This application also is related to, and shares common disclosure with, the following applications:
1. Field of the Invention
The present invention pertains to dual inline memory modules (“DIMMs”) and, more particularly, to a DIMM connector accommodating sideband signals.
2. Description of the Related Art
As the power of individual electronic computing devices has increased, computing systems have become more distributed. Early “personal” computers, although powerful for their time, were suitable for little more than primitive word processing, spreadsheet, and video game applications. More intensive applications, e.g., computer aided design/computer aided manufacturing (“CAD/CAM”) applications were typically hosted on relatively large, more powerful “mainframe” computers. Users invoked applications from time-sharing terminals that served as a conduit for information. However, most of the computational power resided on the host mainframe, where most of the computations were performed.
Stand-alone computing devices eventually evolved from dumb terminals and weak personal computers to powerful personal computers and workstations. As they became more powerful, the computational hours for applications became more distributed. Individual computers eventually became networked, and the networks distributed the computational activities among the network members. Many computations once performed on a mainframe computer, or that were not previously performed, were now performed on networked personal computers. Networks also permitted users to share certain types of computing resources, such as printers and storage.
More powerful computing devices also permitted larger, more complex networks and other computing systems. Small local area networks (“LANs”) became wide area networks (“WANs”). Recently, networks have evolved to produce system or storage area networks (“SANs”). Some of these networks are public, e.g., the Internet. Some may be characterized as “enterprise computing systems” because, although very large, they restrict access to members of a single enterprise or other people they may authorize. Some enterprise computing systems are referred to as “intranets” because they employ the same communication protocols as the Internet.
Most types of electronic and computing systems comprise many different devices that electronically communicate with each other over one or more buses. Exemplary types of devices include, but are not limited to, processors (e.g., microprocessors, digital signal processors, and micro-controllers), memory devices (e.g., hard disk drives, floppy disk drives, and optical disk drives), and peripheral devices (e.g., keyboards, monitors, mice). When electrically connected to a bus, these types of devices, as well as others not listed, are all sometimes generically referred to as “bus devices.” In
For instance, a computer typically includes one or more printed circuit boards having multiple integrated circuit components (or “bus devices”) and connectors mounted to them. The components and connectors are interconnected by and communicate with each other over trace etched into the board. The boards are interconnected by plugging one or more of the boards into another board intended for this purpose. A first component on a board communicates with a second component on the same board over the traces etched onto the board. The first component communicates with a component on another board through the connectors by which the two boards are plugged into the third board intended for that purpose. Thus, both the traces on the boards and the connectors between the boards are a part of the bus. Again referring to
DIMMs are one common type of memory component. A DIMM is simply a printed circuit board (“PCB”) on which a number of memory chips are mounted. The memory chips are usually some form of “volatile” memory, which means that the data stored in them will be lost if power supplied to the chips is interrupted. Many DIMMs therefore include battery packs mounted to them. The battery packs house batteries that provide “backup” power to the DIMM if the primary source of power is interrupted for some reason. The backup power supplied by the battery packs then provides an opportunity to save the data if primary power is restored in time.
DIMMs are widely used in mass storage devices such as redundant arrays of inexpensive disks (“RAIDs”). DIMMs are sometimes used in a RAID controller to implement a type of memory known as “cache,” and DIMMs used in this context are therefore sometimes referred to as “cache cards.” RAID controllers will only allow posted writes to occur when it can guarantee that the batteries can sustain backup for a minimum period of time agreed upon by the user in the event of a power outage.
Still another problem arises from the pin-out on connectors. Using a custom cache card limits the users options for upgrading and replacement of the memory module. By adding battery packs to a standard DIMM module allows the user to use a larger standard DIMM modules if more read ahead caching is required or use the battery backed DIMM if posted write data is required. The majority of the pins defined for DIMM connectors are required for normal operation. Side band signals used for battery status and control need to be shared with low priority DIMM signals or use special signals defined as no-connect. Typically, the industry has used custom non-industry standard cache card outlines and connectors, industry standard cache card without batteries, or industry standard cache card with batteries on main board. Alternatively, the industry has embedded the cache card on the main RAID controller. Each of these approaches has its disadvantages.
The present invention is directed to resolving, or at least reducing, one or all of the problems mentioned above.
In one aspect, the invention is a DIMM connector, including pins for at least one of battery status or control for transmitting side band signals for a battery backed cache.
The invention may be understood by reference to the following description taken in conjunction with the accompanying drawings, in which like reference numerals identify like elements, and in which:
While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof have been shown by way of example in the drawings and are herein described in detail. It should be understood, however, that the description herein of specific embodiments is not intended to limit the invention to the particular forms disclosed, but on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the appended claims.
Illustrative embodiments of the invention are described below. In the interest of clarity, not all features of an actual implementation are described in this specification. It will of course be appreciated that in the development of any such actual embodiment, numerous implementation-specific decisions must be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which will vary from one implementation to another. Moreover, it will be appreciated that such a development effort, even if complex and time-consuming, would be a routine undertaking for those of ordinary skill in the art having the benefit of this disclosure.
Turning now to the drawings,
Referring now to
The Intelligent HBA 200 is intended to be mounted in a server (not shown in
Note that not all the features of the base adapter 205 are shown for the sake of clarity. As those in the art having the benefit of this disclosure will appreciate, such a base adapter 205 will include a number of implementation specific details not germane to the present invention. Such details, because they are routine and well known in the art, have been omitted from the drawing and the discussion herein in order not to obscure the invention.
As mentioned,
The topology 500 includes a SCSI bus 510, a breakout node 515 on the SCSI bus 510; an external SCSI connector 520, an internal SCSI connector 525, a SCSI adapter 530, and a terminator 535. In an actual, physical embodiment, each of the external SCSI connector 520, internal SCSI connector 525, SCSI adapter 530, and terminator 535 could be either a pin of or a pad for a chip. The breakout node 515 could be a via in a printed circuit board (“PCB”) and the SCSI bus 510 could be traces on the PCB (not shown). The external SCSI connector 520 is positioned on the SCSI bus 510 at a first point defined by a first propagation delay td1. The internal SCSI connector 525 is also positioned on the SCSI bus 510, but at a second point defined by a second propagation delay td2. The first and second propagation delays td1, td2 are substantially equal. The SCSI adapter 530 and the terminator 535 electrically tap the breakout node 515.
More technically, the signals traveling on the SCSI bus 510 would ordinarily be expected to travel at the speed of light, but for a number of factors well known in the art. For instance, a signal's propagation through the conductive material that comprises the SCSI bus 510 introduces delay. However, greater delay is introduced by, for example, the electrical loading introduced of SCSI devices (not shown) on the SCSI bus 510, the routing of wires and traces that comprise the SCSI bus 510, and the particular implementation of the connectors, e.g., the external connector 520 or the internal connector 525.
Note that the propagation delay may vary at different portions of the SCSI bus 510. In the context of the invention, the important consideration in determining the first and second points at which the external and internal connectors 520, 525 are located is the propagation delay from the breakout node 515. Thus, the distances d1, d2 at which the external and internal connectors 520, 525 are located is immaterial except to the extent they provide an upper boundary affecting the propagation delays td1, td2. Note that, in some embodiments, the SCSI bus 510 may have a constant propagation delay per unit length such that the distances d1, d2 may be equal because they produce equal delays td1, td2.
In one particular embodiment, the topology 500 is implemented in accordance with the Ultra 2 or Ultra 3 SCSI specification. Generally speaking, in this implementation, it is preferred that the distances d1, d2 be less than 3.5″ and the delays td1, td2 should be less than 525 ps to inhibit significant signal degradation. It is also generally preferred for the same reason that: (1) the distance of the SCSI adapter 530 from the breakout node 515 should be less that 1.5″ and the propagation delay less than 225 ps; and (2) the distance between the terminator 535 and the breakout node 515 should be less than 6.0″ and the propagation delay less than 900 ps.
Note that the propagation delays td1, td2 are “substantially” equal. As will be appreciated by those in the art having the benefit of this disclosure, there are several limitations on the precision with which the propagation delays td1, td2 can be implemented. For instance, variations in bus device embodiments might introduce variation in electrical loading, which affects propagation delay. Similarly, design constraints might limit flexibility in bus layout so that a designer does not have the latitude to achieve precisely equal propagation delays. The propagation delays td1, td2 would ideally be precisely equal, because a difference will degrade performance proportionally to the amount of the difference. However, in various embodiments, some difference can be tolerated in light of variations introduced in the design, manufacturing, and assembly processes.
Returning to
Returning to
The configurability of the Intelligent HBA 200 can be illustrated by considering the implementation of
Both the daughtercard 300 in
The Intelligent HBA 200 is configurable to provide either an otherwise conventional RAID controller functionality or a RAID controller permitting direct attached storage to be shared. The Intelligent HBA 200 is configurable by switching out various implementations of the daughtercard 210. This is done by segregating various “interfacing” capabilities off the base adapter 205 onto the daughtercard 210 so that different implementations of the daughtercard 210 can be used to configure the Intelligent HBA 200 for different uses. Thus, the daughtercard 210 can be used to “modify” a protocol in use on the Intelligent HBA 200.
For instance,
The Intelligent HBA 200b in
Thus, segregating this “interfacing” capability normally found on the base adapter 205 onto the removable, replaceable daughtercard 210, the Intelligent HBA 200 provides numerous advantageous characteristics. This approach can also be used to add upgradeability to a base RAID controller; upgrade from two-channel SCSI to four-channel SCSI; upgrade from two-channel SCSI to two-channel SCSI with Fibre channel; and/or one-channel and two-channel Fibre on a network interface card (“NIC”). This ability to upgrade also has the salutary effect of lengthening the life of the Intelligent HBA 200 and reducing it's cost to upgrade as the technology evolves. This also means that the Intelligent HBA 200 is configurable in the field depending upon the particular computing system being implemented.
The cache card 215 of
The battery packs 410a, 410b house eight batteries 415 each. The batteries 415 power a cache comprising multiple memory devices 417, as best shown in
The number of batteries 415 and battery packs 410a, 410b will be implementation specific. Two battery packs 410a, 410b were chosen in the illustrated embodiment so that each memory module in the cache may be individually powered by a single pack 410a, 410b of batteries 415. In this particular implementation, the battery packs 410a, 410b are redundant, although this is not necessary to the practice of the invention. Thus, in the event one of the battery packs 410a, 410b fails, the other may support the entire load. If both battery packs 410a, 410b are operational, then they will share the load.
Referring now to
The second part 430 includes a lip 435 and a flexible tab 438 terminating in a hook 440. The casing 420 is assembled with the PCB 442 by engaging the lip 435 with one edge 445 of the PCB 442 as shown in
When the flexible tab 438 is inserted into the slot, the hook 440 passes all the way through the PCB 442 and engages the surface 455 opposite the side 460 to which the battery pack 410a, 410b is mounted. The length of the flexible tab 438 should be long enough so that this engagement secures the battery pack 410a, 410b to the PCB 442 snugly in order to facilitate the electrical contact between the battery pack 410a, 410b and the PCB 442. Note that the assembly of the battery pack 410a, 410b to the PCB 442 establishes the electrical contact through which the batteries 415 power the DIMMs. To replace the batteries 415, the battery pack 410a, 410b can be disassembled from the PCB 442 by manually pushing the hook 440 back toward the edge 450 and pushed back through the slot.
Thus, the lip 435 and the flexible tab 438 comprise, by way of example and illustration, means for engaging and securing (through a spring force), respectively, the battery pack 410a, 410b to the PCB 442 of the cache card 215. However, the invention is not so limited. Alternative embodiments may employ alternative, equivalent structures performing this same function. Similarly, the pegs 448 comprise, again by way of example and illustration, but one means for preventing planar movement of the battery pack 410a, 410b relative to the PCB 442 when the battery pack 410a, 410b is secured to the PCB 442. Alternative embodiments may also employ alternative, equivalent structures performing this function. Note, however, that the structure of the lip 435, flexible tab 438, and pegs 448, and any equivalent structure that may be employed in alternative embodiments, is such that they permit the assembling of the battery pack 410a, 410b to the PCB 442 without the use of tools while rigidly securing the battery packs 410a, 410b to the PCB 442.
Referring now to
Thus, the casing 420 includes two retention features that enable the battery packs 410a, 410b to be assembled to a memory module without use of a tool. These two features specifically are the lip 435 that, during assembly, grabs the bottom of the memory module and the plastic hook 440 that flexes during the installation process and “snaps” through a hole in the DIMM memory module, grabbing the underside of the DIMM memory module. These two features ensure that the battery packs 410a, 410b remains secure during any transportation or shipping of the memory module.
The design of the cache card 215 allows it to not only be readily assembled with and removed from the daughtercard 210, but to do so without any tools. Thus, this aspect of the present invention allows new batteries to be replaced on the existing cache card at less than 3% of the cost of a new cache card. Still further, the battery backed DIMM can be replaced with an industry standard DIMM and the battery pack and cache card fit within the envelope specified by PCI specifications and passes the appropriate levels for shock and vibration testing.
The cache card 215 also includes a DIMM connector 415, including pins for transmitting sideband signals for the battery backed cache. The Intelligent HBA 200 of
Conventional DIMM connectors built in accordance with industry standard specifications do not accommodate sideband signals regarding battery status or control. DIMM connectors accommodating such sideband signals employ custom pinouts, which then are incompatible with industry standard specifications. The present DIMM connector overcomes this conundrum by redefining pins that ordinarily are not connected or used for some purpose not presently germane to a different function accommodating the sideband signals) regarding battery status or control. Thus, the present DIMM connector can accommodate such sideband signals using an pinout compatible with industry standards. In the illustrated embodiment, four pins have been reassigned functions from a “no connect” status or other use—pin 61 (PRESENT_/NC), pin 80 (NVRW_/NC), pin 81 (NVLATCH/WP), pin 164 (PWR_GOOD/NC), and pin 165 (NVCS_/SA0). Thus, in the illustrated embodiment, the sideband signals are indicative of indicate battery status, battery life, or battery control.
The PRESENT_signal is communicated on pin 61, which is a “no connect” in the industry standard pinout. The PRESENT_signal, is used to indicate that the cache card 215 is being used with a battery backed memory module, e.g., the cache card 215, in an Intelligent HBA 200 instead of in, e.g., a conventional RAID controller. To this end, the base adapter 205 includes a special pin (not shown) that, when the cache card 215 is plugged into the DIMM connector 415 grounds the PRESENT_signal which is high otherwise. That is, the PRESENT_signal is high unless the cache card 215 is used in an Intelligent HBA 200. Thus the PRESENT_signal on the pin 61 enables some functions of the illustrated embodiment not useful in conventional memory subsystems, such as a server. The PRESENT_signal can consequently be omitted from some alternative embodiments.
The non-volatile random access memory (“NVRAM”) signals NVRW_, NVCS_, and NVLATCH convey selected information about the NVRAM, i.e., the cache. The NVRW_, NVCS_, and NVLATCH signals are communicated on pins 80, 165, 81, respectively, which are a no connect (“NC”), communicate a serial address bit (“SA0”), and communicate write protect (“WP”) signal, respectively, in an industry standard pinout. More particularly:
The power good signal PWR_GOOD is communicated on the pin 164, which is a no connect in an industry standard DIMM connector. This signal is driven by the battery backed cache 430 and received by the base adapter 205 to indicate when the critical voltage level of the VCC power rail has been crossed. The power good signal will drop low immediately when the VCC rail falls below 2.95 V to offer an early warning to the memory controller that the power rail is dropping. The memory controller will use this signal to stop memory activity and place the SDRAM of the cache 430 into self refresh mode. The power good signal will rise high after about 200 mS after the VCC rail increases above 2.95V to allow circuits depending on the VCC power rail to stabilize before exiting the reset state.
Thus, the cache card 215 includes a modified DIMM connector pin-out to support sideband signals for a battery backed cache. Such a custom DIMM pin out allows the battery-backed cache to be used in an industry standard DIMM socket. It also allows industry standard DIMM to be used in a battery-backed cache socket. Still further, users will have a wider variety of cache modules to select for use.
The cache card 215 also includes a variety of features leading to improved power management. The Intelligent HBA 200 includes on the cache card 215 a micro-controller 850. In the illustrated embodiment, the micro-controller is an 8-bit micro-controller commercially available from Microchip Technology Incorporated, USA under the designation PIC12C67X, but any suitable micro-controller known to the art may be used. The micro-controller 850 is used to implement a battery fuel gauge, primarily in software, but which also includes a charging circuit for charging at least one battery and a decrementor circuit for counting the amount of time system power is removed from the battery.
The cache card 215 of the Intelligent HBA 200 in the illustrated embodiment utilizes 3.0–3.6V from the system. It internally generates 8V, 5V, and 3V_REF for its embedded circuitry. During normal operation the batteries 415 will fast charge for 1 minute during each hour for conditioning. Posted-write caching will only be enabled when both battery packs 410a, 410b are good. As a result, both battery packs 410a, 410b should be functional at the start of a power failure. Tables 3–4 provide additional information regarding battery back-up life and battery shelf life, respectively.
In particular, the Intelligent HBA 200 includes a sophisticated power management scheme. The micro-controller 850 detects battery status from two onboard A/D converters (not shown) with 8-bit accuracy. The micro-controller 850 forces a fast charge for one minute during each hour to condition the battery packs 410a, 410b and tracks the battery capacity. The micro-controller 850 also controls battery power enable, and reports battery information across an Inter-IC (“I2C”) bus. The I2C bus is a well-known bus design typically used to connect integrated circuits (“ICs”). An I2C is a multi-master bus, i.e., multiple chips can be connected to the same bus and each one can act as a master by initiating a data transfer.
The micro-controller/I2C memory map is set forth in Table 5.
The health bits in the charge status register (discussed further below) are used to indicate why the packs are not good. If an open condition occurred, an amber status LED (not shown) for that particular pack will blink and the associated bit in the status register will be set. If a short condition occurred, the amber status LED for that particular pack will remain solid. If either a short or open condition exists, the capacity register and good bit in the status register (discussed further below) will be cleared to 0.
The micro-controller 850 voltage threshold determination in the illustrated embodiment will depend on a number of factors. More particularly:
The cache card 215, in the illustrated embodiment, also includes a “fuel gauge” that extrapolates lost battery capacity based on elapsed time during the loss of system power. More particularly, the cache card 215 tracks the elapsed time during the loss of system power to the battery 415. As will be appreciated by those skilled in the art having the benefit of this disclosure, system power can be used to charge and maintain the batteries 415 at full capacity. However, the loss of system power will result in the loss of capacity as the batteries 415 will discharge through a number of physical phenomena such as leakage and self-discharge. The rates of discharge from these phenomena can be estimated. The cache card 215 then extrapolates from the elapsed time the loss in capacity during the elapsed time using such estimates. In the illustrated embodiment, estimates are formulated on a worst-case scenario for a variable load. The fuel gauge will therefore indicate that the batteries 415 have at least the indicated capacity and the batteries 415 will therefore typically have a higher than indicated capacity. However, this is not necessary to the practice of this aspect of the invention. For instance, the estimate may be formulated assuming a fixed load.
More particularly, this fuel gauge comprises a resettable decrementor based upon the amount of time in backup or non-backup (i.e., loss of battery capacity due to leakage and self-discharge). When the batteries 415 are first attached to the circuit 800, shown in
When the system power is removed, the batteries 415 will begin to drain. The fuel gauge decrementor circuit (discussed further below) will operate at a low power state while it counts the amount of time that the system power is removed. In the illustrated embodiment, this is performed by a power reset chip (not shown) and the low power micro-controller 850, which operates from a 32 KHz clock 855. In one particular implementation, the power reset chip is a X24C105 integrated circuit commercially available from Xicor Corporation. The fuel gauge decrementor circuit will continue to track the amount of time in backup until system power is returned to the charging circuit 810.
The fuel gauge decrementor circuit will, in the illustrated embodiment, know if the batteries 415 were enabled to sustain a load during this time. If the load on the batteries 415 is not engaged, the fuel gauge decrementor circuit will equate the final count value to the lost capacity based upon the amount of power consumed to sustain the counter circuit, plus any power loss due to extra components and self discharge within the battery pack. If the load on the batteries 415 is engaged (such as the cache being placed in a low power state to back up data), the fuel gauge decrementor circuit will equate the final count value to the lost capacity based upon the amount of power consumed to sustain the fuel gauge decrementor circuit, plus any power loss due to extra components, self discharge by the battery 415, and the maximum amount of current expected to be consumed by the load.
The fuel gauge decrementor circuit at this point should contain a new capacity indicating that the battery is within the range 0%–100% charge. Since the system power is enabled, the batteries 415 will begin to charge. At this time, in the illustrated embodiment, the fuel gauge decrementor circuit does not increment, although it may do so in alternative embodiments. When the charging circuit 810 reaches charge termination, a signal will indicate to the decrementor circuit that the batteries 415 are at 100% capacity at which time the fuel gauge decrementor circuit will reset the capacity. If the system power is lost prior to the charging circuit reaching charge termination, the fuel gauge decrementor circuit will decrement the existing capacity without resetting the capacity.
The “fuel gauge” is implemented partially in software executed by the micro-controller 850 and 4 hardware, timer registers (not shown) in the micro-controller 850. As noted above, when system power is lost, the micro-controller 850 shuts down all peripherals (not shown) to operate at a low current. The timer registers in the micro-controller 850 are initialized when power is lost. These registers are used to track the amount of time the server has been powered down. An external 32 KHz crystal 855 with a 16-bit timer will overflow at a 16 second rate. After the overflow occurs, the micro-controller 850 will decrement the necessary timers and execute the sleep instruction. When server power returns, the timer registers are used to calculate the amount of capacity lost. The capacity lost will be calculated based upon whether the memory was in backup mode or self discharge mode and whether the DIMM capacity is 32/64 MB or 128 MB.
In the illustrated embodiment, the following registers are used:
This concludes the detailed description. Note that some portions of the present invention might be implemented in software, and hence described in terms of a software implemented process involving symbolic representations of operations on data bits within a computer memory. These descriptions and representations are the means used by those in the art to most effectively convey the substance of their work to others skilled in the art. The process and operation require physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical, magnetic, or optical signals capable of being stored, transferred, combined, compared, and otherwise manipulated. It has proven convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.
It should be borne in mind, however, that all of these and similar terms are to be associated with the appropriate physical quantities and are merely convenient labels applied to these quantifies. Unless specifically stated or otherwise as may be apparent, throughout the present disclosure, these descriptions refer to the action and processes of an electronic device, that manipulates and transforms data represented as physical (electronic, magnetic, or optical) quantities within some electronic device's storage into other data similarly represented as physical quantities within the storage, or in transmission or display devices. Exemplary of the terms denoting such a description are, without limitation, the terms “processing,” “computing,” “calculating,” “determining,” “displaying,” and the like.
The particular embodiments disclosed above are illustrative only, as the invention may be modified and practiced in different but equivalent manners apparent to those skilled in the art having the benefit of the teachings herein. For instance, the cache card 215 might be employed in a laptop computer rather than a RAID controller. This would provide the advantage of being able to port the state of one laptop computer to a second laptop computer provided both employed a sleep state. Furthermore, no limitations are intended to the details of construction or design herein shown, other than as described in the claims below. It is therefore evident that the particular embodiments disclosed above may be altered or modified and all such variations are considered within the scope and spirit of the invention. Accordingly, the protection sought herein is as set forth in the claims below.
We hereby claim the earlier effective filing date of U.S. Provisional Application No. 60/231,384 filed Sep. 8, 2000.
Number | Name | Date | Kind |
---|---|---|---|
4595880 | Patil | Jun 1986 | A |
4737771 | Coash | Apr 1988 | A |
5315228 | Hess et al. | May 1994 | A |
5528460 | Byrd | Jun 1996 | A |
5726859 | Khadem et al. | Mar 1998 | A |
5883497 | Turnbull | Mar 1999 | A |
5962159 | Satou et al. | Oct 1999 | A |
6098140 | Pecone et al. | Aug 2000 | A |
6160718 | Vakilian | Dec 2000 | A |
6243315 | Goodman | Jun 2001 | B1 |
6252511 | Mondshine et al. | Jun 2001 | B1 |
6307377 | Cummings et al. | Oct 2001 | B1 |
6469474 | Bunker | Oct 2002 | B1 |
6492058 | Watanabe et al. | Dec 2002 | B1 |
6608564 | Post et al. | Aug 2003 | B1 |
6747878 | Hipp et al. | Jun 2004 | B1 |
6757748 | Hipp | Jun 2004 | B1 |
6871300 | Irving | Mar 2005 | B1 |
20020056018 | Schumacher et al. | May 2002 | A1 |
20020059492 | Sabotta et al. | May 2002 | A1 |
20020080541 | Bunker et al. | Jun 2002 | A1 |
20030081387 | Schulz | May 2003 | A1 |
Number | Date | Country | |
---|---|---|---|
20020058444 A1 | May 2002 | US |
Number | Date | Country | |
---|---|---|---|
60231384 | Sep 2000 | US |