Active solid-state devices (e.g. – transistors – solid-state diode – Housing or package – For high frequency device
Reexamination Certificate
1998-11-19
2001-07-24
Potter, Roy (Department: 2822)
Active solid-state devices (e.g., transistors, solid-state diode
Housing or package
For high frequency device
C257S708000
Reexamination Certificate
active
06265774
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to millimeter wave devices, and, more particularly, to a housing or package for such devices.
BACKGROUND
Millimeter wave devices employ a monolithic microwave integrated circuit device or “MMIC” as an active element and operate at very high frequencies, 45 Gigahertz to 120 Gigahertz and higher. The devices may be configured as amplifiers, oscillators and the like electronic devices to perform functions at those high frequencies akin to those functions accomplished at lower frequencies with more familiar amplifier and oscillator architecture. Such millimeter wave devices find application, as example, in MMW radiometers.
The elements of the device are physically quite small. At a frequency of 120 Ghz, one wavelength measures a mere one-quarter of a centimeter in length or slightly less than one-tenth of an inch. MMW devices thus are physically small in size, and its components, including one or more MMIC's, are much smaller still. Although small in size, unlike lower frequency apparatus, the physical dimension of the MMIC, the associated electronic components and the electrical lead wires are large in respect to the wavelength of the operating frequencies. As a consequence lead wires and the like, by which the components are wired into circuit, and even the body of the component can impact the electromagnetic characteristics of the electronic circuit defined with those elements. Thus the development of a new MMW device and the proof of the device's design is complicated by that impact.
For one, it is necessary to shield the device's components and/or electrical leads. Thus it is not possible to mount all the circuit components and MIMICS on a conventional printed circuit board or place the device in a conventional housing, such as used at lower frequencies. The MMW signals in one part of the circuit must be limited to propagating only to precisely defined routes and must not “jump” that route and propagate in undesired ways to parts of the circuit where they would cause interference. As example, because the wavelength is so short, a small electrical lead from a component or interconnection may serve as a full wave or half-wave antenna, and radiate MMW energy from that lead into open space. Other electrical leads in the circuit could likewise act as a full-wave or half-wave receiving antenna, picking up the foregoing radiation.
As more specific example, in the familiar superheterodyne type receiver, a mixer receives MMW energy from an external source and mixes the received signal with another MMW signal supplied by a local oscillator to produce a lower difference frequency or “beat” signal, as variously termed, a process called down-converting. The downconverted signal is then coupled to an IF amplifier, which amplifies that signal. If, however, through the spurious radiation process described, the MMW signal is also coupled to the output of that mixer by means of a spurious path, the mixer's output contains not only the IF signal, but an interfering MMW signal. The combination of those two signals may overload the succeeding IF amplifier. Then too, the two signals may beat together in the IF amplifier creating still additional interference signals.
To avoid such spurious signals, it is necessary to essentially shield each component from every other component in the MMW device. The MMW device's housing or packaging provides that shielding. Thus, to house the active MMIC semiconductor device and the ancillary capacitors and resistors, the practice has been to use a thick metal plate that has been machined to form precisely defined compartments or hollow cavities, referred to as “mouseholes”, within the plate's thickness or depth, and provide additional small passages for the interconnecting leads to those elements. The mouseholes are small and narrow, consistent with the small physical size of the components.
The active MMIC devices are formed on substrates. And the MMIC devices and the other electrical components are dropped into their respective mousehole, and fastened in place, suitably with an epoxy adhesive. The open side of the housing is covered with a metal lid, closing the mouseholes.
Typically, machining of the housing to shape is accomplished using Electron Discharge Machining apparatus. That apparatus has the ability to maintain very tight dimensional tolerances and produces cavities with sharply defined square corners, which is desirable. However, discharge machining is an expensive process.
As is familiar to designers of microwave and higher frequency devices, the foregoing assembly of MMW elements into the foregoing package on initial assembly of the first or prototype unit rarely, if ever, produces MMW device performance that conforms to the theoretical operation desired. Although the foregoing package solves one problem, it produces others: the cavity mode resonances.
The adverse effects are due to the fact that the dimensions of the formed cavities, the mouseholes, are all on the same order as the wavelength of those frequencies for which the device is intended to function. Recalling that at 120 MHz, the wavelength is only one-tenth inch and that portions of the electrical leads are located within those cavities, the lead portion again act as an antenna or coupling and the energy radiated therefrom “excites” one or more electromagnetic modes within the respective mouseholes.
A rectangular cavity is capable of supporting, that is, resonating in a number of different modes, including a primary TE01 mode and an indefinite number of higher order modes. The higher the number of the mode, the lower the maximum amplitude of the voltage or intensity. Exact mathematical representations of those modes for a given cavity are available in the technical literature. Any excited mode could cause interference and is undesirable and should be minimized or eliminated entirely.
Although foreseeable, because of the complex nature of mode excitation and lead placement and the shape of the mouseholes, the precise cause of the adverse effects are unpredictable. To the present day, neither close attention to detail in fabricating the completed package or refinement in design procedures have been able to eliminate those effects. As a practical matter, the adverse effects come with the territory.
Thus, following assembly of a prototype MMW device, the procedure is to hunt down and destroy those expected uninvited resonances. This is presently accomplished by inserting high frequency absorbent, “lossy”, material, such as Eccosorb material, at strategic locations in the cavities. The Eccosorb material produces an electrical loss to the incident radiation, thereby reducing or minimizing the offending mode or modes.
Although intended to be identical in construction, in the absolute sense each MMW device in a production run differs in physically minute respects from others in the run. An electrical lead from a component in one device may be oriented slightly in position from the corresponding lead in the corresponding component in the next device, creating a small physical difference. However, measured against the wavelength of the frequencies employed, which is, as earlier stated, one-quarter centimeter at 120 GHz, the difference is significant. That difference results in the excitation of a different mode during operation. Thus each device produced in the run must be tested and the unique resonant modes hunted down and destroyed.
Typically the design of a MMW device is confirmed or “proofed” with the production of the first prototype. Often one finds that it is not possible to sufficiently minimize the unwanted resonances in that production prototype. One must refine the design of the mousehole cavities and, essentially, construct a second iteration of a prototype.
Since the components are all permanently fastened in place in the first prototype, and are nearly impossible to remove without damaging the device or the housing, one is left with the prospect of rebuilding the next prototype from scratch. That procedure
Barber Gregory K.
Gage Charles E.
Mitchell Jeffrey B.
Osgood Bruce E.
Sholley Michael D.
Goldman Ronald M.
Potter Roy
TRW Inc.
Yatsko Michael S.
LandOfFree
Millimeter wave adjustable cavity package does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Millimeter wave adjustable cavity package, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Millimeter wave adjustable cavity package will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2475275