concerned about designing [embedded technologies] out because of the source of supply. A large OEM told me this was a grave concern to them and [asked if] 1 was aware of any public info that would help them make a better decision. That's how I began assembling the information."

That prior art might exist is not unusual, said several persons who have researched the matter. "The Patent Office has no way to check for prior art," said Lee Ritchey, founder of The Speeding Edge, who cited unrelated instances of patents issued that described the work of others from years and in some cases decades earlier.

Reached by phone, Dennis Silverio, now retired, was surprised that the concept he copatented was even in the news. He told PCD&M that the goal of his work with Leary was to create capacitance. "We designed a voltage and a ground plane for four layers to create a capacitance in the board." Due to persistent shorts, "we had a bitch of a time manufacturing the board," he added. Mupac sold over $100 million worth of wire wrap boards with the Leary construction, Silverio said, adding that Mupac never licensed the technology. (Mupac is now a division of Carlo Gavazzi, a Swiss company.)

The significance of this turn of events cannot be understated. Several persons PCD&M spoke with said that OEMs have held off from designing embedded technologies into their products out of fear of conflicting with existing patents--and engaging in expensive legal battles over their validity.

Ritchey, who gives classes on how to make stackups for planar capacitance without using patented material, believes more such revelations are just ahead. "All it takes is a company that's willing to push it."

Sanmina-SCI stands to lose the most financially, should the USPTO or the courts determine that the purported prior art invalidates the patents issued for ZBC. (At press time, there were no known actions disputing the validity of the Howard patent.) In a May interview with EBN, Dudnikov called Sanmina's licensing program "a competitive edge." Although Sanmina-SCI has not disclosed how much revenue the program generates, licenses are said to cost $1 million, plus additional fees based on the amount of material used, one licensee told PDC&M.

Howard says he no longer benefits financially from his earlier work, having ceded those rights under the agreement with Zycon. "I'm very, very proud of the work that I did," he adds.

Of his revelation, Howard says he "knew it would stun the field. It's truly like letting the genie out of the bottle."